The commonly quoted 8% median is a number almost no product actually occupies. Here are the bands by trial model, and what to do at each level.
The trial-to-paid conversion rate benchmark most often quoted is a median of 8%. Almost no product actually sits there, which makes it close to useless as a target.
Find your model's row first. Comparing across models is the most common misreading of this metric.
This page is the numbers. For the method of improving them, see the trial-to-paid conversion guide.
The most credible current dataset comes from ChartMogul, ProductLed and Kyle Poyar, who surveyed 200 B2B software products in January 2026 and defined conversion as a free signup becoming a paying customer within six months.
Their headline findings: a median free-to-paid rate of 8%, trials requiring a card converting around 30% — several times the rate of card-free trials — and a roughly tenfold gap between the top and bottom fifth of self-serve products.
| Model | Typical range | Strong |
|---|---|---|
| Free trial, no card | 8–15% | 15–25% |
| Free trial, card required | 25–35% | 50%+ |
| Freemium | 2–5% | 8–12% |
| Reverse trial | 18–32% | 30%+ |
Ranges rather than single figures, because published studies disagree. A widely cited 2025 dataset put card-free trials near 18%; the larger 2026 study puts them closer to 9%. Both are defensible and they measured different samples.
Conversion rate in isolation is a misleading metric, and the ChartMogul data shows why cleanly.
Modelled per 1,000 visitors, a standard free trial produced around 45 signups and roughly 3.6 paying customers. A card-required trial produced fewer signups — about 35 — but roughly 10.5 paying customers.
The card-required trial converts far better and captures fewer people. In that model it also produced more customers overall, but that is not guaranteed for every product: requiring a card can cut top-of-funnel volume substantially, and for an unknown brand the drop is larger.
The study's typical respondent sits between $1M and $10M ARR. If you have 30 trials a month, three caveats apply before you compare yourself to anything above.
Your sample is too small. At 30 trials, one extra conversion moves your rate by more than three points. Read a rolling three-month figure rather than a monthly one.
Traffic source dominates. Trials from organic search convert far better than trials from a launch spike or paid social. A rate that halves after a Product Hunt launch usually reflects traffic quality, not a product regression.
Activation predicts conversion better than anything else. If under a fifth of trials reach first value, your conversion rate is an activation problem wearing a pricing costume — see the onboarding framework.
In the ChartMogul sample, 14 days was the most common trial length, used by 62% of products. Free trials were the primary entry point for 57% of products, against 26% for freemium and 7% for reverse trials.
That does not make 14 days correct for you — it should follow from your time-to-value rather than from what is common. The trial length playbook covers choosing it, and freemium versus free trial covers the model decision itself.
Before changing anything, check where trials actually stop. Conversion decisions cluster near the end of a trial, so a rate that looks poor at day 5 may be normal. Judge the cohort after the trial has fully expired, not during it.
| Your rate (no card) | Reading | Work on |
|---|---|---|
| Under 5% | Something structural is broken | Activation, then trial length |
| 5–10% | Normal but improvable | The trial email sequence |
| 10–20% | Working | Traffic volume, not conversion |
| Over 20% | Strong | Consider raising the price |
That last row is the one founders skip. A conversion rate well above the band often means the price is below what the market will bear — the price increase playbook covers the sequence.
Conversion benchmarks move quickly and definitions vary — some studies count six-month windows, others count trial expiry. Check the linked source for current figures and methodology before treating any number here as a target.
It depends on your model. For trials without a credit card, roughly 8–15% is solid and 15–25% is strong. Card-required trials sit around 30% at the median with 50%+ at the top end. Freemium-to-paid typically runs 2–5%.
ChartMogul, ProductLed and Kyle Poyar found a median free-to-paid rate of 8% across 200 B2B software products surveyed in January 2026. Very few products actually sit at that median — the spread between the top and bottom fifth is roughly tenfold.
Substantially, at roughly 30% versus several times less for card-free trials in the 2026 data. They also capture fewer signups. Measure paying customers per 1,000 visitors rather than conversion rate alone, because conversion can be improved simply by reducing signups.
Use a rolling three-month figure rather than a monthly one. At 30 trials, a single extra conversion moves your rate by more than three points, so month-to-month movement is mostly noise rather than signal.
That is strong, and it often indicates your price sits below what the market will bear. At that level the useful next move is usually testing a higher price for new customers rather than optimising conversion further.
14 days was the most common length in the 2026 ChartMogul sample, used by 62% of products. What is common is not necessarily right for you — trial length should follow from how long your product takes to demonstrate value.
Bring your trial numbers and activation rate. Marcus tells you whether to fix conversion, activation or traffic — and which one first.
Try GhostCoach free →14-day free trial · cancel anytime · 30-day money-back on Lifetime