Built for SaaS teams scoring free-trial users

The lead-scoring attributes that predict free-trial conversion are not the ones most SaaS teams score on.

Most SaaS free-trial scoring models still weight firmographic fit (company size, industry, title) over product behavior. The 2026 data shows product-usage attributes predict paid conversion roughly 4x better than firmographic signals alone. Strkr's PLG scoring model is designed around that reality.

Start free trial See pricing
What this audience is actually dealing with

The pains that bring buyers here.

Scoring free-trial users is structurally different from scoring marketing-sourced inbound leads. The trial user has already self-selected into the product. The relevant question is no longer "are they a fit" but "are they activating", and most SaaS teams are still running MQL-era scoring models against trial pipelines. The symptoms show up as reps calling trial users who never logged in a second time and ignoring the signed-up account that invited 8 teammates in the first 48 hours.

Firmographic-first scoring misses the signal

Company size does not predict trial conversion.

Scoring models carried over from MQL marketing programs weight firmographic attributes (headcount, industry, title). Published PLG data from Amplitude, Mixpanel, and ProductLed research shows firmographic signals explain roughly 15 to 25 percent of paid-conversion variance for free-trial users. Product usage explains the rest. Teams that score primarily on firmographics prioritize the wrong accounts.

Activation events get ignored

The moment that predicts paid is not in the model.

Every SaaS product has an activation event that strongly predicts paid conversion: integrations connected, teammates invited, first project created, first report generated. Trial users who hit the activation event convert 3 to 8x higher than trial users who do not. Most lead-scoring models do not include activation events as a scoring attribute because the product-analytics data lives in a separate system.

Depth-of-use signals are missing

Session count and feature adoption predict paid.

Trial users who log in on 3 or more distinct days during the trial window convert at roughly 4x the single-session rate. Users who adopt 2 or more features convert at roughly 5x the single-feature rate. These depth-of-use signals are the strongest predictors of paid conversion and the easiest to miss when the scoring model does not pull from product analytics.

Team-expansion signals get buried

A trial user inviting teammates is a buy signal.

When a trial user invites teammates, the account is signaling organizational adoption intent. Teams-invited count is one of the single strongest predictors of paid conversion for team-based SaaS products. The signal often does not make it into the lead-scoring model because invitations are a product-side event, not a CRM field.

Negative signals get weighted wrong

A trial user who logs in once and never returns is cold.

A trial user who signs up, logs in once, and never returns is almost certainly not converting. Many scoring models still show these accounts as 50 to 70 points because the firmographic fit scored high. Reps spend time on no-chance accounts because the model does not decay the score when activation does not happen within the first 72 hours.

Scoring windows are wrong for trials

Trial scoring needs 7-day decay.

Free-trial scoring needs tighter decay windows than MQL scoring. A trial user who was hot 10 days into a 14-day trial but has not logged in since is cooling fast. A rep calling that account on day 13 is too late. Lead scoring models designed for 30 to 90-day MQL cycles do not decay fast enough for trial pipelines.

The attributes that actually predict free-trial paid conversion

The 2026 PLG scoring model.

The attributes below are weighted based on published PLG benchmarks and the baseline conversion lift each signal produces. The cards walk the top predictors in rough weight order.

Activation event completion

Primary predictor, 30 to 40 percent weight.

The product-specific activation event (first project created, first integration connected, first report run, first teammate invited, first data uploaded) completed within the first 72 hours of signup. Users who hit activation convert at 3 to 8x the non-activated rate across published benchmarks. The event is product-specific but every product has one; Strkr pulls it from the product-analytics feed.

Session count during trial

Depth-of-use signal, 15 to 20 percent weight.

Distinct active days during the trial window. Users with 3 or more active days convert at roughly 4x the 1-day rate. The attribute captures stickiness beyond the signup moment. Scoring builder lets the admin set a threshold or run a linear weight per active day. The Strkr model recommends 1 point per active day up to 10.

Feature adoption count

Depth-of-use signal, 10 to 15 percent weight.

Distinct product features used during the trial. Users adopting 2 or more features convert at 5x the single-feature rate. Each feature can be configured as a separate scoring event with its own weight (power features that correlate with paid plans get higher weight than peripheral features).

Teammate invitations

Organizational-intent signal, 10 to 15 percent weight.

