FeverBee – Community Consultancy

Community Platform Overage Fees Are Becoming A Big Problem

Richard Millington — Mon, 20 Jul 2026 11:07:45 +0000

Have You Been Hit With An Unexpected Invoice? What Should You Do?

Over the past six months or so, several organizations have contacted me about billing errors from one specific platform vendor. A handful are being charged $100k+ for a single quarter in additional, unexpected bot traffic.

This wasn’t the result of a new campaign, a favorable change in algorithms, or an influx of new customers. It was sharp spikes that arrived without warning and vanished just as fast.

In recent months I’ve reviewed the data of three clients and came to the same conclusion. The traffic which presents itself as human most definitely isn’t.

But this presents a problem. What should you do in this scenario?

Bot Traffic Now Exceeds Human Traffic

Over the past year or so, most communities reached a point where bot traffic exceeded human traffic.

I analysed a few of our clients and discovered some have seen bot traffic reach 90%+ of all traffic to their community. As best as I can tell, the majority sit somewhere between 50% and 80%. That’s a staggering increase in usage for the vast majority of communities.

And it comes at the exact time that human visitors are declining because of the collapse of search traffic.

For major sites like Reddit and Stack Overflow, the problem became so extreme they’ve struck paid licensing deals with the AI companies, and Wikipedia’s parent now sells a paid enterprise API for bulk access.

Communities don’t have that leverage. And they don’t want to block everything anyway, because they still want the useful crawlers: Google, so members can find them in search, and the AI engines that cite their sources and send readers back.

What they don’t want is to pay to be scraped by the extractive crawlers that take everything and return nothing.

This is the catch with blocking. It isn’t free. Block the wrong crawler and you lose your search visibility or your place in AI answers. So the goal isn’t to shut bots out, it’s to stop paying for the ones that give you nothing back.

Part of the challenge is that building and running crawlers has become cheaper and easier than ever. On top of that, AI agents that fetch pages live, on demand, have multiplied the number of requests hitting your community. User-triggered crawling grew more than 15x last year alone.

Part of it is just the incentive to grab this data before more organisations lock it down.

Two things are happening here.

The baseline is rising permanently. Every new model generation re-crawls the archives, and agent traffic climbs as more people use AI to answer questions, so the floor under your usage keeps moving up and won’t come back down.
Spikes are becoming a problem. They’re the thing clients keep reporting, and the thing that actually blows the budget. These are sharp, sudden surges of traffic pretending to be human, arriving without warning and pushing them over their usage allowance in a matter of days.

What Are These Bots Exactly?

A bot is essentially just software that requests pages in the same way a person does – but there’s no human reading the other end. For the best part of two decades, they were a good thing. Googlebot would crawl your community, index valuable content, and then surface them when people requested relevant information.

Over the past few years, however, training crawlers like GPTBot, ClaudeBot, and CCBot pulls entire archives in bulk to feed their language models. Worse yet, retrieval and agent bots will fetch pages in real time when someone asks a question, rather than just sending them a link to the potential answer. The numbers vary, but AI might be retrieving 30+ pages to effectively answer a single query.

The Real Scourge Isn’t The Good Bots

But the real scourge here isn’t the clearly labelled LLMs or search bots, although they do make a lot of data requests. The real challenge is the growing share of bots, sometimes as high as 30%, that pretend to be humans (or other bots).

This means they use generic browser identifiers and datacenter connections to look like ordinary visitors.

Their goal is to consume knowledge and information without honouring blocks or, in some cases, being forced to pay Reddit, Stack Overflow, or Wikipedia. They don’t honour robots.txt, because that would defeat their entire objective.

And this is where the real problem begins.

Because if you are on a usage-based platform package, as most of you are, and this dark bot traffic causes you to go over your agreed usage allowance, you may get hit with significant overage fees.

There’s a nasty irony here, and it drives everything below. The bots that are honest about being bots are the easy ones to deal with. It’s the ones that lie about being human that end up on your invoice.

Be fair to the vendors here. Most of them aren’t villains and most want to help. But they get charged for this traffic too, and that charge is the same whether the visitor is a human or a bot. When a vendor passes an overage on to you, it’s often a bill that landed on them first.

Historically, most smart vendors, when confronted with obvious proof that traffic is non-human, waive the overage charge for the benefit of the relationship. The vendor, in essence, eats the cost and tries to do a better job filtering and blocking it in future. They do this for the benefit of the client relationship (or, ideally, because of carefully negotiated contractual terms, more on that later).

But not all vendors are alike.

Today a growing number of organisations are being threatened with legal action (or even closure of their community) if they refuse to pay the overage fee. And this creates a big problem, because it messes with budgets and calls the community’s viability into question.

What Should You Do If You’re Hit With An Overage Charge?

The most important thing to do when you get an overage request is not to rely on the platform vendor’s analytics or tools like Google Analytics. Instead, you need to request access to the raw server logs. This is a reasonable request to make.

The reason is simple. Your analytics and your server logs show very different things. Google Analytics and similar tools run on JavaScript tags, and most crawlers don’t execute JavaScript, so they’re invisible to Google Analytics. Your server logs, however, or your CDN logs, see every single request. Every time a bot requests a page or data, it shows up in the logs.

Separate Out The Expected Bot Traffic

Now within this, there are some things you can count exactly.

GPTBot, ClaudeBot, and CCBot all publish their user agent string, so you can filter the logs for them and have a hard floor.

Most importantly, you can verify them against the published IP ranges.

OpenAI, Anthropic, and Google all publish those, or you can use reverse DNS. That separates the real bots from the ones trying to pretend to be them.

Create A Range For Dark Bot Traffic

The hard part is the undeclared ones. These you infer from behaviour rather than counting them, but there are some very specific tells to look out for.

Request rate. A single IP pulling hundreds of pages a minute isn’t a person.
Hitting deep archive threads no human would navigate to.
Fetching only HTML and skipping the CSS and images a browser would load.
Requests coming from datacenter IP ranges (AWS, GCP, Azure) instead of residential ISPs.

So, for example, if you see a datacenter hammering your old forum posts at machine speed, that’s a bot, even if it claims to be Safari. This gives you an estimate, not a clean number.

An even better route, when available, is to have a CDN (content delivery network) or WAF (web application firewall) classify the traffic for you.

Tools like Cloudflare and Fastly are already pretty good at labelling verified bots, AI crawlers, and traffic that is likely automated, and giving you a breakdown. It’s probably the most credible, independent number you can use.

Then you can compare the metering number the vendor charges you against your human analytics. For example, if your invoice is built on one million sessions and Google Analytics shows 400,000, that gap is your automated traffic. It’s not exact, but it gives you a rough estimate to work with.

Three Terms In The Contract Are Critical Here

In the contract (or Master Services Agreement) there are three things to pay attention to.

What is the usage allowance? This is typically how many page views and API calls are allowed per month or quarter. This number is often in the low millions to begin with, but can rise significantly for larger organisations.
How is usage defined and calculated? What counts as a view? Many contracts define a view in a way that is meant to exclude bots, often with a phrase along these lines: Traffic not detected as originating from a browser is not counted as page views.
What happens when that usage is exceeded? Do you simply move up to the next tier indefinitely? Or do you get billed separately (often excessively) for the overage?

Of the three, it’s the definition of a ‘view’ that matters most, and it’s worth understanding why it doesn’t protect you as well as it looks.

A definition like the one above sounds reassuring. Bots are excluded, so you only pay for humans. But look at what it actually does. The honest, declared crawlers, GPTBot and the like, announce themselves as bots rather than browsers, so they’re easy to detect and get excluded.

The disguised ones, the dark bot traffic wearing a browser identity, and the AI agents like ChatGPT’s and Perplexity’s browsers that literally are browsers rendering the page, are all detected as browsers. So they count.

In other words, the standard definition protects you from the bots that tell the truth and bills you for the bots that lie. Which is exactly backwards, because the lying ones are the fastest-growing share of the problem.

The definition also puts the burden of proof on you. By this wording, it’s on the client to demonstrate a spike is dark bot traffic and not real humans. And that usually requires the server logs. If the vendor won’t give you access to them, you can’t make your case.

The vendor also has leverage. If you don’t pay your bills, they can threaten to close your community down.

Negotiate These Conditions At Contract Signing And Renewal

All of which means the critical thing to do at renewal is to negotiate all three of the terms above.

Before the specifics, one shift matters more than any single clause: what counts as proof. You will almost never be able to demonstrate, request by request, that a spike was automated. What you can produce is an independent CDN or WAF classification. So the real win at renewal is getting the vendor to accept that classification as sufficient evidence. That moves you off an impossible standard, proving every request was a bot, and onto a fair one, where the independent tooling makes the call.

