blog · Jul 5, 2026
how to add analytics to an app you built with ai (lovable, bolt, v0, replit)
Arjun Varma · maker of smolanalytics
if you built your app with Lovable, Bolt, v0, or Replit and you do not write code, the easiest way to add analytics is smolanalytics: you sign up, it hands you one prompt, you paste that prompt into your builder, and your app wires itself up. you never open a code editor or a terminal. from then on you get three things. a dashboard that answers plain-English questions like "how many people signed up this week?". a verdict at the top that says what is broken and where people give up. and a short email every morning with what is working and what to fix first. it can run without cookies, so there is no consent banner to add. the trial is 14 days with every feature and no credit card; after that it is $9 a month, covering every app you build, never per site. it is open-source, and your users' data is never sold or used to train anything.
how do i add analytics to an app i built with ai?
you paste a prompt. that is the whole setup, and it is worth being specific about why it is that simple.
when you sign up, smolanalytics generates a prompt written for your builder, with your real keys already filled in. you copy it, paste it into Lovable (or Bolt, v0, Replit), and the builder does the wiring: it loads the small analytics script on every page and tracks your two or three key moments, like signup and checkout. you do not touch code, because the AI that built your app is the thing reading the prompt and making the change.
the prompt looks roughly like this (you get the real one, with your keys, right after signup):
Add analytics to my app: load https://your-app.fly.dev/sdk.js on every
page, then run smolanalytics.init("sa_xxxxxxxx", { host: "https://your-app.fly.dev" }).
Then track my 2 or 3 key moments, like signup and checkout.
the full walkthrough for each builder is on analytics for Lovable, Bolt, v0 and Replit.
do i need to open a terminal or write code?
no. that is the point of the paste-prompt flow.
most analytics setup guides assume you can edit a config file, run a command, or drop a script tag in the right place in your HTML. if you built your app by describing it to an AI, none of those steps are things you do, and being told to "just add this snippet to your <head>" is exactly where these tools lose you. so the setup meets you where you already work: the builder that wrote your app writes the analytics too. one paste, and you are done.
how do i actually read the numbers?
you ask, like you would ask a person.
there are no charts to decode and no report to build. you type a question into the ask bar ("did anyone use the new feature?", "where do people give up during signup?", "how many people came back today?") and you get the real number back in plain English. the number is counted from what actually happened in your app; it is never made up. that matters more than it sounds: the tool is built so it returns the true computed number or nothing, so you can trust the answer enough to act on it.
here is the shape of an answer (these are example numbers, not real traffic; this launched recently):
you ▸ where are people giving up during signup?
app ▸ [140] people started signup this week, [88] finished, 63%.
the drop is on the email-verification step: [34] never
came back after we sent the code.
what is the morning email?
it is a short digest that arrives on its own, every morning, and gets to the point: what is working, what broke, what to fix first. if traffic spiked overnight or signups quietly stopped, you hear about it from the email before your users tell you. you do not have to remember to open a dashboard; the dashboard comes to you.
this is the part that changes the habit. most people add analytics, look at it twice, and never open it again. a morning email you actually read means you notice the drop the day it happens, not the week you finally check.
will i need a cookie consent banner?
no. smolanalytics can run in a cookieless mode, so there is nothing stored on your visitors' devices and no consent popup to add to your app. you still see visitors, signups, and where people give up; you just do not have to wrap your beautiful new app in a banner to get it.
what does it cost?
the trial is 14 days with every feature and no credit card. after that it is $9 a month, and that covers every app you build, never per site, so if you ship three apps this month, it is still $9. if you go over the included events it is $5 per extra million, with a receipt emailed to you, and you are never locked out of your dashboard. if you run a whole portfolio of little apps, that pricing is the entire pitch on analytics for indie hackers.
and if you are the kind of builder who does live in an editor (Cursor, Claude Code) there is a version of this aimed at you too, where you ask your analytics questions right in the editor you code in: analytics in Cursor and Claude Code.
the honest part
this is one person, it launched recently, and there are no testimonials on the site because there are none yet. none will be invented. so the reason to trust it is not a logo wall. it is that the numbers are counted from your real events and cannot be faked, the setup is genuinely one paste, and your data is yours: never sold, never used to train anything.
the code is open source at github.com/Arjun0606/smolanalytics. if you would rather just try it, start the 14-day trial: paste one prompt into your builder, and tomorrow morning there is an email telling you what is working.
smolanalytics is the analytics that tells you what to fix — try the cloud or self-host free.