Most SaaS teams choose their free model the way they choose a pricing page layout: they look at three competitors and split the difference. Free trial, because everyone in the category runs a free trial. Freemium, because the category leader runs freemium. It’s a copy of a copy, and it’s the most expensive default in product-led growth — because the model you pick decides the shape of every conversion mechanic you build on top of it. Get it wrong and you spend the next two quarters tuning email sequences on an instrument that was never going to play.
The choice isn’t a matter of taste. Four questions decide it, and a team can answer all four in about ten minutes with everyone in the room.
The four axes
Wes Bush’s MOAT framework is the cleanest version of this I’ve found, and I use it as an opening exercise in most engagements. Four dimensions, each a continuum, each scored independently:
- M — Market strategy. Differentiated on the left, dominant on the right. Are you winning because the product does something others don’t, or because you’re the biggest name in a crowded room?
- O — Ocean conditions. Blue ocean on the left, red ocean on the right. Is the buyer choosing whether to solve this problem at all, or choosing between you and six alternatives they already have open in other tabs?
- A — Audience and adoption. Top-down on the left, bottom-up on the right. Does the product enter through a decision-maker who signs, or through an individual contributor who drags their team in behind them?
- T — Tech-savvy ICP. More savvy on the left, less savvy on the right. Can your user self-serve their way to a result, or do they need to be walked?
There’s an interactive version of this scorer on the site if you want to run it with your team in the room: the MOAT Self-Score. Four sliders, two minutes, one answer.
Score where the product genuinely sits, not where the deck says it sits. That distinction is the entire value of the exercise. Every founder’s instinct is to mark “differentiated, blue ocean” because that’s the story they tell investors. The scoring only works if someone in the room is willing to say the honest version out loud.
Then read the cluster. All four hard left — differentiated, blue ocean, top-down, technical buyer — and you want a free trial. Sell the full experience for a finite window and let the deadline close. Trials post the highest signup-to-paid rate of any model, typically 15–25%, and requiring a credit card lifts that by 3–5× because it filters for intent before the clock starts.
All four hard right — dominant, red ocean, bottom-up, less-savvy users — and you want freemium. Scale on free, monetize the long tail, and let habit do what urgency can’t. Freemium products run around 25% sustained engagement, roughly 15% higher net revenue retention, and about double the NPS of trial-based peers. Users who graduate from a free tier are habit-converted rather than urgency-converted, and they churn less because the product is already load-bearing in their week.
Like this? I send one idea like it most weeks. One PLG idea a week → subscribe
The middle is where most companies actually live
Here’s the part teams don’t expect: most scores don’t cluster at either end. A company under $10M ARR is rarely dominant and rarely as differentiated as it believes. It usually sells into a category that already exists, enters bottom-up through a practitioner, and serves users of mixed sophistication. Three of four scores land mid-spectrum.
That’s the reverse trial: every signup gets the premium plan from day one on top of a real free tier underneath. When the window closes, the user isn’t comparing free against paid in the abstract. They’re deciding whether to give up features they’ve already built a workflow around. Loss aversion does the closing, and you never have to run the education campaign that explains what’s in the paid tier — they’ve been living in it.
The numbers are the reason it keeps winning the middle: conversion around 15%, matching a free trial, with sustained engagement around 25%, matching freemium. Airtable measured roughly a 20% activation lift when they moved to it. Notion, Loom, Miro, and Superhuman all convert this way today. It isn’t a compromise between two models — it’s the model that fits the position most companies are actually in.
What the wrong model looks like from the inside
A B2B SaaS team I spoke with this month had strong organic top-of-funnel — roughly 60 engaged leads a week with no paid spend — and a conversion problem nobody could name. The onboarding flow was clean. The product worked. The tell came out sideways, in how they described their paywall: everything rested on one report having to justify the upgrade by itself.
When a single artifact has to carry the entire buying argument, the model is usually wrong, not the artifact. Their four scores clustered in the middle. The fix wasn’t a better report — it was exposing every new user to the premium tier for the window, then removing it. You stop having to teach people what they’re missing when they’ve already had it.
A second client, further along, runs the same shape deliberately: a 14-day reverse trial with a genuine free tier underneath, and an internal expectation that about 70% of accounts will stay on free indefinitely. That number reads like failure on a dashboard. It isn’t. It’s the model working as designed. The job on that 70% isn’t a heavier upgrade nag — it’s watching for the behavioral signals that say an account has outgrown free, which is exactly why product-qualified leads outperform marketing-qualified ones. Free tiers aren’t a leak in that design. They’re the top of a longer, better-qualified pipeline.
Model first, then the gate, then the first session
Most teams work this sequence backwards. They pick a model by default, then argue about where the paywall goes, then try to fix the first session with onboarding tweaks when conversion doesn’t move.
Run it in order instead. The model determines what free is for. Only then does where you place the value gate become a solvable question — a free trial gates on time, freemium gates on units of value, a reverse trial does both in sequence. And only then is it worth engineering the first sixty seconds so the user reaches the First Value Moment before any gate is in play at all.
If three of your four scores land mid-spectrum and you’re running a straight free trial, the gap between the model you have and the model that fits is almost certainly the largest conversion leak in your funnel — bigger than any email sequence or onboarding change you could ship this quarter. It’s also the one nobody audits, because the model was decided once, early, by someone looking at a competitor’s pricing page. Re-deciding it deliberately is where I start most trial conversion work. And the model is only the first of ten upstream decisions that set your ceiling — the 10 Laws of Conversion is the scorecard for the other nine.
Running a trial you’re not happy with? Book a call and I’ll walk your funnel live.