Negotiate what counts as a view. It’s not enough to rely on ‘non-browser’ traffic when the worst bots are pretending to be browsers. You need a fair way to determine whether a spike or increase is bots posing as humans, and to have that traffic excluded. Language worth pushing for looks something like this:

Page views shall exclude any traffic the Customer can reasonably demonstrate, through server logs or CDN classification, to originate from automated or non-human sources, whether or not that traffic presents as a browser. Where a sudden increase in traffic is shown to be predominantly automated, it shall be excluded from usage calculations and shall not, by itself, trigger overage fees.

Secure timely access to server logs. On request, you should be able to obtain the raw server logs within a defined window, say [x] business days. Without this you can’t prove a spike is automated, and the definition above is worthless to you.
Agree what happens when you exceed the allowance. Pin down whether you simply move up a tier or whether you’re billed per unit at a punitive rate. Cap the overage rate, and build in a waiver or review process for spikes that are shown to be automated rather than human.

Summary

Overage invoices from community platforms are rarely a billing error. They’re the result of bot traffic, most of it now non-human, and increasingly of dark bots that disguise themselves as human visitors to get past blocks and avoid paying for what they take.

If you’re hit with one, don’t argue from the vendor’s dashboard or Google Analytics. Get the raw server logs, separate the declared bots you can count from the disguised ones you have to infer, and use a CDN or WAF classification as your independent number.

Then fix it in the contract. Negotiate the usage allowance, the definition of a view, and what happens when you exceed it. These are the three terms that decide who pays for the bots.

Right now, the standard ‘non-browser’ definition has you pay the cost; renewal should be the time to change that.

The post Community Platform Overage Fees Are Becoming A Big Problem first appeared on FeverBee - Community Consultancy.

Should You Build Your Community On Gainsight CC?

Richard Millington — Wed, 15 Jul 2026 14:15:00 +0000

Gainsight Customer Communities, which many people call inSided, is the one platform in this series built to run your community as a customer success engine rather than a marketing channel. That single decision shapes everything it does well, and everything it doesn’t.

I scored it on the same nine capabilities I use for every platform in this series, so you can put it side by side with Higher Logic Vanilla, Khoros, and Discourse. It came out as a real all-rounder. Four scores at the top, a solid middle, and two weak spots it shares with Vanilla.

But the most interesting thing about Gainsight CC isn’t any single score. It’s that being part of Gainsight cuts both ways, and whether that helps or hurts depends entirely on who you are.

How Gainsight CC Got Here

It started in Amsterdam as inSided, and for years it only ever built communities. It served both sides of the market: B2B SaaS, but also big consumer brands like T-Mobile and Sonos. It grew slowly across Europe first, then expanded internationally. It was never a general forum that bolted on support features later. It was built from the start for companies running real customer communities, and you can feel that focus all the way through the product.

Gainsight bought inSided in 2021, and the reason tells you what the platform is for. The argument, from Nick Mehta, who was Gainsight’s CEO at the time and has since handed over to Chuck Ganapathi, was that net revenue retention is the single biggest driver of a software company’s value, and that the best way to scale retention is to bring customers into the success process through a community. So inSided didn’t get treated as a side product. It slotted in next to Gainsight’s customer success and product tools as the community layer of the whole journey.

The bigger signal is the team. In 2025 Gainsight acquired Erica Kuhl Consulting and a moderation tooling company called ModerateKit, and put Erica Kuhl in charge of the combined community, learning, and in-app unit. Erica built Salesforce’s Trailblazer Community into one of the largest customer communities on the planet over seventeen years. And as Khoros has shrunk, a lot of the people who built the best Khoros communities have ended up at Gainsight. For a platform you commit to for years, that matters as much as any single feature.

What Gainsight Does Well

Four capabilities score an 8: AI, moderation, ideation, and content.

The AI suite is the broadest in this series. Not one clever feature, but the widest spread: Write with AI for drafting, community recaps and summaries, and four working agents underneath for moderation, search, translation, and support answers. The translation agent is the one I’d point to, because a community that answers and reads across languages without you running it twice is still rare. The breadth is clearly ahead of the market. The track record is not, because a lot of it is new.

Moderation sits joint top with Khoros and Discourse. An AI agent auto-classifies posts as they come in, so obvious spam and abuse gets caught before a human looks. Profile-level moderation removes the spam account, not just the spam post, which matters because spammers come straight back with the same profile. What it lacks is Discourse’s trust-level auto-promotion, so it doesn’t quietly reduce your moderation load as members earn trust, and there’s still no thread merge or split.

Ideation is the best of any community platform here, and the reason is specific. On the Product Feedback dashboard you can filter ideas by company and see the revenue behind a request, not just the vote count. For a B2B SaaS company, where ten votes from your three biggest accounts matter more than fifty from free users, that’s the difference. The catch: the account value needs your CRM connected (Salesforce in the instance I checked, though HubSpot and others are supported), and it still isn’t a dedicated product-feedback tool like Pendo or Productboard.

Content and knowledge is where it pulls just ahead of Vanilla. A proper knowledge base, smart search, content rating, and a product updates hub. The thing that moves it ahead is embedded education. Every platform in this series lost a point on content for the same reason: none of them did courses. Through the Skilljar integration Gainsight shipped this year, courses, certifications, and learning paths now sit inside the community, with one login and one search across courses, articles, and discussions. You can see it live on Cognite and Pipefy.

The Solid Middle And The Weak Spots

Three scores sit in the middle. Discussions and Q&A is a 7: five content types, a real Q&A format with best-answer marking, a good editor, but not the modern-forum polish Discourse has. Analytics is a 7: dedicated dashboards for content, users, Q&A, audience, and product feedback, most needing no integration, but no build-your-own dashboard the way Vanilla gives you. Gamification is a 6: the full points, badges, and ranks, and it works, it just isn’t a reason to choose the platform.

The two low scores are the same two Vanilla has. Mobile is a 3, because there’s no native app, only responsive web. Events is a 4: registration, RSVPs, and attendee management are there, but no streaming, ticketing, or speaker management. Almost nobody chooses a community platform for its events tooling or its mobile app, and for a support and product community neither is likely to decide it.

How Far You Can Shape It

This isn’t one of the nine scores, but it’s becoming one of the most important questions to ask any vendor: how far can you change the platform without hiring developers for six months?

Higher Logic Vanilla led here for a while with its Widget Builder. This year Gainsight shipped its answer, Developer Studio. You can build and version custom widgets in GitHub, or describe what you want in plain language and have it write the code, with proper staging and production environments. You could replace the generic homepage with a search bar and the single most useful unanswered question. You could put a badge on member profiles showing their plan or account health from your CRM.

Here’s the limit worth naming, though. Developer Studio lets you add widgets, but the platform underneath is opinionated and purpose-built. The information architecture is deliberately shallow, the content types are fixed, and the deeper theming is limited. That’s a trade. You give up flexibility to get something faster to stand up and harder to break. If your plan is to bend a community into a shape it was never designed for, this is not the most flexible tool on the list.

Being Part Of Gainsight Cuts Both Ways

The factor that isn’t on the feature list is the one I’d weigh hardest.

If you already run Gainsight on the success side, this is close to a no-brainer. The account context and the success alignment just work, and the best parts of the product, the account-value features especially, are built for exactly your situation.

But if you’re not a Gainsight customer, that same tight coupling can put you off. You might worry it’ll pull you into the whole Gainsight suite over time, or that the features that make it special are really meant for people already in that world. You can run Customer Communities on its own, and plenty of companies do. Its identity is just so tied to Gainsight that it reads as a bigger commitment than a standalone community tool. That helps it with existing Gainsight customers and works against it with everyone else.

The Verdict: Should You Build On Gainsight CC?

Buy Gainsight CC if you’re B2B SaaS, your community exists for support deflection and product feedback, and you want your roadmap tied to your customer accounts. It’s an easy call if you already run Gainsight on the success side, because that’s where the account context comes from.

Look elsewhere if you need a native mobile app, if events are the centre of your community, or if you want to build a completely bespoke, heavily customised experience. Courses, though, are no longer a reason to look elsewhere. The Skilljar layer changed that.

Put the whole card together and you get a real all-rounder with a clear identity: the community platform built for customer success.

Where To Go Instead

If Gainsight CC isn’t the fit, here’s where I’d point you. Pick Higher Logic Vanilla if you want the cleanest question-to-answer-to-knowledge loop for support. Pick Discourse if your community is discussion-led and technical. And if events are the whole point, look at Bevy or Gradual.

If you’re leaving Khoros, the real question isn’t whether to go, it’s which of those platforms you land on. For a customer success community, Gainsight CC is a strong answer.

More In This Series

This is part of my 2026 series reviewing the major enterprise community platforms against the same nine-capability scorecard. If you’re weighing a move, these go together:

The post Should You Build Your Community On Gainsight CC? first appeared on FeverBee - Community Consultancy.

Should You Build Your Community On Discourse?

