Short answer: before launch, a privacy-friendly web analytics tool like Plausible or Umami is all you need. From your first paying users, add product analytics: PostHog's free tier covers most micro-SaaS products for a long time. Once you have dozens of customers and no hours left to read funnels, the bottleneck is not data but decisions, and that is the stage Holdy's Free plan was built for.
Why the usual advice fails solo founders
Most "best analytics" lists are written for teams with a data person. An indie hacker running a micro-SaaS has the opposite constraint: plenty of questions, no hours. Setting up a tool is easy; the expensive part is the weekly ritual of opening dashboards, building the funnel, reading it and deciding what to ship. If that ritual doesn't happen, the fanciest analytics stack is a cost with no return.
So the right question is not "which tool is best" but "what do I need to see at this stage, and who is going to read it". The answer changes three times in the life of a small product.
Stage by stage
| Stage | What you need to see | Good fit | Cost in October 2026 |
|---|---|---|---|
| Pre-launch and launch | Where visitors come from and which page converts to sign-up | Plausible, Umami, Fathom, Simple Analytics | Umami self-hosted is free; the hosted tools start around US$9 to US$15 a month |
| First paying users | Which steps new users complete, where they drop, who comes back | PostHog or Mixpanel free tier | Free up to 1M events a month on both |
| Dozens of customers, no analyst | Why trials don't convert, which accounts went quiet, what to change first | Holdy on top of your events | Free plan with no card, up to 20 active customer accounts, 5 questions and 3 AI-generated improvements a month |
Prices above are the ones each company published in October 2026. Check their pricing pages before deciding.
Stage 1: pre-launch. Web analytics, nothing more
Until people are using the product, there is no product behavior to analyze. What matters is the landing page: traffic sources, which section gets clicks, how many visitors sign up. Plausible, Fathom and Simple Analytics give you that on one screen without a cookie banner; Umami does the same and can be self-hosted for free. Resist installing a full product analytics suite now. An empty funnel teaches nothing and the setup time is better spent on the product.
Stage 2: first customers. Add product analytics, keep it small
When the first users are inside, you need to know three things: which steps a new account completes in the first days, where it stops, and whether it comes back the next week. PostHog and Mixpanel both offer free plans up to 1 million events a month, which is more than a micro-SaaS will send for a long time. PostHog adds session replay and feature flags on the same free tier, so for a solo technical founder it is usually the better default.
Two habits make this stage work. Track accounts, not just people: if your customer is a company, attach every event to the company so you can later ask "which customers activated", not "which users clicked". And keep the plan to five to ten events that mark progress (created the first item, invited a teammate, finished the core flow), not every button.
Stage 3: dozens of customers. The bottleneck becomes the decision
This is the stage where analytics stops paying off for most indie hackers. The data is there; the hours to read it are not. Trials expire without anyone noticing which step they got stuck on. Accounts go quiet weeks before they cancel. The roadmap is set by the loudest support ticket.
Holdy was built for that gap. It reads the events you already collect, by connecting PostHog or Mixpanel in one click or through its own SDK installed with one command, and finds on its own where activation and conversion stall. Each finding arrives as an opportunity with the evidence, the affected accounts and a written plan: what to change, for whom, how it will be measured. The change ships as an in-product hint for the right accounts, a message, or a draft pull request on your GitHub repo, with part of the accounts held out so you can see whether it moved conversion and MRR.
The Free plan has no card and no deadline: Free plan with no card, up to 20 active customer accounts, 5 questions and 3 AI-generated improvements a month. A company counts as one account no matter how many people it has; a person who pays on their own is an individual account, and every 10 count as one. Paid plans start when the base outgrows it: Launch at US$29 (BRL 149) a month up to 100 customer companies, Growth at US$99 (BRL 490) up to 1,000. Seats, events and integrations are never charged.
What to skip
- Enterprise analytics (Amplitude Growth, Pendo paid tiers) until there is someone whose job is to operate them.
- Building your own dashboards on top of your database. It feels free and costs weekends every month.
- Tracking everything. Ten events that mark progress beat two hundred that mark clicks.
- Any tool you won't open weekly. If you know you won't, pick the one that opens the finding for you.
How to choose in 3 questions
- Do you have paying users yet? No: web analytics only. Yes: add PostHog or Mixpanel's free tier.
- Will you read the funnel every week? If honestly not, you need something that finds the problem and proposes the change.
- Can you attach events to a customer account? If yes, Holdy can tell you which customers are stuck and what to ship; if not, start by adding that to your tracking.
Running a micro-SaaS and short on hours to read the data?
Holdy finds what blocks activation and conversion, ships the improvement ready for review, and measures the impact on retention and revenue against a holdout group. Start on the Free plan, no card.