How to Add Analytics to Your Vibe Coded App (2026)
You shipped your first app. Now how do you know what is happening? A practical guide to analytics for vibe coders - from choosing a tool to filtering test traffic.
We evaluate every tool based on published features, real-world usage, community feedback, and independent testing where possible. Affiliate commissions never influence our rankings. How we research · Editorial policy
How do you know what is happening?
You just shipped your first app. Maybe you built it with Cursor, maybe Lovable, maybe Claude Code and a weekend of determination. It works. It is live. You posted it on Reddit and someone said "cool." You feel great.
Now what?
Is anyone using it? Where did they come from? Did that Reddit post actually drive traffic, or was it just your mum checking it on her phone? Which page do people land on? Where do they leave? Is the signup flow working or is everyone bouncing at step two?
Without analytics, you are building blind. And every day without it is data you will never get back. You cannot retroactively learn how people used your product in its first month.
First question: what kind of business do you want to be?
Before you install anything, you need to decide what you actually want analytics for. This is not a technical question - it is a business question.
Do you want to improve your product? Understand where users get stuck? Figure out which features matter? That is product analytics - session recordings, funnels, event tracking.
Do you want to understand your traffic? Where visitors come from, which pages they read, whether your content strategy is working? That is web analytics - pageviews, referrers, countries.
Do you want to retarget users with ads, build audience profiles, and sell that data to third parties? That is the Google Analytics model. There is nothing wrong with it - plenty of successful businesses run this way. It is just a different choice with different trade-offs.
We chose the privacy-first route. Not because the other way is wrong, but because we wanted our brand to be a breath of fresh air for users who are tired of being tracked across the internet. That decision shaped everything that followed.
There is no right or wrong answer here. Just different trade-offs. The important thing is to decide consciously, not by default.
The ecosystem trap
Google Analytics is free. It is also the on-ramp to the Google ecosystem. Google Ads, Google Search Console, Google Tag Manager, Google Optimize - once you are in, everything integrates beautifully with everything else. That is by design.
Amazon AWS works the same way. So does Cloudflare. You start with one service, and before you know it you are five services deep and migration would take months.
For a startup, this matters. You are making decisions now that will be expensive to reverse later. If you choose Google Analytics today, moving to something else in a year means rebuilding your dashboards, losing historical data, and retraining yourself on a new tool.
That does not mean you should avoid these ecosystems. If you want everything under one roof and you trust the provider, it can be genuinely simpler. Google runs the biggest search engine you want to appear in - the integration between Analytics and Search Console is real and useful.
Just go in with your eyes open.
Your actual options in 2026
The analytics market has split into three clear lanes:
Privacy-first web analytics
These tools tell you what is happening on your site without tracking individual users. No cookies, no consent banners, no GDPR headaches. They answer: how many people visited, where did they come from, which pages did they read.
Fathom Analytics ($14/month for 100K pageviews) is the most polished option. Clean dashboard, fast, EU-hosted option available. It does one thing well and stays out of your way.
Plausible ($9/month for 10K pageviews) is cheaper to start with and open-source. You can self-host the Community Edition if you have a VPS and want to pay nothing.
Umami is the fully self-hosted option. MIT-licensed, runs on 512MB of RAM, and you can have it running in 15 minutes via Docker. We use Umami across all our projects. The setup was genuinely easy - Docker container, a subdomain, done. Total cost: nothing beyond the VPS we already had.
Product analytics
If you need to understand how people use your product - not just which pages they visit but what they click, where they get stuck, and which features they ignore - you need product analytics.
PostHog is the standout for indie founders. The free tier is extraordinary: 1 million events, 5,000 session replays, and 1 million feature flag requests per month at zero cost. It effectively replaces Mixpanel, Hotjar, and LaunchDarkly in a single platform.
PostHog can be self-hosted, but the team themselves say they have "literally never seen the math work out" for self-hosting versus their free cloud tier. It requires 16GB RAM and runs 16+ containers. Use the cloud.
Full tracking (the Google model)
Google Analytics 4 is free and gives you everything: user journeys, conversion funnels, audience segments, e-commerce tracking, ad integration. If you are running Google Ads or plan to, the integration is genuinely useful.
The trade-off: your users get tracked across the internet, you need cookie consent banners, and you are adding to Google's data profile of your visitors. For some businesses this is fine. For ours, it was not the relationship we wanted with our users.
Mixpanel (1M events/month free) sits between privacy-first and full-tracking. Powerful product analytics but requires a credit card to start and is designed to upsell you into enterprise plans.
What we actually did
We set up Umami on our existing VPS via Docker. One subdomain, one Docker container, one PostgreSQL database. The whole process took about 15 minutes. It costs us nothing beyond the server we were already paying for.
We check it daily - probably too much, honestly. Visits, countries, referrers. We set up a daily digest email that arrives each morning with the previous day's stats. That is actually more useful than logging into a dashboard because it comes to you.
One mistake we made: we did not filter out our own test traffic early enough. When you are building and constantly refreshing your own site, those visits pollute your data. Set up IP exclusions or localhost filtering before you go live, not after.
If we were not comfortable self-hosting, we would pick Fathom. It is the tool we most often recommend to founders who just want analytics without managing infrastructure.
Set up analytics before you launch, not after. Filter your own traffic from day one. Every day without data is data you will never have.
If you built with Cursor, Lovable, or Bolt
Adding analytics to a vibe-coded app is straightforward. For any Next.js app (which is most of them), you are adding a single script tag to your root layout.
For Fathom or Plausible, it is one line in your HTML head. For Umami, same thing - a script tag with your website ID. For PostHog, you install their npm package and initialise it in your app.
The AI tools (Cursor, Claude Code) are good at this. Ask them to "add Fathom analytics to my Next.js app" and they will put the script tag in the right place. This is not one of the things that breaks.
What AI tools are less good at: deciding which analytics tool to use, understanding the privacy implications, and knowing that you should filter test traffic. That is why you are reading this instead of asking Cursor.
What to do now
Pick a lane: privacy-first web analytics, product analytics, or full tracking. If you are unsure, start with privacy-first - you can always add product analytics later, but you cannot undo giving Google your users' data.
If you want a detailed head-to-head comparison of the privacy-first options, we wrote one: Fathom vs Plausible vs Umami, with real pricing breakdowns and the self-hosted angle that nobody else covers.
You cannot improve what you cannot measure. Set it up before launch, filter your test traffic, and check it regularly. The tool matters less than having one at all.
Frequently Asked Questions
Especially then. The MVP stage is when you need to learn the fastest. Without analytics, you have no idea if your MVP is working. You are making decisions based on feelings, not data.
It depends on your perspective. GA4 tracks users across the internet, requires cookie consent banners in the EU, and feeds data into Google's advertising profile of your users. Some founders are fine with this. Others are not. Neither position is wrong.
You can add it any time, but you cannot backfill data. Every day without analytics is user behaviour data gone forever. If you launch on Monday and add analytics on Friday, you will never know what happened Monday through Thursday.
If you have a VPS and are comfortable with Docker, about 15 minutes. One docker-compose file, one subdomain pointed at it, done. If Docker is new to you, use Fathom or Plausible cloud instead - they are worth the monthly cost for the simplicity.