Richard Millington — Tue, 07 Jul 2026 13:29:16 +0000

Of every platform I review in this series, Discourse is the one I score highest on the thing communities are actually built to do: hold a good discussion. That is a 9 out of 10, the top score I give anyone. It is also one of the three platforms, alongside Higher Logic Vanilla and Gainsight, that I see Khoros communities migrating to most often.

I run it through the same nine-capability scorecard as every episode, so you can compare it like for like. Here is why it is so strong, where it falls down, and the one thing you have to get right if you move to it.

Where Khoros communities are going

I will start with the evidence for that claim, because it is not a hunch.

Look at where some of the biggest communities leaving Khoros have actually landed. Google moved its developer community and its Local Guides community onto Discourse. Shopify runs its community on it. And HubSpot have just relaunched their community on it too.

These are not small forums. As I record this, the Shopify Community on Discourse has around 1.5 million members and 1.8 million posts. Google’s developer forums have over 260,000 members and close to a million posts. HubSpot has around 500,000 members and 800,000 posts. Those are live numbers, and they are some of the largest brand communities on the planet. When they left Khoros, this is where a lot of them went.

So when I say Discourse is one of the three places Khoros communities most often land, alongside Higher Logic Vanilla and Gainsight, that is what I am basing it on. The question for you is not whether serious companies trust it. It is whether it fits what your community is for.

A product-first company

To understand Discourse, you have to understand that it has been a product-first company for its entire life, since Jeff Atwood founded it in 2013.

That came with an advantage and a disadvantage. The advantage is that they built the best forum-centric product in the market, by a distance, because the product was the obsession. The disadvantage is that for years they were behind on the things around the product: support, sales, professional services, the commercial machinery other vendors had figured out. If you wanted a slick sales process and a big account team, Discourse historically was not where you would find it.

That history still shapes the platform you would buy today. The product runs deep. Outside the product, the rest of it is catching up, and it matters if your last impression of Discourse is a few years old.

They have built a proper team, and it is a good one. They are the most technically savvy group I engage with in this space. Almost everyone I have spoken to there actually knows how to use the platform and understands community better than the sales teams I deal with at most other vendors. That is rare, and it is the precise thing people leaving Khoros say they stopped getting.

There are two honest caveats. The product-first focus still lingers a little. And they are still a relatively small team compared to the bigger organisations serving this space. But they punch well above their weight, and where it counts, the people are excellent. That product DNA also shapes where they are strongest: technically minded communities, developers and gaming especially, the places where code blocks and the deeper forum mechanics genuinely matter.

Discussions and Q&A

Discussions and Q&A. I score Discourse a 9, the highest discussion score I give any platform in this entire series.

Discourse is the benchmark other platforms get measured against. It tracks each member’s read position, so it knows where everyone left off, and it lets the community edit posts together in wiki mode. The composer handles code blocks with proper syntax highlighting. You also get real-time presence, topic timers, slow mode, whispers, custom reactions, bookmarks, and tag groups. The admin tooling underneath is excellent: you can merge, split, move, and archive topics, run bulk actions, pull up full revision history, and schedule auto-close and auto-bump. If running a serious discussion is the job, nobody does it better.

On the Q&A side, there is a plugin that gives you accepted answers and a proper question-and-answer format, and it works well. Discourse goes deeper on raw discussion craft than anything else I review.

One recent thing worth knowing: nested replies only shipped in mid-2026. So if you looked at Discourse a year ago and the flat reply structure put you off, look again.

AI and the MCP

Second, and this is where Discourse has genuinely surprised people. AI. I score it an 8.

There is a live MCP, the connection that lets external AI tools read and act inside your community through a standard interface. There are AI Agents, thirteen tools that can close a topic, edit tags, set slow mode, grant a badge, or mark a question as solved, with a human reviewing before anything goes live. There is AI search across the whole community. It writes summaries of long topics, auto-tags new posts, runs AI spam detection, and there is an AI bot that answers questions straight from your community’s content.

Two of those deserve a closer look.

First, the MCP. Plenty of platforms are now bolting on a way for AI to read your community. What makes Discourse’s different is that it does not just let an assistant read, it lets it act, through those thirteen agent tools, and it sits on a community that is already machine-readable from the ground up.

Which is the second, and smartest, thing they have done. Every page in a Discourse community is machine-readable. Add “.json” to the end of any topic’s web address and you get a clean, structured version of that thread: the question, the answers, who said what, and which reply was marked as the solution. You do not switch it on. It is just there. So when someone asks ChatGPT or Claude a question your community has already answered, Discourse is one of the formats those tools find easiest to read and to cite. They have gone further, with support for the newer llms.txt standard so you can hand AI tools a curated map of your best content.

The practical upshot: if being found inside AI answers matters to you, and it should, Discourse is unusually well set up for it. The reason this scores an 8 is that it is shipping, not roadmap. They put this stuff out month after month.

Moderation

Third strength, and it connects straight to the last one. Moderation. Also an 8.

The same AI agents carry a set of moderation-specific tools, paired with that human review queue, so you get AI doing the first pass and a person making the call. Underneath sits Discourse’s trust level system, which automatically promotes members through four tiers as they earn it, which quietly takes a lot of routine moderation load off your team. Add AI spam detection and a deep plugin ecosystem and you have one of the stronger moderation suites in the market.

This matters more than it used to. If you are a large consumer community with real safety concerns, moderation is no longer a nice-to-have. Discourse holds up well here, and it is a common reason I see enterprise teams shortlisting it.

Where it is middle of the pack

Now the capabilities where Discourse is fine, but they are not why you would choose it.

Gamification, a 6. The trust level system is an elegant hybrid: it gamifies participation and reduces moderation burden at once, because earning trust unlocks abilities. There are badges with custom triggers, and an official plugin that adds leaderboards and scoring. But it does not have the deep, native points-and-ranks reputation engine Khoros built, which is why Khoros still scores a 7 here and Discourse sits just behind.

Content and knowledge, a 5. You get the basic building blocks: a Docs plugin that gives a knowledge-base style view, wiki mode so the community can maintain articles, and a proper rich-text editor since 2025. But it is not a real knowledge platform. No native courses, no blog, no video library, and the knowledge-base experience is lighter than the tools built specifically for it. It is adequate. It is not a reason to buy.

Mobile, a 5. A strong progressive web app that feels close to native, a Discourse Hub app for iOS and Android, and the option of a custom branded build. Better than Vanilla at a 3, behind Khoros at a 6. Solid, not a selling point.

Analytics, a 5. There is built-in reporting for content and engagement, extended by plugins. One point in its favour: because Discourse is so open, it is easier to get your data out and into your own analytics tools than with most platforms. But the reporting inside the product is still a bit weak, so if your execs want board-ready dashboards out of the box, you will feel that gap and lean on the API to close it.

Customisation and depth: the thing you have to get right

Now the thing that defines the whole platform, and the single decision that makes or breaks a move to Discourse.

Discourse has some of the most advanced features and functionality out there, and with a deep plugin ecosystem plus the ability to get right into the code, the ceiling on what you can build is effectively unlimited. The critical thing, if you move to it, is that you actually take advantage of that. The whole value of Discourse is that it goes really deep in a handful of areas: discussion mechanics, moderation, AI readiness, the developer and gaming use cases. If you are inside those areas and you use the depth, almost nothing competes. If you are outside them, that depth is not doing anything for you, and the capabilities we just walked through are where you would feel short-changed.

There are two things to consider. First, a lot of that power lives in plugins and custom development, so you want a developer in the loop, or budget for Discourse’s own services team. A large enterprise team currently moving off Khoros told me they had had to install some plugins and do a little custom development, but that it was easier than the heavy-handed custom work Khoros had required. That matches what I hear repeatedly.

Second, it can be a lot. Discourse is so feature-rich it can feel overwhelming, on the front end for members and on the back end for admins. This is not a simplified forum experience. It is a deep forum experience.

So here is the failure mode I would warn you about, because I have seen it. A team buys Discourse for the depth, then treats it like a simpler packaged platform. They never put the technical resource behind it. They do not build the plugins, they do not use the deep mechanics, and they end up with something more complex than they needed and less finished than what they left. The platform did not fail them. They bought the depth and never used it. If you are not going to resource it, a more packaged platform will make you happier.

One practical note for your evaluation. You do not get a sandbox to trial your own configured instance the way the usual enterprise platforms hand you one. There is a generic demo you can poke at, but testing Discourse properly means committing to a real build. That raises the stakes on getting the decision right up front, so do the homework before you sign.

The gaps: events and ideation

Now the two weakest areas, stated plainly.

Events, a 3. Events used to be a bolt-on plugin. It is now bundled into Discourse core, with an interactive calendar, event posts with going and interested RSVPs, recurring events, timezones, even a holiday calendar. That is real, and it is well ahead of Khoros, which has nothing native and scores a 1. What it still is not is a full events platform. There is no ticketing, no registration pages, no check-in. So if meetups and webinars are the point of your community, this will not be enough, and you would want Bevy or Gradual.

