One of the most common problems I see with clients? They measure too many things — and paradoxically get almost nothing out of it.
Dozens of dashboards, hundreds of events in GA4, reports nobody reads.
The problem isn’t a lack of data. The problem is that nobody said what matters right now.
That’s exactly what the Lean Analytics framework addresses. Its core idea is surprisingly simple: at every stage of your business, there’s one key metric worth focusing on. Everything else is a bonus.
The framework is business-model agnostic — it works for e-commerce, SaaS, media sites, and marketplaces alike. Only the specific metrics change depending on your stage. The principle stays the same: focus on what matters most right now.
Why just one metric?
- It forces you to name what’s most important right now
- It aligns the team around a single goal
- It filters out noise from hundreds of dashboard numbers
And here’s the thing — that metric changes. It’s different at every stage. That’s the real power of this approach.
Five Stages, Five Different Questions
Lean Analytics identifies five stages every business goes through. The stages are sequential: there’s no point in skipping ahead, because you’re scaling whatever you built in the previous step. And if you built it wrong, you’re scaling the problem.
1. Idea — Does Anyone Care?
The goal of the first stage isn’t to measure traffic. It’s to validate whether the problem you want to solve actually bothers someone — and whether they’re willing to pay for a solution.
You don’t need GA4 here. You need interviews, surveys, and experiments.
What to focus on:
- How many people are actively trying to solve this problem?
- How badly does it hurt? (Is there willingness to pay?)
- Is the market big enough?
Output: A validated business plan and a clear picture of who you’re building for.
2. Stickiness — Does It Work?
You have an idea. Now you need an MVP — a minimum viable product that proves the whole thing holds together.
Running an e-shop? Then it’s a store with a checkout process. No blog, no extra features. Selling subscriptions? You care whether people accept the offer and don’t churn after the first month.
At this stage you’re solving three things:
- Technical — does it work without critical bugs?
- Usability — can people use the product without calling support?
- Business — does it fit the planned budget?
Key metrics describe whether the product is usable and whether people stick around. For example:
- Daily or weekly active users
- Retention — how many people come back after a week? A month?
- Churn rate — how fast are users leaving?
- Engagement rate — does the website or product capture potential customers’ attention?
- Conversion rate — can people actually complete a purchase?
As long as people are leaving, there’s no point in bringing more in.
Something broken? Go back to the previous stage. Everything checks out? Move on.
3. Grow — Is It Growing?
Have you truly been through the entire Stickiness stage? Everything checks out? Welcome to the growth phase.
If it doesn’t check out and you push forward anyway — you’re scaling a loss.
At this stage, you’re figuring out how to reach a critical mass of users. For an e-shop, that means the budget tips into the black. For a platform, the network effect kicks in — the more people use the service, the more sense it makes to join.
There are multiple growth engines, and they depend on your business type. An e-shop can grow primarily through paid acquisition and retention. A SaaS product might benefit from natural sharing among colleagues. Some products combine both. What matters is knowing what works for you — and investing in that.
Key metrics focus on acquisition efficiency and the ability to retain customers. For example:
- Customer acquisition cost (CAC)
- New vs. returning users
- Visitor-to-customer conversion rate
4. Revenue — How Much Does It Earn?
You’ve been through the growth phase. You have users. Now it’s time to squeeze the most out of your business model.
The goal isn’t just “to make money”. The goal is to understand exactly where value is created and where you’re losing it.
There are many levers you can pull — from increasing average order value to optimizing purchase frequency to reducing acquisition costs. It depends on where your model has the most room.
Key metrics describe how much a customer earns you and how much it costs to acquire them. For example:
- Revenue per customer
- Ratio of customer lifetime value (CLV) to acquisition cost (CAC)
- Conversion rate by segment
5. Scale — Where Next?
The business works, metrics are dialled in. Now you’re thinking about market positioning.
According to Michael Porter, there are three fundamental strategies:
- Segmentation — focusing on a narrow market
- Efficiency — lowest price through optimization
- Differentiation — being unique and hard to replace
At this stage, you need to know whether your chosen strategy is working. Do customers behave the same as early adopters? Does the business work without intensive support?
Key metrics describe ecosystem health and the ability to expand. For example:
- Market share in your target segment
- Margin and its trend over time
- NPS or another customer satisfaction metric
Don’t Skip Stages
I’ve seen a website that poured millions into marketing. But it was unusable — users left faster than they arrived. They skipped Stickiness and went straight to scaling a loss.
I’ve seen an e-shop that started optimizing conversion rates before even launching. Instead of validating whether their MVP actually worked. They ran out of money within a month and the project collapsed.
Every skipped stage is a debt that comes back. With interest.
What Does This Look Like in Practice?
- Figure out what stage you’re at. Honestly. Most companies think they’re further along than they actually are.
- Pick one metric. The most important one for your current stage.
- Display it everywhere. Dashboard, team meeting, weekly report. One metric, one focus.
- Once you’ve nailed it, move on. And pick a new one.
Want to find out what to measure first? Drop me a line — let’s figure it out together.