It always starts the same way.

A whale prospect asks for “one quick add.” Sales swears it’s tiny. Engineering shrugs. Product says “we’ll hide it behind a flag.” You take the deal.

That’s the first shovel of dirt on your own grave.

Here’s the pattern I’ve seen—over and over—running Concord and watching friends build great, then bloated, then dying products.

The pattern nobody talks about

1. The ‘quick add’
   You bend the roadmap for one logo. You tell yourself it’s “strategic.” It’s not. It’s a seed that doesn’t fit your soil.

2. The second order promise
   Now support needs a toggle. Docs need a page. RevOps needs a field. QA needs a matrix. Cost compounds. Value doesn’t.

3. The identity wobble
   Your product stops being opinionated. You’re the Cheesecake Factory menu: something for everyone, great at nothing. (I used to joke about this exact trap; it’s real.)

4. The slow death
   Onboarding drags. Demos turn into tutorials. Dev speed crawls because every change trips five edge cases. Churn rises quietly—users don’t fight you; they just stop logging in.

5. The rationalization
   “We’re moving upmarket.” No, you’re moving sideways to avoid admitting you built for an imaginary customer.

6. The cash illusion
   Those big checks mask the growing tax: support load, roadmap debt, sales commitments you regret on day two.

7. The cliff
   A competitor ships the simple, obvious version. Your team is too entangled to respond. Game over.

I’ve made these mistakes. Then I did the uncomfortable thing: we said no—a lot.

What we killed at Concord (even when customers asked)

Here’s what we actually did, not theory.

• We stopped building enterprise-only detours.
  Post-COVID, we were split across SMB, mid-market, and enterprise. That was our Cheesecake Factory era. We picked a lane and cut entire enterprise-specific developments—even ones already in progress. That decision saved the company.

• We turned off low-value “power” features for new logos.
  If a feature added complexity without moving the needle for our core users, it was gone for new customers. Existing customers kept it; we just stopped spreading the problem. Less to learn, less to break, faster to ship.

• We told big logos “we won’t build that”—and helped them leave.
  When an enterprise asked for something that pulled us off-mission, we declined and even helped them transition to a better-fit vendor. Sounds insane… until you measure the drag avoided. Some stayed anyway because they valued simplicity over “enterprise-grade” bloat.

• We rebuilt our go-to-market to match the product.
  Transparent pricing, self-serve trials, fewer hoops. If you sell to SMBs and mid-market, hiding pricing and locking demos behind four calls is self-sabotage. We stopped acting like an enterprise vendor.

And inside the product? My job became 80% subtraction. We trimmed or hid features that 1% of users loved but 99% tripped on. “Less is more” isn’t a slogan; it’s an operating system.

The non-obvious reason this works

Here’s the thing: most customers don’t want your edge cases. In our world (contracts), more than 90% of agreements on Concord are signed with zero negotiation. Read that again: no edits at all. You don’t need a cathedral of options to serve the vast majority of use cases. You need a paved road.

That’s also why CFOs—not legal—now own contracts in many companies. Contracts behave like financial processes; they should be templatized and fast. Complexity helps almost no one.

I wrote about this math in detail elsewhere: the courage to standardize (and to say no) beats the fantasy of infinite flexibility. Speed and clarity win.

“But we’ll lose revenue.”

Sometimes, yes. And that’s exactly the point.

Saying no to misaligned revenue is how you create durable revenue. Focus compounds:

• Happier users → higher activation and NRR

• Simpler code → faster shipping and fewer bugs

• Straighter stories → shorter sales cycles and fewer discounts

• Cleaner ops → lower support costs and fewer escalations

Focus also gives you spine. Once we picked our lane, “No” got easier—and more believable. The team moved faster because they weren’t living in fear of the next one-off.

A quick diagnostic for founders

Ask your team these five questions. If you can’t answer “yes” to at least four, you’re already in feature-creep territory:

1. Could a new PM explain your ICP and top 3 jobs-to-be-done in one minute, without slides?

2. Can a first-time user activate in under 10 minutes without a call?

3. Does every shipped feature have a measurable adoption target—and a kill switch if it misses?

4. Are you comfortable recommending a competitor when a request doesn’t fit your lane?

5. Do you remove at least one feature (or workflow step) for every two you ship?

If that felt spicy, good. This is where companies are built or buried.

How to say no (without burning bridges)

• Name your lane in public. Put your ICP and “we are not for” list on the website. It’s easier to say no when you’ve already said who you are.

• Offer the “why” and an alternative. “We won’t build X because it hurts onboarding and slows everyone else. If X is critical, Y vendor is excellent.” People respect clarity.

• Instrument “no.” Track saved engineering hours and reduced support load from declined requests. Show your team (and your board) the compounding win.

• Make subtraction a ritual. Quarterly “de-feature” reviews. Adoption thresholds. Sunset plans. The garden needs tending or it becomes a jungle.

The uncomfortable truth

I used to chase every “maybe” dollar. It almost killed us. When we finally cut the bloat, picked our segment, and turned off the noise, everything got easier—product velocity, customer love, even morale. We didn’t shrink; we sharpened.

And yes, a few logos walked. That’s the trade. But the ones who stayed? They’re exactly who we built for. Our roadmap finally looks like a roadmap again, not a ransom note of exceptions.

Courage is a product strategy. The courage to say no to revenue today, so your company still exists tomorrow.

If you’re staring at a “quick add” right now, you already know the answer. Say no. Build the paved road. Let someone else own the gravel.