A free trial is not a marketing tactic, it is a distribution model. Done well it compresses the sales cycle, lowers acquisition cost, and lets the product carry a share of the pipeline that would otherwise sit on a sales rep. Done poorly it floods the funnel with signups who never activate, trains the market to wait for a discount, and masks a product that is not yet ready to sell itself. This guide walks the full launch arc for a B2B SaaS free trial or freemium program: picking the trial model, defining activation events, designing the signup flow, building the onboarding automation, standing up PQL scoring, launching, measuring, and iterating on a monthly cadence.
Time: 6-10 weeks
There are three real trial models in B2B SaaS, and the first mistake most teams make is choosing on vibes rather than on product shape. A time-boxed trial gives the user full access for a fixed window, usually seven to thirty days, and works best when the value is obvious inside that window and the buyer is already shopping. A usage-boxed trial gives the user full access until they hit a cap on seats, records, workflows, or API calls, and works best when value scales with volume and the buyer needs to feel the ceiling before they will pay. Freemium gives the user a permanent free tier with reduced functionality, and works only when you can afford the support cost of a large non-paying base and when the free tier produces a steady stream of qualified upgrades. OpenView PLG benchmarks and Reforge PLG research both land in the same place: the right choice is the one that matches how your product actually produces value, not the one your competitors picked.
Activation is the most abused word in SaaS. Treat it operationally. An activation event is a specific, observable, in-product action that reliably predicts a user will convert to paid. Pick two or three, not twelve. The first is almost always account setup completed: workspace named, a teammate invited, and the first piece of real data imported or connected. The second is the first use of the core workflow with the user's own data, not with sample data. The third, if your product supports it, is a collaboration event: a share, a comment, an export, a workflow run that touches a second human. Appcues trial research is consistent on this point: products that explicitly define activation and instrument it convert trials to paid at multiples of products that leave activation implicit. Write the events down, add them to your analytics tool, and tie every onboarding email, in-app nudge, and PQL rule to them.
Every field in the signup flow is a tax on top-of-funnel volume, and every field you remove is a bet that the sales motion downstream can recover the context without it. OpenView PLG index data is unambiguous: fewer fields produce higher signup rates, and in self-serve categories the lift is large enough to dominate the first-year revenue curve. Start with email and password, or an SSO option, and defer everything else to the first in-product session. If you are sales-assisted, you can ask for company name and company size on step two, but keep step one to the minimum. Verify the email asynchronously rather than blocking the user behind a magic link, and route the user straight into the product the moment the password is set. The one place to add friction is bot defense: a simple captcha or a passive behavioral check costs you almost nothing in human conversion and keeps the trial free of fake signups that will pollute your PQL scoring.
Onboarding is not a welcome email. It is a coordinated program that runs across email, in-app messaging, and sometimes SMS or a scheduled call, and that is scored against the activation events you defined. For a time-boxed trial, build a sequence that lands a message on day zero, day one, day three, day seven, and three days before expiration, with branching logic that skips messages for users who have already completed the related activation event. For freemium, run a continuous drip that nudges users toward the next activation event they have not completed and surfaces upgrade triggers at the moments users hit the free-tier ceiling. In-app messaging is the higher-leverage half of this program because it reaches the user inside the product at the moment of intent, while email is the lower-leverage half that carries the context the user needs when they are away from the app. ProductLed research is clear that teams that run both channels against the same activation goal convert at meaningful multiples of teams that run email alone.
A product qualified lead is a trial or freemium user whose in-product behavior predicts a purchase conversation would land. The point of PQL scoring is to let sales reach the users who are ready and leave the rest to the automation, which is how a trial program scales without burying the sales team in low-fit signups. Build the scoring model on the activation events you defined, plus a thin layer of firmographic context: company size from an enrichment provider, domain match against your ICP list, and seat count or usage inside the trial account. Keep the model simple at launch. A clean rule such as reached two activation events plus company size over fifty employees plus three or more seats invited is almost always a better starting point than a weighted model with ten inputs, because the simple rule is explainable to the sales team and tunable from the data as the program matures. Reforge PLG research is consistent that the first version of a PQL model should fit on an index card, and that teams that overbuild it in the first ninety days lose the plot.
