After every Salesforce launch, someone asks for an adoption report, and the number that comes back is almost always login count. Eighty-five percent of licensed users logged in this month. The room nods, the project is declared a success, and a few quarters later somebody notices the system is not paying back. The problem is the metric. A login measures that a person remembered their password and opened a browser tab. It says nothing about whether the work got better, which is the only thing the investment was supposed to buy.
We measure adoption with three numbers instead. Each one answers a specific question, and together they form a funnel that tells you, at a glance, where a rollout is succeeding and where it is leaking.
The three measures that matter
Think of adoption as a funnel. Everyone with a licence sits at the top. The investment pays back only for the users who reach the bottom, and the job of measurement is to show you where people fall out.
Reach
Share of relevant users who visited the specific page where the new functionality lives. Answers whether the rollout communication worked.
Engagement
Share of those who completed the intended action at least once. Answers whether the interface is usable.
Retention
Share who repeated the action in three of the last four weeks. Answers whether the work actually got easier.
Reach, engagement, retention
Reach is the percentage of users who should be using the new functionality who have actually visited the page where it lives. Not the homepage, not a dashboard, the specific page. Event Monitoring captures this; a custom log record on page visit works as a fallback. If reach is below seventy percent four weeks after launch, communication failed, and that is the first thing to fix.
Engagement is the percentage of those who reached the page and completed the intended action at least once: creating an opportunity, approving an order, logging a case outcome. A Flow updates an adoption record when the action fires, building a clean dataset over time. A big gap between reach and engagement means people arrived, looked, and bounced, which is a usability problem rather than a communication one.
Retention is the percentage of engaged users who completed the action in three of the last four weeks, and it is the measure that matters for return on investment. The system pays back when behaviour changes durably. A user who does the new thing once is curious; a user who does it for four weeks running has changed how they work, and only the second user is delivering the business case.
Reach measures the announcement. Engagement measures the interface. Retention measures whether the work actually got easier. Only the third one pays the invoice.
How to instrument it
The cleanest approach is a custom Adoption Event object. Each row records a user, an action, and a timestamp; the actions you care about live in metadata, and reports aggregate over the object. Capture reach with a small Lightning Web Component that fires once per visit per session, capture engagement from the same Flow that performs the action, and compute retention with a weekly aggregation that tags users who cross the threshold. The instrumentation has to be light enough that adding a new tracked action takes ten minutes. If your adoption tracking is itself a project, it will quietly stop being maintained.
The monthly adoption review
Three numbers, reviewed monthly by the product owner of the rollout, with a fixed shape to the conversation. If reach is low, what was the communication and what changes next month. If engagement is low, what is the friction in the interface and what is the smallest fix. If retention is low, what is wrong with the value proposition itself, and has the team quietly fallen back to a workaround. The meeting exists to leave with exactly one intervention to ship next month, not a list of twelve.
Adoption is never uniform. There is usually a team where retention is half the average. Investigate, do not message harder. The laggard almost always has a faster workaround or an uncovered process step. Find it, then incorporate it or build the missing piece. Communication alone does not move retention.
The honest version
A Salesforce rollout is a product launch inside your own company, and the same rules apply. If usage drops three weeks in, the product is not finished. If the team that built it stops measuring the moment the dashboard turns green, it will quietly stop being used, and the licence becomes a line item nobody can defend. Six months after launch, a well-designed rollout sits above eighty-five percent reach, seventy percent engagement, and fifty-five percent retention. If you are below those, the problem is fixable, but not by sending another email. It is fixed by treating adoption as the feature it always was.