We launched our SaaS with a very generous free tier. We wanted to get as many people using the product as possible to build word-of-mouth. It worked our free user base exploded by 300% in six months, and our brand awareness grew massively.

But I just finished our quarterly financial review, and the numbers are terrifying.

While our paid revenue grew by 15%, our support costs grew by 60%. When I dug into the data, I found that 80% of our support tickets were coming from free users. They were asking for custom features, requesting API access, and expecting 1-on-1 onboarding calls for a product they were paying $0 for.

We are essentially running a free help desk for people who have zero intention of ever upgrading. Our small support team is burned out, our response times for paying customers are slipping, and our margins are shrinking because we have to pay for more support staff just to handle the free crowd.

I asked Claude Code to rename a function across the codebase. It found 12 files, made the change, and reported success.

A thirteenth file sat in a directory outside its search scope. No warning.

An agent wrote a fix for a bug and included a passing test. I was suspicious, so I reverted the fix and ran the test again.

It still passed. The assertion was checking a mock that always returned true, regardless of the code.

We bootstrapped our SaaS to $30k MRR with a very simple pricing model. But last year, we started landing larger mid-market clients. These clients all had one thing in common: they refused to use our automated Stripe checkout. They demanded to be billed via custom invoices, with net-30 terms, purchase order (PO) numbers, and specific line items for their procurement departments.

Because we desperately wanted the revenue, we said yes to every custom billing request. We didn't want to spend engineering time building an enterprise billing system, so my co-founder and I just did it manually.

I finally tracked our time last month. We spent 40 hours (an entire work week) manually generating PDFs, chasing down PO numbers, emailing finance departments, and reconciling payments. We were acting as our own collections agency.

Our CFO (who is our part-time advisor) looked at the numbers and pointed out something terrifying: If we factored in the opportunity cost of the engineering hours we weren't shipping, and the delayed cash flow from net-30 terms, we were actually losing money on those enterprise deals. We were so focused on the top-line revenue that we ignored the massive operational drag.

Looking at our analytics last week, a massive chunk of new signups drop off right at step two of our interactive setup wizard even though it only takes sixty seconds to complete.

we thought adding helpful tooltips and step-by-step guides would make onboarding seamless, but it seems like every extra click just increases overall user friction.

how do you audit your first-time user experience to figure out what is actually causing drop-offs versus what is just passive friction?

I work with B2C products most of the time. I have experience with B2B in sponsorship.

But I can t compare it most comprehensively.

That s why I think some of you have more relevant experience and can compare your work in B2C and B2B.

Adam Cohen is the cofounder of @Weave , an engineering analytics company that helps teams go AI-first. Before Weave, he was head of operations and sales at education software firm Top Hat and VP of operations and revenue at @Causal , which was acquired by Lucanet in 2024.

Someone bought my product a year and a half ago. Paid once. Never upgraded. Never asked for anything.

Then they emailed me out of nowhere with a detailed explanation of a problem they'd solved with it. Something I'd never thought it could do. They'd built a whole process around it and wanted to tell me.

Quick one for anyone doing outreach.

I've spent a lot of years around CRMs and email setups, and this comes up every single time someone starts cold emailing. Send from your main domain, or buy a lookalike domain just for outreach?