Do not open the trial to the whole top of funnel on day one. Launch to a bounded cohort first, usually by holding a flag on paid acquisition for two to four weeks and routing only organic and existing-list traffic through the trial, or by capping daily signups via the signup page. The point of the bounded launch is to stress test the mechanics with a group small enough to debug and large enough to produce signal. Watch five numbers in near real time: signups, email verified, activation rate at day three, PQL volume, and trial-to-paid conversion at day thirty. Pair the quantitative view with a qualitative one: have a human read every churned trial's product telemetry for the first two weeks and talk to five churned users on video. The bugs and friction points that show up in the first thirty days are almost always the ones that would have been masked by a broad launch, and fixing them before you scale paid acquisition is how the unit economics stay honest.
Once the program has thirty days of clean data, run a formal review every month for the first two quarters and quarterly thereafter. Lead with two headline numbers: trial-to-paid conversion over a rolling thirty-day cohort, and days-to-activation for the median activated user. The ratios around them are the diagnostic. If signups look healthy but verified rate is weak, the email verification step is leaking. If verified is healthy but activation at day three is weak, onboarding is not landing the user on the activation event fast enough. If activation is healthy but trial-to-paid is weak, the upgrade experience, the pricing surface, or the sales SLA on PQLs is the problem. Fix one layer per review, run one real experiment against it, and resist the temptation to change three things at once. Appcues trial research and ProductLed benchmarks both land on the same discipline here: durable trial programs are tuned one lever at a time, and the compounding curve shows up only when the mechanics are stable long enough for the market to form a habit around them.
Plan for six to ten weeks end to end if the product baseline, time-to-value, and billing prerequisites are already in place. The arc runs trial-model choice and activation definition in weeks one and two, signup and onboarding build in weeks three through six, PQL scoring and sales handoff in weeks seven and eight, and a bounded launch with monitoring in weeks nine and ten.
Pick the model that matches how your product actually produces value. Time-boxed works when the value is obvious inside a two-week window, usage-boxed works when value scales with volume and the buyer needs to feel the ceiling, and freemium works only when the free tier has standalone value and you can afford the support cost of a large non-paying base.
An opt-in card raises conversion from signups to paid but suppresses top-of-funnel volume. If the sales motion can defend the volume hit and billing is wired for a clean trial-end charge, cards up front make sense. If paid acquisition economics depend on volume at the top of the funnel, defer the card to the upgrade moment.
A product qualified lead is a trial or freemium user whose in-product behavior predicts a purchase conversation would land. Build the first version of the model on your activation events plus a thin layer of firmographic context, keep it as a boolean rule that fits on an index card, and route PQLs to named reps with a short SLA rather than a round-robin queue.
Most self-serve B2B SaaS trials convert in the three to ten percent range on cold traffic, with sales-assisted trials and warm referral traffic converting materially higher. Write your target down before launch, segment the number by signup source and company size, and tune against the diagnostic ratios in the funnel rather than against the headline conversion number alone.
Run a monthly review for the first two quarters and quarterly after that. Pick one layer of the funnel to improve each cycle, run one real experiment against it, and resist changing three things at once. Durable trial programs are tuned one lever at a time, and the compounding curve shows up only when the mechanics are stable long enough for the market to form a habit.
Strkr gives product, marketing, and sales a shared view of signups, activation events, PQL triggers, and trial-to-paid conversion, so your free trial program stops living in a dashboard spreadsheet and starts compounding inside the system of record.
No sales call, no migration consultant, no four-month implementation. Enter your card, get 14 days of the full Pro tier, cancel any time before day 14 with zero charge. Spin up a workspace, import your CSV, and have something useful before lunch.
