Home › SaaS Resources

B2B SaaS Analytics: Usage, Revenue and Honest Numbers

B2B SaaS analytics is the practice of measuring a subscription product from two sides. One side is behavior: what accounts do inside the software. The other is money: who joined, who stayed, who grew and who left. The skill lies less in collecting data than in agreeing what the numbers mean.

The two halves of the picture

Product measures watch behavior inside the software. Which screens people open, where they stop, and which features they never touch at all.

Revenue measures watch the money. New accounts, renewals, upgrades, and the accounts that quietly slip away without saying why.

You need both at once. Usage without money hides churn, and money without usage hides the reason behind it. Most arguments about analytics are really arguments about which half somebody happens to be looking at.

The measures worth your time

Most teams track far more than they use. This short list covers almost all of the work that matters:

• Activation. The share of new accounts that reach a first real result inside the product.
• Retention. Who is still with you months later, group by group.
• Expansion. The accounts that grow after they sign, and the reason they grew.
• Churn. The accounts that leave, and the reason each one gave.
• Time to value. How long the first real result takes to arrive for a new account.

Pick the two that matter most this quarter and keep them in plain sight. The rest can live in a report that nobody has to read weekly.

Account-level thinking

In B2B, the customer is a company with several users. A single login going quiet may mean little, while a whole account going quiet is a warning. Group users by account, so you see the company's behavior, and look at the number of active people against the number who were given access. Wide gaps often point to a rollout problem, not a product problem.

Cohorts: the fairest comparison

Comparing this month's customers with last year's customers mixes different people at different stages. A cohort view groups customers by when they joined and follows each group over time. It answers the question that matters: are customers who joined recently staying as long as those who joined earlier?

How to keep the data honest

Name every event once and write the name down where the whole team can see it. Two events called the same thing will ruin a chart six months from now.

A report is only useful when everyone in the room reads it the same way. Agree what each number means before anybody builds a dashboard on top of it. Write down how each figure is worked out in the same place the figure appears, so a new colleague can check you.

Watch your marketing site too. Scripts that slow your pages or set cookies your visitors did not expect are a problem long before they are a measurement problem.

When analytics is not the next step

A product with very few customers does not need a sophisticated setup. Talking to each customer, and noting why they joined and what they use, will teach you more than a dashboard. Add instruments when the conversations become too many to remember.

Example: steady revenue, quiet accounts

A vendor notices that revenue is steady but several accounts have logged in far less than before. A review of account-level activity shows that one department, which championed the product, has lost its sponsor. The vendor contacts the remaining users, offers to retrain them, and learns why the product slipped. Revenue alone would not have raised the alarm until the renewal date.

Questions to ask a vendor about your own data

• Can I see usage by account and not only by individual?
• Can I export the underlying records, not just reports?
• How far back does history go on my plan?
• Can I define events and measures myself?
• Who at the vendor can see my analytics data?

Analytics is only as trustworthy as access to its source. A product whose reports you cannot trace is asking for faith.

Reports to build first

If you are starting from nothing, build three: new accounts by week and where they came from; the share of new accounts that reached a first result; and accounts that have gone quiet in the last month. Review them with the people who can act, for example the person who onboards customers and the person who answers support. If a report does not lead to action in that conversation, drop it. Fewer, better reports are used more than many unused dashboards.

Related reading

Learn how one platform counts visits in its server-side analytics feature, view the reporting dashboard, or read about the B2B SaaS platform the numbers come from. The glossary entry on conversion rate covers a measure used on marketing sites.

How Microsite Generator handles this

On sites built here, page views and conversions are counted on the server. No JavaScript and no cookies are added to your site to make that happen.

You can also run an A/B test on a page, which changes the headline, lede and answer capsule and leaves everything else alone. Sites are capped at up to 25, 50 or 100 pages each, every page passes the compliance gate, and each one is checked against Google Search Essentials and modern web standards.

That keeps your marketing site light and your measurement simple, which is usually the right trade for a small team.

Updated October 2026