Almost every trial conversion conversation starts in the wrong place. A founder tells me the trial isn’t converting and asks me to look at the emails. Sometimes it really is the emails. Far more often, the number was set three decisions earlier, before a single user ever signed up.
That’s the thing most teams have backwards. Trial conversion gets treated as an optimization problem — better subject lines, a cleaner onboarding checklist, one more nudge before the trial ends. It’s an architecture problem. Ten upstream decisions determine whether a product converts at 3% or 15%, and every tactic you ship afterward moves you around inside the ceiling those decisions already set. You can run a flawless email sequence against a broken architecture and gain almost nothing.
The 10 Laws of Conversion is the scorecard I use to find which of those decisions is holding the ceiling down. It’s the diagnostic backbone of every Trial Conversion Engine engagement, and it’s the first thing I run before anyone writes a word of copy. Here it is, grouped the way it actually gets used.
Layer one: who gets in, and how long they have
Law 1 — Trial Model. Red ocean, freemium. Blue ocean, free trial. If the market has already trained users to expect free access, freemium lowers the barrier to habit. If the category is new and value has to be proven, a time-limited trial creates the decision pressure users need. The spread here is enormous: median free-to-paid runs around 2.6% for freemium against 15–25% for an opt-in trial. Those aren’t competing benchmarks, they’re different machines. Four questions decide which one fits your product, and most teams have never answered them deliberately — they copied a competitor’s pricing page.
Law 2 — Credit Card Gate. Requiring a card roughly triples conversion and cuts signups in half. Across roughly 200 B2B products, card-required trials convert near 30%, with an estimated 40–60% reduction in top-of-funnel volume. So the gate is not a conversion decision. It’s a filter for intent. Only require a card when the product can demonstrate real value in the first session — otherwise you delete the volume that would have converted through habit.
Law 3 — Trial Length. Your trial length is wrong if you didn’t set it with data. 62% of B2B products run 14-day trials, and most picked 14 because everyone else picked 14. The correct length is two to three days past the point where 70–80% of your eventually-activated users have activated. That’s a number sitting in your event data right now. Fewer than 3% of companies have ever looked at it.
Like this? I send one idea like it most weeks. One PLG idea a week → subscribe
Layer two: whether they ever reach value
Law 4 — Time to Value. You have about four hours to prove the product works. Category-leading PLG companies get users to the First Value Moment inside four hours of signup. Median is two to three days. Users who hit first value in session one retain at roughly three times the rate of everyone else. One B2B SaaS client I worked with had a first value moment that took most users close to a month to reach; restructuring the first session around a single forced first step pulled it under a minute. Nothing about the product changed. The first sixty seconds did.
Law 5 — The Blank Canvas. Never show a new user an empty product. The blank canvas is the single most common cause of trial abandonment and it is nearly universal in SaaS. Pre-built templates, sample data, and routing based on the job the user came to do produce the largest activation lifts on record — templates alone cut time to first value by 40–60%. This one is cheap to fix and almost always underbuilt.
Law 6 — Activation Metric. If you don’t have a defined activation metric, you’re guessing. Every product has exactly one event that predicts long-term retention. Most companies either haven’t defined theirs or have mistaken a correlative proxy for a causal one — “logged in twice” is not an activation metric. Average B2B SaaS activation sits near 36% against a top quartile above 60%, and a ten-point gain in activation drives a 20–30% gain in 30-day conversion. This law is the load-bearing one. Laws 7 through 10 all depend on knowing who activated and who didn’t.
Layer three: what you do with what you learn
Law 7 — Behavioral Segmentation. The same email to every trial user is an admission of defeat. A user who never logged in needs a re-engagement hook. A user stalled mid-activation needs help at the exact stall point. A user who activated but didn’t buy needs proof or urgency. Three completely different emails. Behavioral segmentation is worth a 20–40% conversion lift, and the cost of skipping it is worse than it looks: one client’s trial base of 443 users turned out to be roughly 80% people who never reached value at all. Nurturing that list as one audience means re-emailing hundreds of users who never got in the door. The 4-Profile System exists to keep that from happening.
Law 8 — In-App Prompts. Email does not drive most of your conversions. Your product does. In best-in-class PLG companies, 60–70% of self-serve conversions happen inside the product — limit-hit prompts, intent-based prompts, value-moment prompts. One-click upgrade beats a redirect to a pricing page by around 40%. If your conversion strategy is email-first, you are optimizing the minority lever.
Law 9 — Value Fence. If fewer than 10% of your free users hit your limits, your fence is in the wrong place. The Goldilocks zone is 20–30% of free users hitting the wall. Under 10% is revenue left on the table; over 50% is frustration before value is felt. Most companies set the fence once and never revisit it, which is why the math behind the value gate is usually the highest-yield hour in a diagnostic.
Law 10 — Post-Expiry. The email you’re not sending is the one that tells you why. A “what happened?” email 48 hours after the trial ends is the most under-used lever in the whole sequence. Reply rates run 5–15%, recovery on expired trials 2–5%, and the qualitative answers are worth more than either number. Nobody sends it because it feels like admitting the trial failed. It already failed. The email is how you find out what to fix.
How to actually run the scorecard
Score each law as one of four things: in place, partial, gap, or can’t tell without data. (There is a one-page reference card with the sourcing for all ten here.) Be honest about the fourth — half of what teams call “in place” turns out to be an assumption nobody has measured. Then rank by expected lift, not by how bad the grade looks. A gap on Law 3 in a product where users activate on day one is worth almost nothing. A partial on Law 6 poisons four other laws downstream.
You are not going to fix ten things. You shouldn’t try. The anonymized B2B SaaS client behind the result I lead with — 0.03% to a sustained 5% trial-to-paid — did not fix ten laws. The scoring exists to find the two or three that are actually capping the number, so everything else can be left alone on purpose.
Run it in order, too. The first three laws define the machine you’re operating. The middle three decide whether users reach value inside it — that’s your Golden Path — the one action your paying users all took. The last four are how you segment, prompt, gate, and learn. Working backwards, which is what optimizing emails first amounts to, is how teams spend two quarters improving the least important layer. And once you know who activated and who didn’t, you have the raw material for product-qualified leads, which is where the same data starts paying a second time.
If you’ve never scored your trial against all ten, the exercise is worth an afternoon on its own. Most teams find their real constraint is two layers upstream of where they’ve been spending.
Running a trial you’re not happy with? Book a call and I’ll walk your funnel live.
