Analytics for Beginners: What to Measure in Your First App

Once your app is live, you'll want to know: are people coming? Are they signing up? Are they actually using it? Analytics answers those questions. The trap is measuring everything and learning nothing. Here's how to set up just enough.

Two kinds of analytics

Web analytics — about visits: how many people came, from where, which pages they viewed, how long they stayed. Examples: Plausible, Fathom, Umami, Simple Analytics, Google Analytics, Cloudflare Web Analytics.

Product analytics — about what users do inside your app: signed up, created a project, invited a teammate, came back next week. Examples: PostHog, Mixpanel, Amplitude.

A marketing site mostly needs web analytics. An app needs a little of both.

The numbers that matter for a new app

Ignore most of what dashboards offer. For your first months, watch:

Visitors — are people arriving at all? From where? (Search, a Reddit post, a friend's tweet?)
Sign-up rate — what share of visitors create an account. Low means your homepage isn't convincing.
Activation — what share of new users do the main thing your app is for: create their first plant, send their first invoice. Low means the first experience is confusing.
Retention — what share come back a week later, and a month later. The most important number. People returning means you've built something useful.
Where people drop off — the step in your main flow where most people give up.

That's five numbers. Everything else is detail.

Privacy-friendly analytics

Traditional analytics tools use cookies to track individuals, often across sites. In the EU, UK, and elsewhere, that generally requires consent — the cookie banner. (What Is a Cookie?.)

Several tools — Plausible, Fathom, Umami, Simple Analytics, and Cloudflare Web Analytics — are designed to count visits without cookies or personal data. Many site owners use them without a consent banner, though you should check the rules that apply to you and mention analytics in your privacy policy. (GDPR Basics for App Builders.)

For a first app, a privacy-friendly web analytics tool is usually the simplest choice: one script tag, a clean dashboard, and no banner.

Tracking events

Page views don't tell you whether someone used your app. For that, you record events — named actions:

analytics.track("plant_created", { species: "fern" })
analytics.track("reminder_enabled")
analytics.track("subscription_started", { plan: "pro" })

Tips:

Name events consistently: object_action in past tense — plant_created, invite_sent.
Track the few that matter — sign-up, activation, the core action, payment. A dozen well-chosen events beat two hundred random ones.
Track important events on the server where possible (payments, sign-ups), since ad blockers stop many browser-side scripts.
Don't send personal data — names, emails, message contents — in event properties.

Your own database is analytics too. "How many users created a plant this week?" is one SQL query away. (SQL for Beginners.)

Funnels

A funnel shows how many people make it through each step of a flow:

Visited homepage     1,000
Started sign-up        180   (18%)
Finished sign-up       120   (67% of started)
Created first plant     60   (50% of signed up)

The biggest drop is where to focus. Here, half of new users never create a plant — so the first screen after sign-up is the thing to improve.

Where your visitors come from

Most tools show referrers automatically. To track your own links — a newsletter, a social post — add UTM parameters:

https://yourapp.com/?utm_source=reddit&utm_campaign=launch

Then you can see which launch channel actually brought sign-ups, not just visits. (Launch Your First App.)

Mistakes to avoid

Counting your own visits. Exclude yourself (most tools have a setting) or your numbers will be flattering nonsense early on.
Vanity metrics. Page views and total sign-ups only go up. Retention tells the truth.
Too many tools. One web analytics tool and, later, one product analytics tool is plenty.
Collecting data without saying so. Your privacy policy should mention it. (Does My App Need a Privacy Policy?.)
Watching numbers instead of talking to users. With 50 users, a conversation with five of them teaches more than any chart.

EasySpawn runs your app on your own domain with SSL, so first-party, privacy-friendly analytics work cleanly — and your database is right there in the workspace when you want to ask it a question directly. See how it works or join the waitlist.