Everyone’s telling you to be “data-driven”. Your analytics dashboard has 47 tabs. Your weekly reports are overwhelmingly long. Yet you're no closer to knowing if anyone actually wants what you're building.
Here’s what you’ve been missing: the more data you have, the less clearly you see. Yep, ignore most of your data. Sounds crazy? Well, at Lumi, we’ve seen time and time again that startups with simpler analytics find product/market fit faster.
Read along to stop measuring everything and start measuring what predicts success.
Why your current analytics setup is probably lying to you
42% of startups fail because they build something nobody wants.
Your analytics should be your early warning system – the canary in the product coal mine. Instead, we’ve seen a lot of founders focus on numbers that look good in investor presentations but reveal nothing about real user value.
→ There’s a huge disconnect: traditional analytics measure activity, not value creation. They track motion, not progress.
Most tracking focuses on what's seductively easy to measure (traffic, signups, clicks) rather than what's essential: whether you're building something people desperately need.
So what are the metrics that matter? ⬇️
The only 3 metrics that matter before product/market fit
Sure, exponential growth curves make for great pitch decks. But when you're still validating your core hypothesis, you need honest feedback loops, not feel-good dashboards.
👀 Read our full guide on how to test your idea before building it
Here are the numbers that actually predict success ⬇️
1. Users who experience your product's core value
Forget total signups. Forget app downloads. Even daily visitors can mislead you.
Track users who actually do the thing your product is built for. For Uber, it's getting to where you need to be. For Spotify, it's playing the music you love. What's yours?
Take a typical B2B startup. 10,000 "registered users" looks impressive until you realize only 1,500 have ever completed a meaningful action. Those 1,500? That's your real user base. The rest is just database inflation.
💡 What's the "moment of value" for your users? Can't answer instantly? That's your first problem to solve.
2. How many users stick around (and when they stop leaving)
Your investors care about revenue churn. But pre-PMF, you need to focus on something simpler: Do users come back?
Every product loses users over time, but eventually this loss slows down and stabilises. Some products level off at 65%, others at 50% (what you’d aim for typical B2B SaaS). The exact percentage doesn't matter as much as finding where YOUR curve flattens.
Why? Because the users still around after that point are your true believers. They're the ones who genuinely need what you've built. Double down on understanding these core users.
What this can look like ⬇️
You launch with 1,000 users. Here's what typically happens:
- Week 1: 400 users return (60% gone)
- Week 4: 200 users return (80% gone)
- Week 12: 180 users return (82% gone)
- Week 16: Still ~180 users (curve flattens)
When that retention curve flattens, it’s usually a pretty sturdy sign of PMF. 🙌
3. The "would you miss us?" test
Ask your users: "How would you feel if you could no longer use our product?"
When 40% or more say "very disappointed," you're onto something. Below that? Keep iterating.
The key: look at different user groups separately. Sometimes you've built something perfect for a specific type of user you haven't identified yet.
Your lean analytics toolkit (without the enterprise price tag)
You don't need expensive analytics software yet. Here's the simple stack that gets the job done ⬇️
1. Google Analytics (free and good enough)
Set up tracking for these key moments:
- Someone signs up
- They complete setup
- They use your core feature
- They come back a second time
- They invite someone else (if relevant)
Skip the demographic reports. Focus on the path from signup to actually using your product.
2. Mixpanel or Amplitude (for understanding user behaviour)
The free plan is plenty when you're starting out. Use it to:
- See how user behaviour changes over time
- Track which features people actually use
- Identify your most engaged users
- Compare users from different sources
💡 Your signup-to-first-value flow is everything early on. If users aren't experiencing value quickly, nothing else matters.
3. Hotjar or FullStory (to see what's really happening)
Recording actual user sessions shows you what people really do, not what they say they do. Watch 10 recordings of users who left after one visit. You'll learn more than from 100 surveys.
Look for:
- Where users click frantically (frustration)
- When they abandon important processes
- How long they pause (confusion)
- Which features they completely ignore
4. Simple feedback tools
In-app surveys beat email surveys every time. Ask:
- What's the main value you get from this?
- What's missing?
- How did you handle this before finding us?
- Would you recommend this to a friend?
Keep it short. 2-3 questions maximum. Response rates plummet after the third question.
Pre-product/market fit website analytics checklist
Core value validation
☐ What specific action shows a user gets value from our product?
Define your "aha moment" in one sentence. This should be measurable and happen early in the user journey.
Example: "User creates their first project" or "User completes first transaction"
Activation metrics
☐ What percentage of users complete this action within a week?
Track: [___]% of new users activate within 7 days
Benchmark: Aim for 20-40% depending on product complexity
If below 10%, your onboarding needs immediate attention
Retention analysis
☐ At what point does our user retention stabilise?
Week [___] is when our retention curve flattens
Percentage at stabilization: [___]%
This is your true user base – optimize for them
User sentiment
☐ What percentage of users would be "very disappointed" if we shut down?
Current percentage: [___]%
Target: 40%+ indicates product/market fit
Run the Sean Ellis test monthly with active users
Segment performance
☐ Which user segment has the best retention?
Best performing segment: ________________
Their 30-day retention: [___]%
Key characteristics: ________________
Consider pivoting to serve this segment better
Growth velocity
☐ How fast are we growing in engaged users?
Week-over-week growth: [___]%
Target: 5-7% weekly growth pre-PMF
Focus on engaged users, not total signups
Channel quality
☐ Which channels bring users who stick around?
Top channel by retention: ________________
30-day retention rate: [___]%
CAC for this channel: $[___]
Double down on quality channels, not quantity
The bottom line
Before product/market fit, your analytics should be simple, focused, and brutally honest. Track the few things that matter. Ignore everything else. And when the data tells you something uncomfortable, listen.
Need help figuring out which metrics actually matter for your startup? Let's talk. 💌