Ideation, a 3. There is a Topic Voting plugin that creates vote-sorted idea categories, and real communities use it in production, but it is not a purpose-built ideation workflow. If feeding your product roadmap is the main point of your community, Gainsight is miles ahead at an 8, with idea scoring weighted by account value.

Neither is fatal for a discussion-led community. Both are dealbreakers if they are central to what your community is for. Know which you are before you sign.

Is Discourse a healthy vendor?

In the Khoros review I said you buy the company, not just the software, for three to five years. That cuts the other way here.

On the things you can actually observe, Discourse looks like a healthy vendor. They are putting out new product on a steady cadence, the AI work especially. They are actively winning new enterprise logos, which is exactly the signal I told you to look for and exactly the signal Khoros is not giving. And the team is technically excellent, and people like working with them.

There is one thing to weigh. Discourse also cut staff in late 2025, which they framed as adjusting their cost structure for long-term sustainability. The difference, and this is my read, is the situation around it. Khoros cut deep after a private-equity acquisition while customers were leaving. Discourse cut from a smaller base and has kept shipping and winning customers since. Same action, very different context. I would weigh the context, not just the headline, and I would ask any vendor, including this one, what their team looks like today.

That is why Discourse is still one of the three platforms, alongside Vanilla and Gainsight, I see Khoros communities migrating to most often. The discussion quality is part of it. The bigger part is that the company behind it is pointed in the right direction, shipping toward where the market is going rather than away from it. For a multi-year bet, that matters as much as any single feature.

So Who is it For?

Buy Discourse if you recognise this:

Discussion is the heart of your community, and you want the best discussion experience on the market.
Your members are technically minded, developers, gaming, power users, the audiences Discourse was built for.
You have, or can get, a little technical resource, and you are willing to actually use the platform’s depth.

Look elsewhere if your community runs on events, if product ideation is the main job, if board-level analytics out of the box are non-negotiable, or if you need a stripped-back, simple experience your members can use without any learning curve. Discourse is a deep forum, not a simplified one. For the right community that is the whole point. For the wrong one it is too much.

Price and Tiers

On commercials, Discourse is more transparent than most of this market, which is refreshing.

They sell hosted plans in clear tiers. Pro and Business are publicly priced, roughly a hundred and five hundred dollars a month respectively, and the full AI feature set is available across them, not locked to the top tier. Above that sits Enterprise, which is custom-priced and where you would land as a large organisation, with the security, scale, and support a big community needs. For that one you talk to sales and get a current quote, and as with every platform in this series, do not take a number off a slide, including mine, because it moves.

There is also a self-hosted option for teams who want to run it themselves, but for an enterprise the realistic conversation is the Enterprise tier. To give you a rough anchor, enterprise and RFP-level engagements have started in the low thousands of dollars a month with a dedicated success manager attached. Treat that as a floor rather than a quote. The point worth taking away is that the lower tiers being openly priced tells you something. You can see what you are getting before you ever speak to a salesperson, which is not true of most of the heavyweights.

The Verdict

Discourse is the best discussion platform in this series, full stop. It scores a 9 on discussions and an 8 on both AI and moderation. If your community is built around conversation and knowledge, especially with a technical audience, it is at or near the top of your shortlist, and the AI readiness is a genuine reason to prefer it right now.

It is not an all-rounder in the way Vanilla or Gainsight are, and it is not trying to be. It is a deep specialist. Events and ideation are its weakest areas, and content, mobile, and analytics sit in the middle of the pack rather than out front. That is the honest picture.

So rank it like this: number one for discussion-led communities, especially technical ones, a strong choice for anyone leaving Khoros, and the wrong tool if events, ideation, or out-of-the-box analytics are the point. If you are leaving Khoros and discussion is your core, this is the first platform I would look at. Just go in committed to using what makes it special.

The full scorecard behind this series is free. Download it, check it against your own shortlist, and tell me in the comments where you think I have got Discourse wrong. And if you want help choosing a platform, or getting off Khoros cleanly, you can find me at feverbee.com.

The post Should You Build Your Community On Discourse? first appeared on FeverBee - Community Consultancy.

Khoros in 2026: An Honest Review (And Why I’d Leave)

Richard Millington — Wed, 01 Jul 2026 13:48:30 +0000

Khoros is the platform I get asked about more than any other, and the one that has caused the most upheaval in our space over the last few years. So let me give you an honest review of where it stands in 2026, and whether you should stay or go.

This one is different from every other platform I review. The product is strong. In some areas it is still the best on the market. And I would still tell most teams on it to leave as soon as they reasonably can.

One thing you should know going in: Khoros has sent me legal threats over my past commentary about them. So I am going to be careful, frame this as my opinion, and stick to what I can stand behind. Two things are true at once. The product is good, and I would leave. The gap between those two things is the whole story.

How Khoros got here

For most of the last fifteen years, Khoros was the premium community platform in our space. If you were a big brand running a big support community, you were probably on Khoros, or on Lithium before it. It earned that. The feature set was deep, and it ran at a scale almost nobody else could touch. A lot of us still carry that version of Khoros around in our heads.

The product still deserves a fair amount of that respect. The company around it, in my view, no longer does.

Khoros came out of Lithium, one of the original enterprise community platforms, which merged with Spredfast in 2019 and rebranded as Khoros. In my experience, that merger started a rough stretch. The attention drifted toward social media management, the community product did not move forward the way it should have, and the cost and service charges came up again and again as a real source of frustration. Then in May 2025, Khoros was acquired by IgniteTech. That is the moment everything else here hangs on.

What Khoros still does well

Before the rest, the product deserves its due, because the scorecard is genuinely good. On discussions and Q&A it scores an 8: deep threaded forums, a proper question-and-answer format with accepted answers, and strong enterprise admin tools. Moderation is an 8, with a mature suite, AI models for spam and abuse, and support for dozens of languages. Analytics is an 8, with deep Adobe and Google Analytics integration. Gamification, ideation, and content all sit at 7. The one real hole is events, which I score a 1.

Two things stand out. The first is scale. Khoros has run some of the largest brand communities on the planet, with enormous traffic, deep permissioning, and complex multi-board structures, for years and reliably. The second is gamification and reputation. Lithium, and then Khoros, with Michael Wu, Joe Cothrel and the team, wrote the playbook for it. A lot of platforms are still copying the model.

So none of what follows is that the product is bad. On its own terms, it is strong. The problem is that you do not buy a community platform on its own terms.

You are not buying a product, you are buying a company

You buy the company behind the platform, and you commit to it for three to five years. Over that window you are not betting on the feature list, you are betting on the vendor: will they support you, keep the lights on, invest in the product, and treat you fairly when something goes wrong. That is where, in my opinion, Khoros has become a bad bet.

Khoros is owned by IgniteTech, part of ESW Capital, which has bought well over a hundred software companies. As best I can tell, the model is consistent: acquire enterprise software with customers locked into longer contracts, cut costs hard, and run it for cash. This is not only my read. IgniteTech’s own CEO has spoken about laying off close to 80% of the company after staff resisted an AI-first mandate.

Within weeks of the acquisition, almost the entire Khoros team was let go. That matters, because the feature set I just praised was built and maintained by people, and much of the work is now done by contractors with far less knowledge of community or the clients. In my experience, that kind of gap shows up slowly, then all at once, usually at the worst possible moment.

All-in on AI, and the Aurora question

What has replaced those people, in the messaging at least, is AI. In April 2026, Khoros launched Aurora AI, described not as an upgrade but as a brand-new, AI-native platform, with a suite of tools the parent company rolls out across its acquisitions. I am not anti-AI; we do a lot of work helping clients get AI ready. The open question is whether this is AI that helps a team do more, or AI positioned to replace the team. Judge it on what it actually does in your community, not on the launch copy.

If you are a current customer, pay close attention here. You are probably on the older platform, Khoros Communities Classic. Khoros is offering free migrations from Classic to Aurora through the end of 2026. Aurora may turn out to be good, but it is a relatively new platform, it does not have anywhere near the feature set of Classic, and any migration is a real risk. As far as I can tell, there is no published end-of-life date for Classic. You would be migrating either way, so if you are going to take on the cost and risk, at least look at what else is out there before defaulting to Aurora.

What customers are telling me

I get messages almost every week from Khoros customers describing real pain, and most of the ones I have spoken to are actively planning to leave. The first theme is support: account managers and commercial teams going quiet for weeks, issues left unanswered, and how hard it is to get anyone on the phone. The second is reliability: outages and platform issues, with some clients left running their own analysis to work out what went wrong. I cannot independently verify every incident, so take it as what customers report, but I am hearing it from enough of them that, in my opinion, it is a trend rather than bad luck.

