Build & Grow Your Way
To More MRR

Jason's core point was simple: obsess about churn. The reason behind it, the one I kept circling back to in my notes: churn rate compounds with size, new MRR doesn't. 5% churn costs you $500 a month at $10.000 MRR – and $5.000 a month at $100.000 MRR. Same rate, 10× the real dollars gone. Whatever you're adding in new MRR each month, by contrast, is roughly the same whether you're at $10k total or $10M.

So every point of churn you remove stays removed as you scale – and gets more valuable the bigger you get. Every growth channel you open has the opposite property: a half-life. It saturates, gets noisier, gets more expensive, needs constant re-investment. You earn growth over and over. You don't earn retention twice.

That's why Build comes first. Not because it matters more. Because you can't out-grow a leaky bucket.

Four moves from MicroConf that actually lower churn:

Watch the churn signals before they leave. Jason's list of leading indicators: going dark, never completing onboarding, too much time on basic tasks, switching from annual to monthly, exporting data, unsubscribing from notifications, ignoring error alerts, failing an invoice. These are the moments to intervene – not the day someone hits the cancel button. Data & analytics is where you catch them.

Price for retention, not just acquisition. Alex Pham – formerly at Clio and Shopify – ran his C.R.E.A.M. playbook on raising prices without blowing up your base. The headline number: a 10% price increase with 5% logo churn equals roughly $1M of enterprise value at a 5× multiple, on a $4M ARR business. No new customers, no new features. Just a price you should've raised already.

Operational excellence as a silent moat. Anthony Eden at DNSimple: reliability prevents the slow, quiet churn that never shows up in a support ticket. Monitoring, testing, clean defaults, boring uptime. Not glamorous, but the bucket stays sealed.

Save fifteen minutes, not two. Rob Walling's rule for AI features: if a feature doesn't save the user 15+ minutes of real work, it's a sparkle icon. Real retention comes from a product that keeps delivering time back to the user – ingestion, categorization, analysis, the kind of work they used to dread.

What these moves share: they all live inside the product or the billing system. They don't decay. They compound. That's the foundation half.

Grow is the half I've been teaching for years. None of what follows makes it smaller or less important. It makes it honest about its nature.

Channels saturate. Creatives fatigue. Algorithms shift. Whatever brought you this month's new MRR will work a little worse next month and need more of your attention the month after. That's not a flaw – it's how acquisition works. You stay in motion by earning it again.

Five moves from MicroConf that actually raise new MRR:

Position as a painkiller, not a vitamin. Gia Laudi's diagnostic: most growth problems are messaging problems, most messaging problems are positioning problems, most positioning problems are customer understanding problems. In a saturated market – and 2026 is saturated – only painkillers cut through. If you can't name the 2am problem you solve, the loudest objection you face, and the specific dream you deliver, you're a vitamin.

Write the homepage like a trustworthy colleague. Nick Disabato runs his homepages through the Pain-Dream-Fix frame (originally Amy Hoy's): start with the pain people feel before your solution, move to the dream of what life looks like without it, then the fix. Talk about the customer – “you” far more than “we.” Outcomes before features. The Content & Story chapter goes deeper.

Publish for zero clicks. Amanda Natividad's SparkToro data: 58.5% of Google searches end in zero clicks. Platforms suppress external links – posts without them get up to 10× more reach. Her rule: 5 deposits of native value for every 1 withdrawal (a CTA, a subscribe). Starve the algorithm and it starves you back.

Price into a better market. Jason Cohen again, on the Grow side this time: find the segment that loves you most, and optimize everything for them. Tweet Hunter went from $9 to $49/month by focusing on business owners who wanted results, not spenders who wanted to save $20. Same product. Different customer. Different ceiling.

Mind the growth rate if you ever want to exit. Einar Vollset – General Partner at TinySeed alongside Rob Walling – laid out the valuation math. 20% YoY growth is the cliff: drop below it and your multiple can fall from 4× ARR to 1–2×. If an exit is on your horizon, growth rate may matter more than Max MRR itself.

These moves keep new MRR honest. Pick one, ship it, measure, pick the next.

Back at the hotel in Portland, I wrote down three lines:

Build lowers churn – raises one half of the ceiling, permanently.
Grow adds new MRR – raises the other half, continuously.
Repeat runs both halves at real output without growing the org chart.

The first two are the ratio. The third is what makes it reachable for a small crew. I wrote about what that looks like in the Augmented-Human Company piece: documented processes running as agent skills, Claude and Codex multiplying output, autonomous agents like OpenClaw doing work a teammate would do – with a human in control of guardrails, quality, and soul. That's the Repeat layer. It's what lets two or three people genuinely operate both halves of the formula.

What surprised me most at MicroConf wasn't a single talk. It was how little AI came up on stage. At an event full of SaaS operators in 2026, I expected half the programme to be “here's how AI changes everything.” Instead, speaker after speaker walked us through churn math, pricing levers, positioning, homepage copy, customer interviews – the fundamentals of running a SaaS business the way they've always worked. AI was in every hallway conversation. Almost nowhere on stage.

Almost. Craig Hewitt of Castos closed the conference with SaaS in the Agent Economy – a talk that felt like a transcript of the last twelve months of our Vibe Coding Cologne meetups. The agent-readiness checklist (working API, clean docs, MCP server, CLI, pay-per-outcome pricing). Tokenmaxxing culture. One shared home for team prompts, skills, and agents. Everything the builders in our community have been quietly doing finally got the main-stage summary it deserved.

That re-grounded me. The Max MRR formula is older than the AI wave and will outlast it. Whatever tools you use to move the two halves, the ratio itself is the thing. Jason's slide would've held in 2015 and will still hold in 2035. What vibe coding changes is only my reach – not the underlying physics.

So here's what I'd do if I were you, right now, today. Scroll back up. Plug your real numbers into the calculator. Look at your ceiling. Ask the honest question: which half am I on? Churn compounds with size, so the answer probably tilts toward Build – but the math on the screen is more useful than my guess. Pick one move. Ship it this week. Next week, do another.

That's the whole job.

Cheers,
Ben