Metrics from Day Zero: Funnels, Churn and Monthly Reports for Websites and Systems

Metrics from day zero means a website or system goes live already measuring what matters: where users come from, at which step they give up, which features they use, who joined, who left and why. Without it, every product decision is an opinion. With it, the conversation moves to concrete, real data.

In this guide you will see why to measure from launch, what to measure on a website and in a system, how to read a conversion funnel, how to track churn, how numbers turn into decisions and what a good monthly report should include.

Why set up metrics from day zero?

Because data that was not collected never comes back. If a website launches without measuring its funnel, nobody can say three months later whether the new contact page converts better than the old one. If a system goes into production without events on its main features, there is no way to know what is used and what just takes up space in the menu.

Measuring from the start brings three gains:

• A baseline: you have the before to compare with the after of every change.
• Problems show up early: a step where many people give up appears in the first weeks, not after a quarter of lost sales.
• Priorities with criteria: the backlog is ordered by measured impact, not by whoever asked loudest.

Adding metrics later is possible, but it costs more: you have to change code already in production, and the history starts from zero.

What to measure on a website

On a corporate or lead generation website, metrics revolve around acquisition and conversion. The most useful website KPIs are usually:

• Traffic source: organic search, social media, campaigns, referrals and direct visits, with campaigns tagged so you know what brought each visit.
• Entry and exit pages: where people arrive and where they leave.
• Conversion funnel: from visit to service page, to the form and to the submitted contact.
• Conversion rate by source: which channel brings visits that turn into conversations.
• Content engagement: post reads, clicks on calls to action and paths to contact.
• Speed and errors: slow pages hurt conversion. The post on endless loading shows why.

Connections to analytics and search tools are part of the initial setup, along with the consent banner.

What to measure in a system

In a web system, an app or a platform, the question is no longer just "who arrived" but "who uses it, how they use it and who keeps using it". The most important product metrics are:

• Active users: how many use it per day, week and month.
• Task funnels: sign up, first use, purchase, issuing a document, any flow that creates value.
• Steps where users get lost: the screen where most people abandon the flow.
• Most and least used features: what deserves investment and what can be simplified.
• Joins and exits: new users, returning users and users who stopped.
• Churn and retention: how many stop using or cancel in each period.
• Errors and slowness by flow: where the experience breaks.

The user journey designed in the project becomes the basis for the events: every important step in the journey gets a measurement.

How to read a conversion funnel

A funnel shows how many people go through each step of a flow, in order. Reading a funnel well means looking at three things:

1. The biggest drop between two steps: that is where the best opportunity is. If 60 out of 100 people open the form and only 10 submit it, the problem is the form, not the traffic.
2. The drop by segment: the same step can work on desktop and fail on mobile, or work for search visitors and not for ad visitors.
3. The trend: a funnel that gets worse after an update signals that something broke.

The numbers in the example above are illustrative. The point is that the funnel tells you where to act before it tells you what to do. What to do comes from combining it with other sources: session recordings when consented, user conversations and tests.

Churn: who left and why

Churn is the rate of users or customers who stop using or cancel in a period. In a subscription system it is cancellation. In an internal system it may be the team that went back to the spreadsheet. In an app, the user who installed it and never came back.

To track churn usefully:

• Define what "left" means for your product: cancelled, 30 days without access, stopped using the main feature.
• Look by cohort, that is, by group of users who joined in the same period. That shows whether recent changes improved retention.
• Cross it with behavior: people who leave often skip a setup step, never use a key feature or run into errors more often.

Churn alone only says there is a problem. Crossed with the funnel and feature usage, it points to where to fix.

How metrics turn into decisions

A nice dashboard without decisions is a cost. Metrics are only worth it when someone turns them into priorities. In practice, the cycle looks like this:

Examples of decisions that come from data:

• Simplify the sign up step where most people give up.
• Bring the feature everyone looks for to the first screen.
• Remove or rethink a feature almost nobody uses that costs maintenance.
• Strengthen the content that brings the most qualified leads from search.
• Fix first the error that appears in the highest value flow.

What a good monthly report includes

In UEEK's ongoing support, the monthly report is when data turns into a conversation with the client. A good report has:

• A one page summary: what improved, what got worse and what needs attention.
• Main funnels: the rate at each step compared with the previous month.
• Product usage: active users, most and least used features.
• Joins, exits and churn: who arrived, who returned, who stopped.
• Technical health: availability, errors, slowness and updates made.
• Prioritized opportunities: what to do next month, with the reason for each item.
• Results of previous changes: whether what was done last month worked.

The last item closes the loop. Without it, the report is a snapshot, not a follow-up.

Privacy and LGPD

Measuring does not mean collecting everything. In general terms, a good setup:

• Respects consent: measurement with cookies and identifiers only after acceptance, with basic anonymous measurement when the person does not accept.
• Collects the minimum necessary and keeps personal data such as names, emails or ID numbers out of events.
• Masks sensitive fields in session recordings.
• Defines who can access the data and how long it is kept.
• Is clearly described in the privacy policy, in line with LGPD, Brazil's data protection law.

This care is part of the initial setup, not a later fix. For the security side of the product, see the post on DevSecOps.

Frequently asked questions

Are product metrics and website metrics the same thing?

No. Website metrics look at acquisition and conversion: who arrives and who becomes a lead. Product metrics look at usage and retention inside the system: who uses it, what they use and who stays.

Can metrics be set up in a system that is already live?

Yes. UEEK can take over an existing system, including one built by another vendor, and add events and funnels as part of ongoing support. The history starts from the setup. The post on software support explains how taking over a third-party system works.

Do I need a data team for this?

No. The goal is precisely to deliver the analysis ready in the monthly report, with prioritized opportunities, without the company having to build an analytics team.

Which tools are used?

It depends on the product and where it is hosted. We use market analytics and product analysis tools, set up in the client's environment with the data under the client's control.

Does this only apply to products hosted by UEEK?

Day zero setup is standard on the websites and systems UEEK builds and hosts with ongoing support. For products hosted elsewhere, we assess the access needed case by case.

How does this relate to squads?

In a dedicated squad, metrics guide backlog priorities. The monthly report shows the effect of what the team delivered.

How UEEK can help

UEEK has worked with digital products for 14 years, for more than 200 clients, in 6 countries, across more than 500 projects, from local businesses to Itaú. On the custom websites and systems we build and host, metrics go in on day zero. With ongoing support, we look after the live product and deliver a monthly report with funnels, usage, churn and opportunities. When demand is continuous, development squads turn those opportunities into deliveries. Custom software is born with this measurement in place.

Talk to UEEK about ongoing support with metrics