Your B2B freemium tier is selecting for the wrong customers

At some point in the life of most B2B SaaS companies, someone makes the case for a free tier. The reasoning arrives pre-packaged.

A free plan lowers the sign-up barrier. It creates a top of funnel that doesn't require a salesperson. It lets users experience the product before committing. And if a small fraction upgrade, you've acquired paying customers at near-zero sales cost. The maths looks defensible: B2B freemium products typically convert 2–5% of free users to paid plans. Drive 50,000 sign-ups, and that's 1,000 to 2,500 paying customers without a single sales call.

Then you build the free tier. And things go wrong in a way the spreadsheet didn't predict.

The conversion rate is not the problem

Every retrospective on a failed freemium tier leads with conversion rate. The free tier attracted 80,000 users. Only 1,200 paid. Infrastructure costs were substantial. Support was stretched. Conversion rate was 1.5%, against the 3% the model assumed. The free tier was shut down.

This analysis is accurate as far as it goes. B2B freemium conversion rates cluster at the low end of the 2–5% range, and the infrastructure and support costs are real. Notion and Slack spent years cross-subsidising free users before their paid bases could absorb that overhead, and they had distribution advantages most B2B products don't have.

But the conversion rate is the obvious failure mode. Founders who've shut down a free tier can point to it and feel they've explained what happened. They usually haven't. The conversion rate is a symptom. The cause runs deeper.

The feedback problem

Here is what happens when 97 of every 100 users are on a free plan.

Your NPS surveys get answered by free users. Your support tickets come from free users. Your feature requests come from free users. Your product team sees which features get used most, and most of the usage comes from free users. The people who submit detailed feedback, who join your community, who say yes to user interviews: they are free users.

This is not a support overhead problem. It is a signal problem.

Free users in a B2B product have a specific profile. They are often individuals or very small teams who cannot justify a paid plan at your price point. They are price-sensitive by definition: they chose the free tier because they were not willing to pay yet, possibly not ever. The features they value most are the features that let them stay free. The limitations they complain about most are the limitations that create pressure to upgrade.

When your product team acts on this feedback signal, the roadmap bends toward the 97%. Paying customers, the 3% who represent 100% of revenue, find the product increasingly misaligned with their actual needs. Churn from the paid tier rises. The company doubles down on improving the free product to drive more sign-ups, hoping to move the conversion needle. This accelerates the distortion. It is a feedback loop that takes about eighteen months to fully recognise, by which point a lot of engineering effort has been spent on the wrong problem.

What free users actually want

A free user in a B2B product typically falls into one of three groups.

The first group is evaluating. They want to understand whether the product fits their needs before spending money. These users convert reasonably well and relatively quickly, within 30 to 60 days of signing up. They are the users freemium is designed to attract, and they are not the problem.

The second group is permanently price-constrained. They like the product. They use it regularly. They will not pay, because the product doesn't solve a problem painful enough to justify budget allocation at their organisation. These users look good on engagement metrics (daily active rates, feature adoption breadth) and are economically worthless to you.

The third group is passing through. They signed up because the friction was low, haven't logged in since the first week, and won't return. They inflate your sign-up numbers and skew your funnel metrics.

The problem is that your evaluators (the people who will convert) need a different product experience than your permanent free users. Evaluators want to see what the paid plan actually does. Permanent free users want a comfortable resting place just below the paywall. You cannot optimise for both simultaneously. Most companies, under pressure from headline sign-up numbers, optimise for the second group without meaning to.

The pricing signal

There is a subtler problem that takes longer to notice.

When you offer a permanent free tier, you tell the market that your product has zero value at the entry level. Consider what that means in a procurement conversation. A potential customer's finance team asks: 'Why are we paying £200 per month for this when there's a free version?' Your sales team explains the limits of the free plan. Sometimes they win. Sometimes the answer is: 'Can we just use the free version for now?' This often means indefinitely.

Enterprise buyers, who require reassurance that software is worth budgeting for, read the free tier as a market-positioning signal. A free tier says the product is an individual tool that people try before buying. An enterprise-grade product doesn't typically position itself that way.

Removing the free tier later is painful. Users who signed up free feel entitled to remain free. You get public complaints, negative reviews, and the occasional social media escalation. Qualaroo discontinued their free plan and moved their entry point to $100 per month; the founder was direct about why: the free tier was anchoring the product at a price point that didn't support their revenue objectives. That transition is achievable, but it's harder than not building the tier in the first place.

When freemium actually works

Freemium is not universally wrong. It works under a specific set of conditions, and those conditions are rarer than the marketing suggests.

The product needs genuine virality, not manufactured virality through referral programmes, but natural spread where one user's activity creates value or visibility for another user who doesn't yet have the product. Slack worked partly because a message in Slack is useless if the recipient isn't on Slack. Figma worked because a design file shared externally required the reviewer to open Figma. The product itself recruits new users as a side effect of being used.

The upgrade path also needs to be intrinsic. Not 'you've hit the five-project limit, upgrade to continue' but 'the features you need for your team's actual workflows are in the paid plan', where the user naturally grows into paid usage as their use case matures. And the addressable market needs to support the volume that makes the conversion maths tolerable. At 3%, you need tens of thousands of active free users to build a meaningful paid base. Most B2B products serve a narrow vertical. They will not reach that volume.

What to do instead

The alternative is not removing free access entirely. It is a constrained trial: time-limited access to the full product, or usage-limited access with a genuine ceiling.

The difference matters. A time-limited trial, 14 to 30 days, creates urgency without creating a permanent underclass of free users. The evaluator experiences the paid product, not a stripped-down version of it. When the trial ends, they convert or they don't, and there is no ambiguity. There is also no persistent population of free users shaping your feedback signal toward their own needs.

Usage-based access, limited to a certain number of documents, API calls, or team members, can work if the limits are calibrated so evaluators can complete a genuine assessment before hitting the ceiling, but strict enough that ongoing use requires payment. This is harder to calibrate than a time limit, but it preserves some of the discovery value of a freemium model. Either approach avoids the feedback distortion that a permanent free tier creates, because neither creates a large, stable population of users whose interests are structurally opposed to your pricing model.

One question before you build the tier

If you are still considering a free tier, ask this before you build it: which of your current paying customers would have stayed free indefinitely if you had offered them that option?

If the answer is 'most of them,' the free tier is a revenue reduction, not an acquisition channel. If the answer is 'none: they converted within 30 days because the problem was urgent,' then a time-limited trial will do the same job without the permanent overhead.

The conversion rate will look worse on a trial than on freemium. That is the point. Users who don't convert in 30 days probably weren't going to convert in 300. Better to know that at the start of the funnel than after eighteen months of building for the wrong audience.