Monday, June 15, 2026
jynlab

notes on building, judging, and selling small software

Case Study · Build-to-Exit

PostHog vs Plausible vs Mixpanel vs GA4: The Solo Builder's Analytics Guide

You shipped the app. Users are signing up. But you have no idea what they do inside it. Website analytics vs product analytics, real pricing, and which to pick.


You shipped the app. Users are signing up. But you have no idea what they actually do inside it. Time to add analytics. The question is which kind.

The short version
  • PostHog: product analytics plus session replay plus feature flags. Free up to 1M events per month. The all-in-one for SaaS.
  • Plausible: privacy-first web analytics. No cookies, GDPR-ready. $9 per month. The GA4 replacement.
  • Mixpanel: funnel and retention analytics. Free up to 20M events per month. Best pure product analytics.
  • GA4 (Google Analytics): free, powerful, complex. Privacy concerns. The default nobody loves.
  • Recommendation: PostHog for product analytics. Plausible for website analytics. You might need both.

Two types of analytics, two different jobs

Before picking a tool, decide what you are measuring:

Website analytics (page views, traffic sources, bounce rate): tells you how people find and move through your marketing site. Plausible or GA4 handle this. These tools care about sessions and sources, not individual user actions.

Product analytics (user actions, funnels, retention, feature usage): tells you what users actually do inside your app after they sign up. PostHog or Mixpanel handle this. These tools care about events, cohorts, and conversion paths.

Most SaaS builders need both. Website analytics for marketing: what content drives signups, which ad converts, where people drop off on the landing page. Product analytics for the app: which features stick, where users abandon onboarding, what power users do that free users don't.

Product Analytics

PostHog: the all-in-one for solo SaaS

Pricing: Free tier includes 1M events per month and 5,000 session replay recordings. After that, $0.00031 per event. Self-hosting is available with no event limits.

Best at: Covering the full product analytics stack in one tool. Funnels, retention charts, user paths, session replay, feature flags, and A/B testing live in the same dashboard. For a solo builder, it eliminates the need for three separate tools.

Limitation: The initial setup requires adding an SDK and defining your event schema. The dashboard has a learning curve. If you want something you can install and understand in ten minutes, PostHog is not that.

A SaaS with 1,000 users generating 50 events per session, 10 sessions per month equals 500,000 events. Well within PostHog's free tier.

Website Analytics

Plausible: the simple GA4 replacement

Pricing: $9 per month for up to 10,000 monthly pageviews. $19 per month up to 100,000 pageviews. Annual billing saves 33%.

Best at: Replacing GA4 for marketing site analytics without the cookie banner, the complex UI, or the privacy headaches. The script is under 1KB, GDPR-compliant out of the box, and the dashboard shows everything you need in a single view.

Limitation: Plausible is website analytics only. No funnels, no user-level data, no event tracking beyond goals, no session replay. If you try to use it as your product analytics tool, you will hit a wall fast.

Product Analytics

Mixpanel: the best pure product analytics

Pricing: Free tier includes 20M events per month with full access to funnels, retention, and user-level analytics. Growth plan at $28 per month adds data exports and more history.

Best at: Deep funnel analysis and retention charts. If you need to understand exactly where users drop out of onboarding or which cohorts retain, Mixpanel's charts are cleaner and more actionable than PostHog's.

Limitation: No session replay, no feature flags, no A/B testing. You get product analytics and nothing else. For website analytics, you still need a separate tool.

Website Analytics

GA4: the default nobody loves

Pricing: Free with unlimited data. GA4 360 (enterprise) starts at roughly $50,000 per year, but no solo builder needs that.

Best at: Marketing attribution and BigQuery export. If you run paid ads across multiple channels and need to tie revenue back to campaigns, GA4's conversion modeling is still the most capable free option.

Limitation: The UI is genuinely confusing. Cookie consent is required under GDPR, which adds friction to your site. Data gets sampled at high volumes. Google can change the product or pricing at any time. Many builders replace it with Plausible and do not look back.

Honorable mentions

Amplitude: Similar feature set to Mixpanel. Free tier covers up to 50,000 monthly tracked users (MTU). Slightly more enterprise-focused UI, but a strong alternative if Mixpanel's pricing ever becomes an issue.

Fathom: $15 per month, privacy-focused website analytics, simpler than Plausible. EU isolation option keeps data off US servers entirely. Worth considering if you serve European users and need strict data residency.

Comparison table (mid-2026)

ToolTypeFree tierPaid starts atBest for
PostHogProduct1M events, 5K sessions/mo$0.00031/event afterAll-in-one product analytics
PlausibleWebsiteNone (30-day trial)$9/mo (10K pageviews)Simple GA4 replacement
MixpanelProduct20M events/mo$28/moFunnels and retention
GA4WebsiteUnlimitedFree (GA4 360 = enterprise)Multi-channel attribution
AmplitudeProduct50K MTU/moCustom pricingEnterprise-style product analytics
FathomWebsiteNone (7-day trial)$15/moPrivacy-first, EU residency

Decision tree

Do you have a marketing site AND an app?
  Yes -> Use Plausible (site) + PostHog (app)
  No, just a marketing site -> Use Plausible
  No, just an app -> Use PostHog

Do you need session replay?
  Yes -> PostHog (replay included in free tier)
  No -> Mixpanel (larger free event tier)

Do you run paid ads across multiple channels?
  Yes, and you need attribution -> Consider GA4 alongside Plausible
  No -> Skip GA4

Do you serve EU users with strict data residency needs?
  Yes -> Consider Fathom instead of Plausible

FAQ

Is PostHog really free?

Up to 1M events per month, yes. After that, $0.00031 per event. A SaaS with 1,000 users generating 50 events per session and 10 sessions per month equals roughly 500,000 events. That is well within the free tier. Most early-stage SaaS products never hit the limit.

Should I use GA4 or Plausible?

Plausible if you value simplicity and privacy compliance. The setup is faster, the dashboard is readable without training, and you skip the cookie consent requirement. GA4 if you need advanced marketing attribution across paid channels and you don't mind the complexity. Most solo SaaS builders choose Plausible and don't regret it.

Do I need both website and product analytics?

Yes, if your SaaS has a marketing site and a separate app. Plausible on the marketing site tells you where signups come from. PostHog inside the app tells you what those signups do next. They answer different questions. Using only one means you are blind on one side of the funnel.

Can I add analytics later or should I start from day one?

Start from day one. You cannot retroactively see what users did before you added tracking. The first 100 users are your most valuable learning signal. If you wait until you "have real users," you will already have missed the data you needed most.

Before you track users

Analytics tells you what users do. Validation tells you whether the right users will show up at all. Make sure the niche is real before wiring up the dashboards.

Validate your idea free →

Get tool guides and teardowns in your inbox