Analytics for a new SaaS: PostHog vs Google Analytics vs Plausible vs Umami
What each one is good at, what it costs at startup scale, and the handful of numbers worth watching in the first month.
Short answer: PostHog is the best fit for a product: sign-ups, where people drop off, session replays and feature flags, with 1 million events a month free. Plausible or Umami are simpler and privacy-friendly for your marketing site. Google Analytics is free and powerful but built for marketing, not for understanding how people use your app.
Compared
| Tool | Best for | Free | Paid | Pros | Cons |
|---|---|---|---|---|---|
| PostHog | Product usage | 1M events, 5,000 session replays, 1,500 survey responses a month | Usage-based after that | Funnels, replays and flags in one place | Lots of features to learn |
| Google Analytics | Marketing traffic and ads | Free | Enterprise version for large companies | Free, connects to Google Ads | Complex; usually needs a cookie banner in the EU |
| Plausible | A simple marketing site dashboard | 30-day trial | From $9 a month for 10,000 page views | No cookies, no banner needed, one-page dashboard | Not built for in-app events |
| Umami | Simple, privacy-friendly stats | Cloud Hobby: 100,000 events a month, 1 site; free to self-host | Pro $20 a month | Open source; cheap | Fewer product features |
The numbers I’d watch in month one
- Visits to sign-ups. Is the page clear about what you do?
- Sign-ups to first success. Did they reach the moment the product is for, like sending a first invoice?
- Week-one return. Did they come back?
- Where they stop. The screen most people leave from is where to look first.
Questions
Do I need a cookie banner?
It depends on the tool and where your users are. Cookie-free tools like Plausible say no banner is required. For others, check the rules for your users’ countries.
Should I use two tools?
Many startups use a simple tool on the marketing site and PostHog inside the product. Don’t add more until you look at what you have weekly.