Then there is getting your data out. From what I have heard, a full, usable export of your content, members, and structure can be more of a fight than it should be, and some customers say they have been charged for additional exports. And there are the invoices: large, unexpected overage bills, sometimes into the hundreds of thousands of dollars. It has become a common story that bot traffic has surged on a community and the customer is then charged for it, with customers telling me they believe much of that billed traffic is bots and automated crawlers, not the human visitors their contracts were written around.

The verdict: should you stay on Khoros?

This is why I score platforms on nine capabilities but tell you to weigh a tenth thing that is not on the sheet: the vendor. Khoros still scores like a top-tier platform, and on product it nearly is. But the scorecard measures what the software can do, not whether the company behind it will still be supporting you fairly in three years. On that question, in my opinion, Khoros is now one of the riskiest choices in the market.

My answer is no. If you are on Khoros today, start planning your exit now, not in a panic, but deliberately. The worst position to be in is deciding to leave the day a surprise invoice lands or your support stops responding, because that is when you have no room to negotiate and no time. Almost everyone I know on the platform is already looking, and many have moved.

Where to go instead

The right alternative depends on what your community is for. If product feedback or Gainsight CRM integration is the point, look at Gainsight. For a general-purpose enterprise community built around Q&A and knowledge, Higher Logic Vanilla is one of the strongest all-rounders. For a daily, social, members-hanging-out community, look at Higher Logic Thrive or Bettermode. Discourse is where a lot of Khoros communities are already migrating and is hard to beat on discussion depth. And if you are built around events or want a vendor investing heavily in AI, look at Bevy.

Map your must-have capabilities first, then test these against your list, not against a demo. If you do move, do it properly: get your data out early, and test that you can export your content, members, and the SEO value tied to your URLs cleanly.

You can see the full nine-capability scorecard and platform comparison at feverbee.com/communityplatforms. And if you want help getting off Khoros or choosing what comes next, you will find me at feverbee.com or richard@feverbee.com.

The post Khoros in 2026: An Honest Review (And Why I’d Leave) first appeared on FeverBee - Community Consultancy.

Higher Logic Vanilla: An Honest Review

Richard Millington — Wed, 24 Jun 2026 15:55:07 +0000

Higher Logic Vanilla: An Honest Review

This is the first in a new series, Enterprise Community Platform Reviews 2026. I’m reviewing the major enterprise community platforms one at a time, using the same nine-capability scorecard each time so you can compare them like for like. First up is Higher Logic Vanilla.

Before we get into it, the context matters, because the ground has shifted under this whole decision. Choosing a platform used to be a five-year bet you could get slightly wrong and survive. That has changed, and the reason is Khoros.

Khoros was, for years, the default heavyweight for enterprise communities. Now it has faded. They wound down Khoros Care, customers are migrating off, and when your platform vendor is in trouble, your roadmap is too. At the same time, AI has raised the stakes on getting the choice right, because your platform is now feeding answers to your members and to their AI tools. Pick the wrong one and you feel it faster than you used to. So the pressure to choose well is higher than it has been in years.

That is also why Vanilla is interesting: it nearly missed its moment. Vanilla started in 2006 as forums that looked good and were easy to stand up, a small team winning enterprise logos it had no business winning. Then Higher Logic acquired it in 2021, which made sense on paper. Higher Logic was strong in associations and weak in the enterprise. Vanilla was the opposite.

Then it drifted. For a couple of years after the acquisition it felt like Higher Logic had taken its eye off the ball. When we ran our Buzz Report a couple of years ago, Vanilla was one of the platforms starting to look left behind. The product wasn’t keeping pace.

That is no longer the case. Vanilla is shipping again, and shipping the things that matter for where the market is going. That turnaround is why, for a general-purpose enterprise community, I now put Vanilla in the top two all-round platforms you can pick. Here’s the evidence, and I’ll be honest about what’s still missing.

Vanilla, not Thrive

One thing to clear up first, because it constantly confuses buyers: Higher Logic sells two community platforms, not one.

Higher Logic Vanilla is the one I’m reviewing today. It’s the B2B and B2C product, built for support, knowledge, and product feedback, with Q&A at its heart.

Higher Logic Thrive is different. Thrive started out aimed at associations and now goes after the enterprise market too. It’s a different experience: the platform you pick if you want members hanging out and chatting every day, rather than the more focused Q&A experience that defines Vanilla.

They’re also run quite separately inside the company. They don’t even share a support team, and they’re marketed and sold separately. So if a daily, social, members-hanging-out community is what you’re after, Thrive is probably your conversation, and I’ll review it on its own. Everything from here is about Vanilla.

So, should you buy Vanilla? You should if you recognise this: you’re running an enterprise B2B community for support, knowledge, and product feedback; you need deep permissioning, with different members seeing different things; you want a real path from a question to an answer to a knowledge base article; and you may need multiple brands or languages on one platform.

Look elsewhere if your community is built around events, or if most of your members live on their phones. Those are the two places Vanilla is genuinely weak, and I’ll show you the numbers rather than bury them at the end.

Discussions & Q&A

Let’s start with the core. Discussions and Q&A. I score Vanilla a very solid eight out of ten here.

The standout is the Q&A. Vanilla has the cleanest separation between an open discussion and a question that needs an answer of almost any platform I reviewed. You get accepted answers, AI-suggested answers, and a genuine route from a resolved question into a knowledge base article. If your community exists to deflect support tickets, that loop is the whole game, and Vanilla runs it well.

It also has the best Zendesk integration of the group, which matters more than it sounds if your support team lives in Zendesk. You can see this working on the Smartsheet community, which runs on Vanilla: 80% of questions there are answered by other members, only 20% by moderators. That ratio is what good Q&A design buys you.

Here’s the honest counterweight. On raw forum depth, Vanilla is a notch behind Discourse, which I’d put at a nine. Discourse has the read-position tracking, the wiki mode, the code blocks, the real-time touches. If a pure, forum-only experience is what you want, Vanilla probably isn’t the best option. But if you want that integrated support experience, Vanilla is fantastic, and for most enterprise teams this isn’t the deciding factor.

AI & the MCP

Second, and the most forward-looking one: AI. I rate it a seven for Vanilla.

At the time of writing, Vanilla is one of only four platforms with a live MCP, the Model Context Protocol. It’s a connection that lets your AI tools read and act inside your community. Vanilla’s is the only one of the four that’s free on every tier and can take member-facing actions, not just answer questions. It can send a DM, award a badge, and tag content. That’s a different category of useful.

Two honest caveats, because I haven’t been able to test it hands-on yet. It’s still in beta, and access is gated behind a beta agreement, so the rollout is phased.

When I spoke to Higher Logic’s own team about this, it’s clear they’re rebuilding around it. They’re rebuilding their moderation tooling to fill the gap enterprises felt when Khoros wound down Khoros Care. They’re adding automated demographics, so the platform can segment members by tenure, role, or activity and then act on those groups. And they’re building what they call data shipping, so enterprise customers can get all of their data out. That’s a credible roadmap. It’s also a roadmap, so judge it on what actually ships.

Other strengths

Vanilla has some other strengths worth considering.

First, customisation. You get more here than on most platforms, more than on Gainsight and others. The layout editor lets you drop in and target widgets by role without a developer. The Widget Builder goes further, syncing files to your local environment and giving an AI coding tool the guardrails and data types it needs to build a custom widget for you. The ceiling is high, though reaching it still rewards someone with design and technical skill.

Second, permissioning. Vanilla goes deep: specific roles, private areas, category-level access, beta groups, and more. If different members should see different things, this is one of the best in the market.

Third, scale. It handles multiple brands, sub-communities, and languages from one instance about as well as anything out there.

Fourth, analytics. You can build your own dashboards, filter deep by category, and pull everything out through the API, which Higher Logic’s own team will tell you is one of the strongest parts of the product. There’s an ROI calculation tool for case deflection, and analytics that flag advocates in the making and accounts drifting toward churn. The limit is manual export: getting data out by API is easy, a clean manual export less so. The MCP tackles a lot of this.

Fifth, support. Like the better platforms in this market, Vanilla has a strong support team. Customers consistently rate the people they deal with very highly, and that matters more over a multi-year contract than any single feature.

The gaps

Now the holes, the places Vanilla doesn’t score well.

The first is events. I’d score Vanilla a 3 out of 10. If your community runs on meetups, webinars, and an events calendar, this is not your platform. Bevy and Gradual are far better suited to that.

The second is mobile, also a three. The experience is desktop-first. If your members live on their phones, you’ll feel it. Mighty Networks and Hivebrite are best in class here; Vanilla is not.

Neither of these is fatal for a classic enterprise support community sitting on the desktop. But both are dealbreakers if they’re central to how your members behave. Know which you are before you sign.