Number of teammates invited during the trial. For team-based SaaS, inviting teammates is one of the single strongest paid-conversion predictors. Users inviting 1 teammate convert at 2x the solo rate; users inviting 3 or more convert at 5 to 7x. The attribute flows from the product's invitation event.

Integration connections

Workflow-depth signal, 5 to 10 percent weight.

Number of integrations the trial user connects. Integrations are a strong intent signal because the user is investing time to wire the product into their workflow. One integration lifts conversion roughly 2x; 2 or more integrations lifts 4x. The attribute pulls from the integration-connection event on the product side.

Firmographic fit

Secondary signal, 10 to 15 percent weight.

Company size, industry, and title aligned with the ICP. Firmographic fit matters as a tiebreaker and for prioritizing which hot trial accounts get the deepest rep attention. The weight is lower than product usage because firmographics do not predict trial conversion on their own; they predict ACV once conversion happens.

Content and support interactions

Engagement signal, 5 percent weight.

Help-center article views, documentation visits, support-chat conversations during the trial. The signal captures users who are investing time to learn the product. Weight is modest but the signal is a useful tiebreaker when two accounts have similar product usage.

Billing-page visits

Buying-intent signal, 5 percent weight.

Visits to the pricing or billing page during the trial. The signal is a direct buying-intent attribute: the user is comparing plans or evaluating budget. A rep follow-up triggered on billing-page visits tends to land well because the user is already in the decision moment.

Negative signals (score decay)

Inactivity decays the score.

Scores decay when the trial user goes inactive. Default decay rule: minus 10 points per day of no login after day 3 of the trial. The decay stops reps from chasing accounts that have gone cold. The decay rate is configurable; the default is tuned for 14-day trials and should be tightened for shorter trials.

How to build the scoring model in Strkr

The implementation path.

Strkr's scoring builder reads product-analytics events alongside CRM fields. The implementation flow below is a 4-week path from zero to a working PLG scoring model.

Week 1: data wiring

Connect product analytics to Strkr.

Native integrations with Amplitude, Mixpanel, Segment, Heap, and PostHog. The integration pulls the event stream for trial users and maps events to Strkr's scoring builder. Alternative: direct product-event ingestion via Strkr's event API if the team is instrumented on a custom stack. Setup is roughly 2 to 4 hours for a mid-size SaaS product.

Week 2: baseline analysis

Identify the activation event.

Strkr's AI runs a retrospective analysis on the past 90 days of trial-to-paid conversions. The analysis identifies the single event that most strongly predicts conversion and the depth-of-use thresholds that correlate with paid. The output is a draft activation-event definition and recommended scoring thresholds.

Week 3: model configuration

Build the attribute weights.

Configure the scoring attributes using the recommended weights from the baseline analysis or set them manually. The scoring builder is a drag-and-drop interface with weight sliders, threshold rules, and decay settings. The admin can run the model against historical data to see which accounts would have scored hot and compare to actual conversions.

Week 4: rep routing

Wire score tiers to rep queues.

Hot trial accounts (score 70+) route to assigned rep queues with task auto-creation. Warm accounts (40-70) land in a lower-priority queue. Cold accounts (under 40) stay in email nurture. The routing is configurable per segment: enterprise trials route to AEs, mid-market trials to SDRs, SMB trials to self-serve nurture.

Ongoing: model refinement

Weekly review of score accuracy.

Strkr's model-accuracy dashboard shows which score tiers converted and which did not. The team adjusts weights monthly based on observed accuracy. Over 6 months the model typically tightens by 20 to 40 percent on precision at the hot tier.

What free-trial scoring looks like in practice

The trial-rep workday.

Scoring is only useful if it changes the rep's workday. The cards below describe how the model surfaces in day-to-day sales motion for a PLG-focused rep.

Morning hot-trial queue

Overnight activations surface at 8am.

Each morning the rep opens the hot-trial queue to find accounts that activated overnight or crossed scoring thresholds during off-hours. The queue is ordered by score with tie-breaking on signup recency. The rep works the top of the queue first because fresh activation energy is the highest-value window to engage.

Context-rich account view

Every scoring signal visible on open.

