Product-led growth (PLG) is a go-to-market strategy where the product itself does the work of acquiring, converting and expanding customers, usually through a free trial or free plan instead of a sales call.
On the conversion side, PLG comes down to three numbers: how many visitors sign up, how many sign-ups reach value, and how many of those users pay.
Real A/B tests show the biggest gains usually come from the middle step, activation, not from the pricing page.
This page looks at PLG through the tests in our database: what changed, what it moved, and who ran it.
For the wider picture of SaaS testing, start at the SaaS CRO hub.
What PLG means, and where the term came from
The term was coined at the venture firm OpenView by partner Blake Bartlett, who settled on the name in 2015-2016, according to OpenView's own history of the term (archived copy, as OpenView's site now redirects).
The early definition was "a go-to-market strategy that relies on product features and usage as the primary drivers of customer acquisition, retention and expansion."
Before the name existed, people called these businesses "freemium" or "bottom-up."
OpenView's PLG guide stressed that PLG is broader than freemium: a sales-led company can still adopt product-led principles, such as removing barriers between a user and their first solved problem.
For a CRO team, that changes what a "conversion" is.
In a sales-led funnel the website's job ends at a demo request.
In a PLG funnel the website hands off to the product, and the product has to finish the sale.
The PLG funnel as a set of conversion points
Every PLG motion has the same skeleton, and each step can be tested on its own.
| Step | What converts | Typical metric | Where it lives |
|---|---|---|---|
| Sign-up | Visitor to account | Sign-up or trial-start rate | Homepage, pricing page, sign-up flow |
| Activation | Account to engaged user | Share of users who complete a key action | First-run experience, onboarding |
| Monetization | Free user to payer | Trial-to-paid or free-to-paid rate | Trial limits, paywalls, upgrade prompts |
| Expansion | Payer to bigger payer | Net revenue retention, seat growth | In-product upsells, team invites |
The trap is optimizing one step at the cost of the next.
A test that doubles sign-ups by attracting people who never activate can lower revenue.
So when you read the results below, check which metric each test reported.
Product-led growth strategy: trial, freemium or reverse trial
The first strategic call in PLG is the free offer itself.
Each model trades volume at the top of the funnel for conversion further down.
A benchmark from First Page Sage (86 SaaS companies, Q1 2022 to Q3 2025, updated September 5, 2025) puts numbers on that trade-off:
| Model | Visitor to sign-up (organic) | Sign-up to paid (organic) |
|---|---|---|
| Opt-in free trial (no card) | 8.5% | 18.2% |
| Opt-out free trial (card required) | 2.5% | 48.8% |
| Freemium | 13.3% | 2.6% |
The report counts a single paid month as a conversion, so it doesn't show retention.
Read it as a picture of where each model loses people, not as proof that one model wins.
A third option is the reverse trial. Elena Verna, writing for Amplitude, defines it as putting users "in a freemium experience with a timed free trial with access to paid capabilities which reverts to a traditional freemium product after the free trial ends."
We don't yet have a controlled reverse-trial test in the database, so treat it as a promising pattern, not a proven one.
What our database does have is evidence that switching models is high-stakes and context-dependent:
- Freemium hurt one app. A subscription app advised by Phil Carter moved from a hard paywall to freemium and saw more than 50% lower subscriber conversion. It reverted after a couple of weeks. This was a before-after change, not a split test.
- Freemium "lost" and still won for another. Opal replaced its hard paywall with a free tier, and pay penetration fell from 20% to 9% while daily active users passed one million. Opal judged it a long-term win. Whether that's right depends on what a free user is worth to you.
- Adding a trial next to a free plan. emaze added a 30-day premium trial alongside its free plan and reported more paid upgrades, though it published no figures.
For a deeper comparison of the two main models, see freemium vs free trial.
Sign-up: cut the steps between interest and the product
PLG only works if people get into the product, so the sign-up is where most early tests start.
The pattern in our data is consistent: fewer steps and fewer doubts.
- Encyclopaedia Britannica cut one whole page from its registration path. In a test documented by MarketingExperiments, that produced +117% free trial conversion, +61% paid subscriptions and +130% revenue. The gain carried through to money, which is the part that matters.
- Teamleader added "No credit card required" to its trial sign-up page. Conversion rose from 48.7% to 54.8% on the Netherlands site and from 33% to 38.6% in Belgium, a +12.5% headline lift in sign-ups.
- Kareo Marketing removed fields from a doctor sign-up form and got +30% new sign-ups.
Not every friction cut works. Patrick McKenzie removed a "log in as guest" link from Bingo Card Creator's sign-up and saw no change in trial sign-ups.
Small elements that nobody uses won't move the number.
Activation: the step PLG teams underrate
In a sales-led company, a salesperson walks the buyer to value.
In PLG, the product has to do it alone.
The largest lifts in our SaaS data come from this step.
- AURUM (Astrea) added a floating checklist that guided trial users through their first tasks. Activation rose 187%. AURUM ran it as a controlled rollout rather than a strict A/B split.
- AURUM also made legal clippings available on first login instead of up to 48 hours later. The share of trial users engaging with a clipping rose 16 percentage points. That's a product change, not a UI tweak, and it's the purest time-to-value test we have.
- Citrix added an in-product guidance layer to the GoToWebinar 30-day trial. Trial-to-paid conversion doubled and product usage rose 40%.
- VWO built a central "Get Started" checklist for its own free trial and reported +127% trial-to-paid conversions. The new flow bundled other UX changes, and VWO was testing its own product, so weigh it with that in mind.
MarketingExperimentsThe common thread: these tests changed what users did inside the product, and the paid conversion followed.
If your trial-to-paid rate is low, look at what share of trials reach the first meaningful action before you redesign the pricing page.
Monetization: where to put the limit
A free plan or trial has to leave people wanting more.
Set the limit too generously and nobody needs to pay.
Set it too tight and nobody sees the value.
Patrick McKenzie's test on Bingo Card Creator is the cleanest example.
Cutting the free trial from 15 cards to 8 raised trial-to-paid conversion from 1.04% to 1.89%, with a sample in the 5,000s at 99% confidence.
A tighter limit worked here because 8 cards were still enough to see the product's value.
The opposite can also be true. Headspace locked all content behind the subscription and reported a significant rise in paid subscribers, while OnSwitch locked free content for three days and found users came back slightly more but bought less.
The right limit depends on how quickly a user can feel the value before hitting it.
For limit patterns tested in apps, see freemium limits. For trial-specific tactics, see free trial conversion.
How to test a PLG funnel without fooling yourself
PLG tests are harder to read than landing page tests, for three reasons.
- The real outcome is late. A trial might convert 14 or 30 days after sign-up. Decide the primary metric before launch, and wait for the full trial cohort to finish before calling it.
- Model changes are rarely split tests. Several of the freemium results above are before-after comparisons, which mix in seasonality and other changes. If you can, split new sign-ups by cohort rather than switching everyone at once.
- Bundled changes hide the cause. The VWO and AURUM onboarding results shipped several changes together. That's fine for a business decision, but don't assume each piece would lift on its own.
Use a guardrail metric one step downstream of your primary metric.
If you test sign-up friction, watch activation. If you test activation, watch paid conversion.
To compare your numbers with other companies, see the SaaS conversion rate benchmarks, and browse all tests for more examples by page type.