Ideation

One more, smaller, thing to know about: ideation.

I don’t think ideation is as important to most companies as it used to be, but if it is for you, I score it a five out of ten. You get ideation boards with voting, but the workflow is less sophisticated than the best-in-class platforms. For a B2B SaaS community where product feedback is the point, Gainsight is well ahead here, with idea scoring weighted by account value. If your community’s main job is feeding the product roadmap, weigh that hard.

Worth flagging too: Vanilla finally offers email digests, which corrects a long-standing gap.

The verdict

Put it all together. For a general-purpose enterprise community, I think Vanilla is in the top two, probably alongside Gainsight. The specialists beat it on their one thing: Discourse on the depth and quality of discussions, Gainsight on product feedback, Bevy and Gradual on events. But across everything an enterprise community needs at once, Vanilla doesn’t have an obvious weakness, and that breadth is what most teams really need.

So is it for you? For most enterprise B2B community teams, yes. At the very least, put it on your shortlist and get to know the team, particularly if you’ve already ruled out the heavier legacy platforms and you want a vendor shipping toward the AI era rather than away from it.

Say no if events or mobile are core to your community, or if you have one overriding need that a specialist serves better and you’re willing to give up breadth to get it. The real trade-off with Vanilla has always been the same: ease of use versus customisation. It strikes a better balance between the two than almost anyone, but if you sit hard at either extreme, you’ll feel the pull.

On commercials, a couple of things to know. Vanilla is enterprise-priced and sold on multi-year contracts. I’m not going to quote a number, because what you pay depends on your size and your negotiation, and anything I put here would be out of date by the time you read it.

Honestly, I find it striking that if you’d gone back five years and said Vanilla and inSided, now Gainsight, would be the top two all-round platforms, most of us would have been surprised. But that’s how it has played out. Vanilla is exceptional at the Q&A-to-knowledge loop, strong across the middle, forward-leaning on the AI connection, and held back mainly by events and mobile, neither of which I think is a severe issue. More than any other platform right now, Vanilla has momentum, and it’s winning a lot of the major clients out there today, especially those moving off Khoros.

Before you go

Next in the series, I’ll take the same nine-capability scorecard to the next platform so you can compare them like for like.

The full scorecard behind this video is free. Download it, check it against your own shortlist, and tell me in the comments where you think I’ve got Vanilla wrong. I read them, and I change my scores when the evidence says I should.

And if you want help choosing a platform or getting a community ready for what’s coming, you can find me at feverbee.com.

The post Higher Logic Vanilla: An Honest Review first appeared on FeverBee - Community Consultancy.

Is Community a Dead-End Career?

Richard Millington — Mon, 08 Jun 2026 17:04:23 +0000

Is working in community a dead-end career?

It is one of those questions that matters to everyone working in this space, which is exactly why so much bias creeps into the answer.

What we want to be true and what is actually true can be many different things.

So I want to explore what the data actually says. We can all think of people who worked in community and then climbed to a really high level. But that runs into survivorship bias. For every person who made it, how many didn’t? How many people are stuck in the same job year after year? How many haven’t made any progress?

I will give the answer first, then explain how we got there.

Is community a dead-end career?

For many people, somewhat. For some people, absolutely not. And for a few, the answer doesn’t really matter.

Let’s go into the data.

Why This Question Is Coming Up Now

I spent some time pulling together all the CMX industry reports from the last four or five years. The picture is consistent: the community industry is contracting. I personally think we have been in a community recession for a couple of years now, and it only seems to be getting worse, especially for the B2B enterprise communities we typically work with.

Fewer people answering these surveys are working full-time on community than before. And a lot of the career advice in our space is now out of date. What we used to share from 2018 to 2022 doesn’t apply the same way today. It was written before AI changed the entry rung, before the contraction, before the big tech layoffs.

When you look at the CMX data, three trends stand out:

The number of full-time community roles is down.
Confidence in increased investment is significantly down from the post-pandemic peak.
Interest in community from other departments has dropped dramatically in recent years.

I really wish CMX still published the interest-from-other-departments number, because it is the one I find most worrying. None of these three measures point to anything positive about our space.

The Three Data Sources

For this analysis we drew on three sources.

One, the CMX industry reports. Their 2026 report is due soon and I’ll take a look at it when it lands.

Two, an 11-year cohort study. More on that in a moment.

Three, the broader labour market data: public salary numbers, layoff trackers, and what is happening at the company level for community pros.

The Community Pyramid Is Unusually Flat at the Top

The first finding is structural. If you look at community job titles, the pyramid is far flatter at the top than in most professional fields.

Roughly 21 managers for every one VP.

I need to caveat that. “Manager” means different things in community from elsewhere, so the comparison is not perfectly clean. But 21 to 1 is still a very big number. There just aren’t that many senior roles to climb into.

Where 86 Community Pros from 2015 Ended Up

The problem with all CMX survey data is that it is a snapshot. You are measuring the people who took the survey that year, not following the same people over time. Different respondents in different years, so what looks like a trend might just be sampling variation.

So we did something different. Back in 2015 we hosted an event called FeverBee Sprint, 11 years ago. We had all the attendee information: their job titles, the companies they worked for. We selected a sample from that list and tracked what happened to them.

Of the 86 we could verify on LinkedIn:

8% climbed within community to a Sr Director or VP role
1% became senior at a community-as-product company
24% are still in the same role today (this one absolutely shocked me)
41% moved laterally out of community into a different function
20% went independent or founded their own thing
6% left the field entirely

The 24% is the number that stayed with me. I will caveat it: this is based on LinkedIn, and if someone hasn’t updated their profile in years they would still be counted. But as best we can tell, those people are in the same job today as they were 11 years ago.

I would love to do this with a bigger sample. We could include all the attendees from all our events over the years. But it takes a serious amount of time and energy. Maybe at some point.

The Single Most Useful Number: Stay Versus Leave

What happens to people’s careers when they leave community compared to staying in it?

This is the most definitive metric we have for the original question.

If you stayed in community, there is a 28% chance you reached Director or VP level after 11 years. If you moved out to another function, there is a 60% chance you did. Twice as likely to climb if you leave.

You could argue that more ambitious people are also more likely to switch roles, so this is not a clean causal story. Fair point. But the difference is striking.

What we found is that community seems to be a 3 to 7 year window of doing the core work. Either you get stuck in it, or you need to leave community after that time if you want to progress.

Only 8% climb within community itself. Only 8%. Despite all the articles written about VP of Community roles, there really aren’t that many roles to climb into.

The Three Paths That Actually Work

If community is your career, three paths genuinely work.

Path One. Big Org, Manage a Team by Year Four

If you want to reach VP level, you need direct profit-line responsibility and you need to be managing a team. You won’t get there without that. Position yourself to work within a team if you want to manage a team one day. If you are solo or part of a team of two, the progression prospects are limited. We’ve written more about the strategic options for enterprise communities here.

Path Two. Work Where Community Is the Business

Some of the people doing very well are at organisations where community itself is the product or central to the business. Stack Overflow, Reddit, Discord, Discourse. Senior community roles exist there because the company depends on them.

Path Three. Consultancy, Agency, or Fractional CCO

FeverBee is one example. There are plenty of other consultants in this space. I’ll say this having watched around 100 people try to transition into consulting over the last 15 years: it isn’t easy. Very few make it. Part of the reason is that there aren’t many clients to go around, and it’s hard to differentiate from people who have been doing this for a long time. Not impossible, but you need a clear answer to why you are the best option and how you will attract clients. Do you have unique relationships or a reputation no one else has? Most people believe they can do the work. That is a separate topic. Attracting clients is the harder problem. If you make it work, brilliant. The overwhelming majority of people I have seen try haven’t been successful.

Five Takeaways for Anyone Doing the Work

Treat community as a 3 to 7 year career window. You are probably not going to be in this for 20 years unless you want to be. Some people enjoy the core work and stay in it forever. No shame in that.
If you want to grow, move to an org with a real community team. You need to be in a place that invests in community: good platforms, dedicated headcount, room to climb.
Pick a community type that maps to a revenue function. Customer marketing, product, advocacy, customer experience. Support is fine, but I noticed that people doing pure online support communities tend to progress less.
Own a metric the business actually cares about. Pipeline influenced, expansion revenue, deflection. Not engagement.
If you are five years in and nothing has changed, the data is telling you something. You are probably in the situation many people are in: there is nowhere to go from where you are. If you want to grow, you will probably need to make a lateral move.

The exceptions exist. Big org with a real team. Community as the product. Going independent. There are opportunities. But if you stay in the role and hope to progress through doing community work alone, the data says you will most likely be disappointed.

The Picture in One View

The headline numbers, the industry pyramid, the climb rate, and the career outcome breakdown:

So, Is Community a Dead-End Career?

It can be. It definitely can be.