When the rep opens a hot trial account the record shows the full scoring breakdown: which events fired, which depth-of-use thresholds the user crossed, which features they adopted, which integrations they connected. The rep opens the outreach with specific context ("I saw you connected Salesforce and invited 4 teammates yesterday") rather than a generic trial check-in.

Billing-page alerts

Real-time nudge for buying intent.

When a trial user hits the pricing or billing page the assigned rep gets a Slack alert within 60 seconds. The response window for billing-page intent is tight; teams responding within 15 minutes see 2 to 3x the connect rate versus teams responding within 4 hours. The alert routing is per-segment configurable.

Decay-triggered pause

Cold trials drop out of the queue.

When a trial score decays below the warm threshold the account drops out of the rep queue automatically and routes back to email nurture. The rep is not blocked from re-engaging but the default workflow stops burning rep time on cooling accounts. The decay rule prevents the "80-point zombie account" problem that plagues static scoring models.

Weekly trial retrospective

Rep sees which signals paid off.

Each Friday the rep reviews the week's converted trials and the hot-queue accounts that did not convert. The retrospective surfaces which scoring signals predicted correctly and which did not. The rep feeds the signal back to the admin and the model tightens over time.

Trial-expiration intervention

Last-chance outreach on hot accounts.

For hot trial accounts approaching trial expiration without converting, the system creates a last-chance task for the rep 48 hours before expiration. The outreach often lands a conversion because the deadline forces the decision. The task is only created for accounts above the warm scoring threshold; cold trials expire silently.

Head-to-head

Strkr PLG scoring vs legacy MQL-era models.

The comparison below is against the typical scoring model that marketing operations teams carry forward from MQL programs. The product-usage layer is the structural difference.

Feature Strkr Legacy MQL scoring model
Product-usage attributes First-class, 60 to 70 percent of model weight Rarely included
Activation-event tracking Native integration with product analytics Not scored
Depth-of-use signals Session count, feature adoption, integration count Not available
Score decay Fast decay on trial inactivity Slow decay (30 to 90 day window)
Teammate invitation signal Weighted scoring attribute Not tracked
Billing-page alerts Real-time rep notification Delayed or missing
Model retrospective Accuracy dashboard on converted vs not Manual export to BI
Weight calibration AI-recommended weights from retrospective Hand-set by marketing ops
Routing by segment Enterprise, mid-market, SMB separate thresholds Single model for all
Rep-visible scoring breakdown Every signal shown on account open Score only, no breakdown
How teams use Strkr

SaaS teams scoring free trials on Strkr.

The playbooks below are live free-trial scoring motions on Strkr customers in 2026. Each is a specific PLG scoring pattern that lifted trial-to-paid conversion.

Horizontal SaaS, 180-rep team

Rebuilding scoring on product usage lifted paid conversion 34 percent.

A horizontal SaaS company serving sales and marketing teams moved from a legacy firmographic MQL model to Strkr's PLG scoring in Q2 2026. The old model gave high scores to large-headcount accounts regardless of trial behavior. The new model weighted activation events, session count, and teammate invitations. Trial-to-paid conversion lifted from 11 to 15 percent over the following quarter. The team attributed the lift to reps spending time on accounts that were actually using the product.

Vertical SaaS, 40-rep team

Billing-page alerts drove a 2.1x lift in connect rate.

A vertical SaaS company serving a industry turned on real-time billing-page alerts routed to Slack. Reps responded within 10 minutes on average versus the prior 3-hour baseline. Connect rate on billing-page triggers doubled and conversion within 48 hours of the alert ran at roughly 2.1x the baseline. The team now treats the billing-page alert as the single highest-signal event in the model.

Product-led startup, 12-rep team

Decay rules cut rep wasted time 40 percent.

A 12-rep PLG startup had reps chasing "80-point" accounts that had gone cold weeks earlier because the scoring model never decayed. After implementing 10-point daily decay on trial inactivity, roughly 40 percent of previously-hot accounts dropped out of the queue. Reps redirected the saved time to the actually-hot trial pipeline and conversion on worked accounts lifted 22 percent.

Collaboration SaaS, 25-rep team

Teammate-invite scoring surfaced the whales.

A team collaboration SaaS discovered through retrospective analysis that teammate-invite count was the single best predictor of paid conversion. The weight on invitations jumped from 5 to 25 percent of model total. Hot trial accounts shifted toward team-expansion accounts and ACV on converted trials lifted 60 percent because the model was surfacing larger-team accounts.

