--- title: The science of scaling according to Mark Roberge description: "It still happens. A founder raises a $10 million Series A and a board member says, “Hire 20 reps in November.” The founder is 30 years old, has never hired a rep, and the company barely has a repeatable process. Yet this is still how many startups scale, with no real evidence they’re ready." canonical: "https://vasco.app/blog/the-science-of-scaling-according-to-mark-roberge" date: "2025-10-11T00:00:00.000Z" author: Mark Roberge jobTitle: "Co-founder at Stage 2 Capital, prof at HBS, founding CRO at Hubspot" readingTimeMinutes: 3 contentType: article intent: strategy-insights pillar: "RevOps maturity & scaling journeys" --- # The science of scaling according to Mark Roberge _From product-market fit to go-to-market fit, led by retention._ ### After working with hundreds of post-seed companies, I’ve seen the same pattern over and over. The winners treat scaling like a science. The rest confuse early momentum with readiness. So let’s talk about when you’re actually ready to scale. ## Product-market fit is not a feeling Ask a room of founders what product-market fit means, and you’ll get a hundred different answers. Some say $500k in ARR. Others say “six happy customers.” Or a steady stream of inbound leads. Those are good signs, but they measure _market-message fit_, not product-market fit. Being good at sales or marketing can get you contracts, even if customers don’t truly need what you sell. That’s selling ice to Eskimos. If you want a quantifiable definition, it’s simple: **retention.** When customers stay, expand, and get ongoing value, you’ve built something real. I like to see **net dollar retention above 100 %** before declaring product-market fit. Otherwise, you’re just filling a leaky bucket. ## The problem: retention is lagging In early-stage SaaS, you can’t wait twelve months to see who renews. You need a _leading indicator_ of retention, a way to know today if customers will stick tomorrow. Here’s the framework I use: > **P % of customers do E event every T time** Three variables: - **P** = the percentage of customers - **E** = the key event that signals value - **T** = the time window That’s your **Leading Indicator of Retention (LIR).** ## Examples of LIR in action - **Slack:** 70 % of customers send 2,000 messages per month. - **Dropbox:** 85 % of customers back up their device every day. - **HubSpot:** 80% of users adopt 5 or more features in a 25-feature platform. ![](https://cdn.sanity.io/images/ys8gstp8/production/ce4a1059d612c7c1b82367a0f8db8f1bbec2f0df-1920x540.png?w=1600&fit=max&auto=format) Those companies didn’t set revenue goals like “hit $1 million ARR.” They set _usage_ goals like “get 70 % of customers sending 2,000 messages.” That second goal builds a foundation you can actually scale. You don’t need regression analysis to start. Just track, cohort by cohort, what percentage of new customers hit your event in month 1, 2, 3… and keep pushing that number up. When it climbs and stays there, you have real product-market fit. ## From product-market fit to go-to-market fit Even then, you’re not ready to scale yet. Product-market fit proves that customers find value. **Go-to-market fit** proves you can deliver that value _profitably and repeatedly._ The metric for that is **unit economics**: your CAC, LTV, and payback period. But again, those are lagging indicators. You need to _extract_ them into variables you can measure today: average deal size, close rate, sales cycle, cost per lead, rep ramp time.If the algebra behind those numbers works out to a healthy LTV : CAC > 3, you’re on the right path. ![](https://cdn.sanity.io/images/ys8gstp8/production/2d75db2dbbe1f18d3f0fc8c3ab1f63f4aab2b875-1920x1080.png?w=1600&fit=max&auto=format) The key is sequencing. Work on product-market fit first, then go-to-market fit. If you optimize both at once, you risk building a repeatable motion on the wrong market. ## Scaling is a pace, not an event When both fits are in place, the next question is speed. How fast do you scale? Most companies treat it like a light switch: raise capital → hire 20 reps. That’s not scaling, that’s gambling. Think of it as _pacing._ Maybe start with two reps every other month. Watch your leading indicators of retention and unit economics. If both stay green for six months, double the pace. If they turn red, slow down, fix it, then accelerate again. Those metrics become your **speedometer. **Many startups only realize they were going too fast when churn hits nine months later. You’ll know nine months _earlier._ ## How to operationalize the science of scaling 1. Define your LIR (the single behavior that predicts retention). 2. Instrument it in your product logs. 3. Track cohorts monthly until the trend line turns upward. 4. Back-solve your unit economics into present-day controllables. 5. Scale gradually, using your LIR and unit-econ dashboards as the green light. Do that, and you won’t need to argue with your board about whether you’re ready to scale. You’ll have data that speaks for itself. ### Final thought Product-market fit proves you’ve built something people need. Go-to-market fit proves you can deliver it efficiently. Scale only when you have both, and let retention, not revenue, tell you when that day has come. ## FAQ ### What does “the science of scaling” mean? The science of scaling means treating growth as a disciplined, data‑driven process rather than a one‑off hiring spree. It’s about proving product‑market fit, then go‑to‑market fit, and finally scaling at a pace your retention and unit economics can sustain. ### How is product‑market fit different from market‑message fit? Market‑message fit means people respond to your messaging or sales motion; product‑market fit means customers actually stay, expand, and get ongoing value. A quantifiable sign of product‑market fit is net dollar retention above 100%. ### Why is retention the best measure of product‑market fit? Retention is the best measure because it shows whether customers truly need what you sell. If they renew, expand, and keep using the product, you’ve built something real. Revenue alone can be misleading if churn is high. ### What is a Leading Indicator of Retention (LIR)? A Leading Indicator of Retention (LIR) is a key behavior that predicts whether customers will stick. It’s defined as: P% of customers do E event every T time, where P is the percentage, E is the key event, and T is the time window. Track this cohort by cohort to see if it improves over time. ### How can I define my own LIR? Define your LIR by identifying the behavior that signals real value for your product. For example, Slack’s LIR is 70% of customers sending 2,000 messages per month; Dropbox’s is 85% backing up daily. Track this event for new cohorts and push the percentage up. ### What is go‑to‑market fit, and how is it different from product‑market fit? Go‑to‑market fit means you can deliver your product profitably and repeatedly. Product‑market fit proves customers find value; go‑to‑market fit proves you can acquire and serve them efficiently, with healthy unit economics (LTV:CAC > 3). ### How do unit economics support the science of scaling? Unit economics support scaling by turning lagging indicators (CAC, LTV, payback) into present‑day controllables like average deal size, close rate, sales cycle, cost per lead, and rep ramp time. If the math works out, you’re on the right path to scale.