But if you want to progress, you have to align yourself with a function where progression is possible. Either work within a larger team where you can climb, or make a lateral move out of community where you can take the connections, experience, and expertise you’ve built into the next level.

Good luck.

The post Is Community a Dead-End Career? first appeared on FeverBee - Community Consultancy.

The Way We Compare Community Platforms Is Broken (new tool)

Richard Millington — Wed, 03 Jun 2026 15:52:31 +0000

TL:DR – We’ve launched a new platform feature selection tool – you can find it here.

Tell me if this sounds like a familiar story.

You decide to migrate your community. You speak to stakeholders, create a list of requirements, and compile them into a weighted RFP. This may run from 20 to over 200 rows.

You send the RFPs to platform vendors and get responses within a week.

As you review the list, you soon discover that nearly every community platform vendor claims to do everything you want. What luck!

So you select a platform based on price or other nebulous attributes and get started.

And that’s when problems begin. Suddenly, the features don’t quite work as planned. Some aren’t quite as ‘out of the box’ as you were led to believe. Others just look a little ugly and outdated. Sometimes the functionality is atrocious. The problems begin to accumulate as does the cost.

The brutal reality begins to dawn on you…you’ve picked the wrong platform, and now you’re stuck on it for the duration of a 3-year contract.

The lesson is that there’s a gaping void between a platform vendor saying, “Yes, we definitely have that feature,” and that feature being any good.

And the problem right now is that it’s very difficult to tell the two apart because people aren’t even sure what to ask for.

How Do You Compare Platforms That Offer Similar Feature Sets?

This confusion has existed for decades. The difference today is that the new breed of platforms is nipping away at the old guard, promising an equivalent feature set at a fraction of the price.

Which leads many organizations to ask: can a platform a few years old really have the feature set of one that’s decades old?

The RFPs we’re using do a terrible job of distinguishing between features and functionality. They ask about features instead of going granular with functionality.

This lets any platform, regardless of how new or lightweight it may be, claim near-feature parity with the old guard.

This creates a lot of confusion, not only about what separates the legacy platform vendors from one another, but also about what separates them from the newcomers.

And this problem is particularly acute when the project is being run by a non-community stakeholder (i.e., IT, marketing, or customer support). Their platform requirements are often aligned with their goals rather than with the best community experience.

To Help You Navigate This, We’re Publishing Our Platform Feature Selection Tool

To help you select the features that you need and have the tricky internal discussions that you need to have, we’re introducing our new platform feature selection tool.

It’s a tool that will help you clarify the difference between vendors who seemingly offer feature parity. And explain why sometimes it might be worth paying more for one platform than another, even though the feature set is the same, the functionality is completely different.

You can either use this to guide your thinking on the functionality you want or use it to select the right functionality for your use case.

Click here to explore it.

Tell me if this sounds like a familiar story.

You decide to migrate your community. You speak to stakeholders, create a list of requirements, and compile them into a weighted RFP. This may run from 20 to over 200 rows.

You send the RFPs to platform vendors and get responses within a week.

As you review the list, you soon discover that nearly every community platform vendor claims to do everything you want. What luck!

So you select a platform based on price or other nebulous attributes and get started.

The brutal reality begins to dawn on you…you’ve picked the wrong platform, and now you’re stuck on it for the duration of a 3-year contract.

The lesson is that there’s a gaping void between a platform vendor saying, “Yes, we definitely have that feature,” and that feature being any good.

And the problem right now is that it’s very difficult to tell the two apart because people aren’t even sure what to ask for.

FeverBee's Community Platform Feature Comparison Tool

Discover the difference between good and great platforms

Click Here

The post The Way We Compare Community Platforms Is Broken (new tool) first appeared on FeverBee - Community Consultancy.

You Can’t Fight The Tide: Why Sephora Closed The Most Totemic Brand Community Of The 2010s

Richard Millington — Tue, 12 May 2026 06:45:21 +0000

There’s been plenty of chatter the past week about Sephora closing its community.

This is sad news for anyone who’s been in the community space for a while.

For many years, Sephora was held up as the totemic example of a thriving brand community that wasn’t centered on customer support.

It’s genuinely hard to think of others (not impossible, mind, but certainly difficult).

Which is why its closure is worth considering.

After all, if the best-in-class example can’t thrive in the modern era, which community can? Does the past approach to building brand communities work in the new era?

Why The Sephora Forums Are Closing

While plenty of folks have mourned the loss of Sephora’s forums, surprisingly few have reflected on why they are closing.

(Note: FeverBee has consulted with Sephora several times over the years, and we were among those who recommended closure for the forums.)

Closing a community is almost always a last resort. You want to fix the problem first.

But there are some problems you simply can’t fix. And that’s what makes the Sephora community a totemic case for a different reason today.

Sephora Forums Should Be Thriving

If you look at sales figures alone, the Sephora community should be on a tear. Revenues are at record highs. There should be more people eager to discuss Sephora products than ever before.

And yet, it’s no secret that engagement in the forums has declined over the past five years.

Which raises one of two possibilities:

Customers don’t want to talk about Sephora products anymore.
Customers want to have those discussions elsewhere.

The first is possible, but unlikely. It’s hard to think of a reason why customers would stop wanting to talk about beauty products en masse.

This makes problem two far more likely.

The Rise Of Reddit

Whenever we face an enticing community mystery (p.s. I love this talk), it’s always helpful to look at some data. And the data in this case puts the smoking gun very clearly in the hands of third parties, notably Reddit.

Let’s examine the growth of the Sephora subreddit over the past few years. Archive.org isn’t a perfect tool, but it lets us compare snapshots of the community at different times (alas, not at identical times of year).

But taking snapshots of Reddit by year and comparing membership growth shows some rather startling results.

That’s a staggering level of growth, especially between May 2023 and Oct 2, 2024 (approximately the time when Google began diverting large amounts of traffic from other channels to Reddit).

But you can argue that figure pales in comparison to the 5m+ members the Sephora community might have. What we need isn’t just membership growth, but the level of engagement.

Let’s begin with the official community.

I can’t share the data Sephora has provided, but we can create a ‘back of the napkin’ estimate by looking at the total number of posts in the Sephora approximately a year ago (3,216,003) and subtracting that from the number today (3,184,561) and then averaging out the difference between the two dates (31442/354 days).

This gives us a post-moderated average of 89 posts per day …or approximately 621 posts per week.

As of today, May 11, 2026, the Sephora subreddit attracts 6,400 posts per week…

That’s 10x higher than the official community volume!!

That’s about as strong a smoking gun as you’re likely to see in data terms.

And this is before we factor in the post-pandemic rise of influencers, Facebook groups, and the plethora of other places where people can discuss Sephora outside of the official community.

The problem isn’t that people didn’t want to talk about Sephora; it’s that they wanted to do it in places Sephora didn’t control.

Why Are External Channels Preferred?

It’s worth looking at feedback on the closure announcement on Reddit to understand why external channels are preferred.

A (Claude) analysis put the responses into a few clear categories.

You can read this and think well, if more people knew it existed, then more people would use it.

And that’s a valid argument, but it lacks nuance. The reality is, Reddit has become the de facto place for people to discuss topics that interest them.

It overcomes the single biggest problem all non-support-centric communities face: finding a trigger to get people to visit. People are already there, and they get alerts, notifications, and prompts to visit the other subreddits they are a part of. It has a built-in trigger.

The Sephora community offered a range of unique customizations, benefits, and rewards for engagement. But ultimately, none of that came close to matching the one that members valued most: ease of use.

As I’ve written before, members are like water. They follow the path of least resistance to achieve their goals. Reddit is the path of least resistance. It’s one of those rare 5 to 7 websites people visit every day to see what’s new. Most brand communities are not.

And that’s a killer when members don’t need to visit your site to solve a problem.

Five Strategic Options

It’s good to reflect on the five strategies available to enterprise community pros right now.

Double down on AI and focus on the community being part of the AI tools your organisation is developing?
Focus on building a sense of belonging that AI can’t compete with?
Optimise what you’re already doing and ignore the noise?
Embrace ‘Community Everywhere’ and engage across multiple platforms?
Five Strategic Exit.

Of the five, I’d much prefer that Sephora pursue option four rather than a strategic exit. But that’s easier to recommend from the outside than to implement from the inside.

That path forward isn’t always viable within unique organizational structures and constraints.

So rather than renew for another three years on an increasingly abysmal community platform, a strategic exit makes sense.

Strategies From 2010 Flounder In 2026

The reality is, times have changed. Case in point, back in 2021, a former Sephora exec moved to Athleta and tried to replicate the same Sephora template there by launching a near-identical community.

I predicted at the time it would fail, which it did.

The reason is that the strategies that succeeded in the 2010s don’t work as well in the 2020s.

Times and technology have moved on – and we need to move with them.