Developer tooling, 18-rep team

Integration-connection events predict paid best.

For a developer tooling product, integration connections were the strongest conversion predictor (users who connected 2+ integrations converted at 7x the single-integration rate). The team weighted integrations at 30 percent of total model. Trial-to-paid lifted 28 percent and rep efficiency (dollars-per-SDR-hour) lifted 44 percent because the model was accurately identifying workflow-committed users.

Score free-trial users on signals that actually predict paid.

Start a 14-day trial with native Amplitude, Mixpanel, and Segment integration, AI-recommended scoring weights, real-time billing-page alerts, segment-specific models, and the retrospective accuracy dashboard.

Start free trial See pricing
Common questions

What buyers in this bucket ask most.

How is this different from the lead scoring my marketing-automation tool already does?

Marketing-automation lead scoring (HubSpot, Marketo, Pardot) is built around MQL marketing programs where the inputs are email opens, form fills, and web visits. The outputs serve a top-of-funnel qualification motion. Free-trial scoring is structurally different: the inputs are product-usage events (feature adoption, session count, activation events, teammate invitations, integration connections) and the output serves a bottom-of-funnel conversion motion. Most marketing-automation tools do not ingest product-analytics events natively, so teams end up with two scoring models in two systems that never reconcile. Strkr's PLG scoring reads from Amplitude, Mixpanel, Segment, Heap, or direct event ingestion and unifies the trial-user view in the CRM.

What if we do not have product analytics instrumented?

If the product does not emit events to Amplitude, Mixpanel, or Segment, the first move is to instrument the 5 to 10 events that matter most for trial scoring: signup, login, activation event, feature use, teammate invite, integration connect, billing-page view, upgrade click. Direct event ingestion to Strkr's event API works for custom stacks. The instrumentation effort is typically 2 to 4 developer days. Without product analytics, free-trial scoring falls back to firmographic signals which, per published PLG research, predict roughly 15 to 25 percent of conversion variance. Teams that invest the instrumentation step see the ROI within the first quarter.

How often should we recalibrate the scoring weights?

Monthly recalibration is the standard cadence for PLG scoring models. The signal landscape shifts as the product evolves (new features change which usage patterns predict paid) and the customer base shifts (new segments change which firmographic weights apply). Strkr's model-accuracy dashboard shows the precision and recall of each score tier against actual conversions. Teams review the dashboard monthly and adjust 2 to 4 weights per cycle. Over 6 to 12 months the model tightens significantly as the weights converge on the true predictors.

Should different segments have different scoring models?

Yes, for teams serving distinct segments with distinct trial behaviors. A single scoring model that applies to enterprise, mid-market, and SMB trials is a common mistake because the signals that predict paid conversion differ by segment. Enterprise trials often show lower session counts but higher activation depth; SMB trials show the opposite. Strkr supports segment-specific scoring models with different attribute weights per segment. Teams typically start with one model and split by segment at 1,000 to 2,000 monthly trial signups.

How does Strkr handle very short trials (7 days or less)?

Short trials require tighter decay windows and faster rep response times. The default decay rule is minus 10 points per day of inactivity after day 3; for 7-day trials this tightens to minus 10 points per day after day 1. Billing-page alerts and activation-event alerts route to reps within 60 seconds versus the 2 to 4 hour window that mid-length trials can tolerate. Short trials also benefit most from the trial-expiration last-chance task because the deadline pressure is immediate. The scoring attributes stay the same; the decay and routing speeds up.

Can we score free-trial users and freemium users in the same model?

The attributes are similar but the thresholds and decay rules differ. Freemium users are not time-boxed, so decay runs slower and the activation window extends. Freemium scoring also emphasizes expansion signals (adding users, hitting usage limits, approaching tier caps) that time-boxed trials do not have. Strkr supports separate scoring models for trial vs freemium with shared attribute definitions and separate weight configurations. Teams running both motions typically find the models share about 60 percent of attributes and differ on the remaining 40 percent.

Also built for

Related audiences.

Lead Scoring for SaaS Companies Lead Scoring Models for SaaS Self-Service Sales AI Lead Score Value Lead Status

Try it free. Bring your team next week.

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.

Start free trial See pricing