Product analytics for SaaS
Datalyze is a product analytics consultancy for B2B and B2C SaaS companies. We find out why trials don't activate, why paying customers churn and which features keep users coming back, using Mixpanel, Amplitude or PostHog and your warehouse.
What SaaS teams can't answer
Why is our activation rate low?
What's usually going onActivation is often defined as "signed up" or "logged in twice", so the number looks fine while users leave before they reach the feature worth paying for. Or there's no agreed activation event, so nobody can say which step loses people.
Why don't trials convert to paid?
What's usually going onThe paywall shows up before users have seen value, or after they've already got what they needed for free. When plan limits and upgrade prompts aren't tracked, pricing gets decided by opinion.
Why are paying customers churning?
What's usually going onChurn shows up in billing a month after the user stopped showing up in the product. Billing sits in Stripe, usage sits in the analytics tool, and nobody joins them, so the early warning is missed.
Which features actually keep users coming back?
What's usually going onFeatures are tracked as page views, or not at all, so you can't compare users who adopted a feature with those who didn't. The roadmap ends up following the loudest customer.
Why can't we see usage by account?
What's usually going onAnalytics tools count users, but B2B SaaS sells to companies. One power user can hide nine seats that never logged in, so expansion and renewal risk stay invisible.
Which channels bring users who pay, not just users who sign up?
What's usually going onMarketing reports sign-ups, product reports usage and revenue lives in billing. Cheap channels often bring users who never convert, especially in B2C, and budget keeps flowing to them.
What we do about each one
- 01Activation
Define activation from your data, then track every step to it
We compare users who stayed with users who left, find the early actions that predict staying, and agree one activation event with your team. Then we track every step to it, find the step that loses the most people.
- 02Conversion
Track the paywall, not just the payment
We track plan limits, upgrade prompts and checkout, so you can see which trial behaviours lead to paying and which prompts appear too early. Then we run pricing and paywall experiments on the segments that matter.
- 03Churn
Join billing to product usage and flag risk early
We bring Stripe or Chargebee data in under the same user and account ID as your product events. Then we build a health score from the usage signals that drop before cancellation, so customer success hears about risk weeks earlier.
- 04Retention
Measure retention by feature, not by page view
We track core features as named events with properties, then build retention cohorts split by which features each user adopted. You get a ranked list of the features that keep users, and the ones nobody comes back for.
- 05B2B accounts
Report on accounts, not only users
We attach every event to both the user and their company, sync plan and MRR from billing and CRM onto the account, and model account-level tables for seats used, features adopted and the usage trend before renewal.
- 06Acquisition
Tie the first touch to revenue
We carry UTM and referrer data from the first visit through sign-up, then join it to billing, so each channel is judged on paid conversion and revenue instead of just sign-ups.
SaaS teams we've worked with
Buildern, a B2B construction SaaS, couldn't trust its attribution or see which accounts were at risk. We modelled the warehouse in dbt, fixed attribution on top of it, and built an account health score the team now uses to spot churn early.
Read the case studyFRAI had no clarity on which user segments converted or why others dropped off. We ran deep behavioural analysis and designed the A/B tests, and paid conversion doubled, with a repeatable experimentation process left behind.
Read the case studyMost VideoTap signups never reached the dashboard. We rebuilt the upload UX from 4 steps to 2, taking onboarding completion from 28% to 80% and activation from 7% to 19%.
Read the case studyThese are real results from B2B and B2C SaaS teams.
Book a callWhat SaaS teams say after working with us
SaaS analytics questions
Mixpanel, Amplitude or PostHog for a SaaS product?
All three handle SaaS funnels and retention well. Mixpanel and Amplitude are easier for product and growth teams to explore on their own. PostHog bundles session replay, feature flags and experiments, and suits engineering-led teams. For B2B, check that account-level (group) analytics is included in your plan. We implement all three, and usually recommend keeping the tool you already have.
How do you define activation for our product?
From your data, not a template. We compare users who were still active after 30 days with those who left, find the actions the retained group took in their first days, and pick the one that separates them best while still being something a new user can reach quickly.
Do you work with both B2B and B2C SaaS?
Yes. The funnel is similar, but the unit is different. B2C SaaS is analysed per user and depends on self-serve conversion. B2B needs account-level reporting, CRM and billing joined in, and usage tied to renewal. Buildern is B2B; FRAI and VideoTap are B2C.
Can you predict churn before customers cancel?
Yes, once product usage and billing are joined. We build a user or account health score from the signals that drop before cancellation, such as fewer active seats, a core feature no longer used or a failed payment, and put it where your customer success team already works.
How long until we see results?
Most teams get their first useful finding within two to three weeks. An audit of your current tracking takes one to two weeks, and a full setup takes one to two months.