The Crest Of The Community Everywhere Wave

A few years ago, I gave this talk about Community Everywhere.

Essentially, members’ preferences have changed, and we now prefer to meet the majority of our engagement needs on platforms outside a brand’s direct control. For many brands, especially those with non-support goals (loyalty, sales, retention, etc.), the new game was to engage in and support ecosystems comprising different groups, platforms, leaders, and more.

This approach is harder to measure and organize (it doesn’t fit neatly into most departments), but also the most aligned with current audience preferences.

And as strategists, it’s always critical to go with the tide vs. fight against it.

Now, the crest of the community everywhere is beginning to crash against some of the larger rocks in the community industry, and Sephora shutting down its forums is the result.

The Blockbuster Problem

There have been and will be some folks who cling to a purist idea of community – a place which a brand controls and where you (theoretically) own your data.

And there will be a few who will look at this Sephora news with a thought of;

“Why didn’t you try [x]?”

And, I’ll bet pretty good money. Whatever X is, it has already been considered, tried, and/or incorporated. The problem isn’t the engagement activities the Sephora team was or wasn’t doing. The problem is quite simply that the forums are on the wrong side of trends.

I remember reading an article many years ago about him closing his Blockbuster video store. Everyone had some advice about how to save the store. Video clubs, membership schemes, social media advertising, etc., and yet, if it were that easy, there would be more than one Blockbuster store in the world.

The only thing that would’ve saved Blockbuster was to acquire Netflix and change the entire business model.

There are some trends you simply can’t fight. You can either completely revamp your approach to incorporate them or admit it’s a losing battle.

The same is true here. No brand community will last forever. They rise and fall like the tide over years, potentially decades. The best run lasts for decades. And Sephora lasted for decades. That’s an incredible achievement.

Can Forum-Centric Communities Survive?

I made similar comments about forums during a debate with Marius at Higher Logic’s Superforum in Washington, DC, a few weeks ago.

I noticed in the poll responses, one skeptical attendee said

“People have been predicting the death of forums for over a decade!”

Which is valid, but also misses the point.

It’s not that binary. It’s not about death (non-existence of forums) vs. life (existence of forums).

There will always be forums in some capacity (people still use videotapes, too!), and there are plenty of degrees of nuance along the continuum between the two extremes.

It’s about the increase or decrease in their popularity and, subsequently, when and where they are the best tool to achieve our goals.

And over the past decade, we’ve seen a sharp decline in the popularity of forums for most engagement-related activities.

For customer support, forums are fantastic. They may even see an AI-driven revival as authoritative and useful sources of information.

But for almost any other needs, the Community Everywhere wave is about to crash upon them and carry the remnants out to sea.

The Community Everywhere Era Is Here To Stay

If you want a simple takeaway from all of this, it’s this.

You need to keep your community fully aligned with your audience’s needs and preferences. And when those needs and preferences shift, so does your approach. That often means changing the technology, goals, or approach. But the moment you try to cling to what you have, seek control, or try to fight against the changing tide is the moment your community begins its long journey towards irrelevance.

And if you want help knowing if your community is aligned with your audience’s needs and preferences, drop us a line.

Good luck!

The post You Can’t Fight The Tide: Why Sephora Closed The Most Totemic Brand Community Of The 2010s first appeared on FeverBee - Community Consultancy.

Stop Letting Defaults Dictate Your Community’s Design

Richard Millington — Mon, 04 May 2026 20:54:17 +0000

I’ve now sat through hundreds of UX sessions watching members navigate enterprise communities. The same pattern surfaces every time.

There’s a wide gap between how members say they want to engage in a community and how the homepage is actually set up to let them. The community team isn’t usually at fault. The platform vendor’s professional services team configured a template, the homepage went live, and no one has questioned the defaults since.

So to save some time, I mocked up what a modern enterprise community homepage should look like:

Notice the seven things on this page. Each one replaces a default that’s quietly costing communities their members.

1. A powerful, value-based message

Most homepages open with “Welcome to the [Brand] Community” or some variation of “Connect, share, learn.” This tells the member nothing about why they should be here rather than on Reddit, Stack Overflow, or a Slack group their colleague set up.

The message should make the unique value of this specific community obvious in one line. What can a member do here that they can’t do anywhere else?

The homepage messaging should generally answer these community positioning questions.

2. Task-based search

A typical community search returns 25 to 50 threads sorted by recency. The member then has to read through them to find the one answer they need. Half the time, they give up and post a duplicate question.

Modern RAG-based search should be doing the work the member is doing manually.

It should synthesize the answer from the best threads, surface the source posts underneath, weight recent and authoritative content above stale ones, and let the member ask a question in natural language.

The technology has been available for two years. It’s about time it was embedded in your community.

3. A small number of trending topics

On most community homepages, there’s no navigation help at all. The member lands on the page and has to figure out where to click to find anything relevant.

The category tree is buried, the topics are alphabetical, and the discovery experience is “scroll until something catches your eye.”

A small, curated set of trending topics solves this. You ideally want five to seven topics, refreshed daily, surfaced prominently.

It guides 80% of members to the location they want to go to most. It tells the member what the community is talking about right now and gives them an obvious entry point.

4. Value-based engagement stats

Plenty of community homepages display vanity metrics. “47,000 members. 380,000 posts. 1.2 million replies.”

These numbers mean nothing to a member trying to decide whether this is the best place to solve their problem and achieve their goals.

Replace them with stats that signal value. This includes questions answered in the last 24 hours, average response time, active superusers online right now, problems solved this week etc. The member is asking, “Will I get help if I post here?” The homepage should answer that question in the data it shows.

5. An AI summary of the past 24 hours

This isn’t a radical idea, but it hasn’t been executed well yet.

Regulars are the engine of every healthy community. They want to know what they missed since their last visit without scrolling through 80 threads to find out.

A short AI-generated digest of the most important discussions from the past 24 hours, with links to the full threads, gives them exactly that. It takes ten seconds to read and pulls them straight back into the activity that matters. Every platform has the technology to build this. Almost none of them have shipped it.

6. A clear path to becoming a superuser

Superuser programs exist in most enterprise communities, but the path to becoming one is not visible on the homepage. It’s buried in a help doc, gated behind an internal nomination process, or restricted to people the community manager already knows.

The homepage should make the path visible. “Here’s what superusers do, here’s how you become one, here’s what you get.” Communities that surface this generate a self-selecting pipeline of high-intent members who want more responsibility.

Communities that hide it rely on the community manager’s personal knowledge of who’s been active lately.

7. A clear list of what members can actually do here

Almost every community offer the following actions: ask a question, share work, find an expert, and join an event. The actions just aren’t visible on the homepage. They’re tucked into a navigation menu or surfaced through a “+New Post” button that doesn’t tell the member what kinds of posts they could make.

The homepage should make the available actions explicit. For example, a short panel with three or four primary actions, each with a one-line description.

It removes ambiguity and gives the member permission to participate in a specific way.

Stop Relying on Defaults For Your Community Design

Too many enterprise community homepages are designed by the platform vendor’s professional services team. That’s not a design discipline. It’s an implementation discipline.

The result is a homepage that reflects what the platform does by default rather than what the community needs. The vendor has no incentive to question whether the design is right for the audience, because their job is to launch the platform on time and on spec. Once it’s live, the community team inherits a homepage that nobody really designed, and the defaults stay in place for years.

This is why we keep having poorly designed community homepages.

Get a community design that actually works

If you’re staring at a homepage you didn’t design, that doesn’t reflect how your members want to engage, and that your platform vendor isn’t going to fix, that’s the kind of problem FeverBee solves.

We can undertake a community design project (est. cost $20k to $30k) that includes:

Community Experience Diagnostic: a scored audit of your current experience against 40+ design principles, benchmarked against the top quartile of enterprise communities we’ve assessed.
Member Behaviour Analysis: analytics, session recordings, and structured interviews with newcomers, regulars, and lapsed members to surface the gap between how members say they engage and how they actually do.
Vendor Capability Assessment: an objective evaluation of what your platform can and can’t deliver, separating platform limits from configuration choices your vendor hasn’t surfaced.
Community Design Index Score: a proprietary score against our benchmark of enterprise communities, giving you a number you can take to leadership.
RAG search architecture spec: the technical specification for how your search should be configured, weighted, and integrated with your knowledge sources.
Information architecture for the full member journey: wireframes for the homepage plus 4 to 5 key templates, mapped against newcomer, regular, and superuser behaviour.
Onboarding flow design: the sequence of prompts, notifications, and CTAs that turn newcomers into regulars across their first session, first week, and first month.
Implementation brief: a document your platform team or vendor can build from directly, removing the ambiguity that lets implementations stall.

If you’re interested, contact us or reply to this email.

Good luck!

The post Stop Letting Defaults Dictate Your Community’s Design first appeared on FeverBee - Community Consultancy.