The short answer. To measure an early-stage startup's progress, a dashboard of about eight figures reviewed every month is enough: cash collected, customers who come back, cost to win a customer, gross margin, net burn, runway, progress toward your milestones and one metric specific to your model. What matters isn't how many numbers you track, but reading trends over several months rather than the mood of the week.
What you'll learn
-
Why measuring progress matters even if you're not raising.
-
The eight-figure dashboard that works for almost any early-stage startup.
-
Which extra metric to add for your model: subscription, marketplace, e-commerce, services, hardware.
-
How to read the numbers without fooling yourself: averages, trends and cohorts.
-
A quarterly review in five questions.
Measuring is the starting point for any serious diagnosis (Startup health check: where you stand and what's missing). This guide gives you a method to do it yourself every month.
Why measure progress if you're not raising?
Because without a round, nobody forces you to look, and nobody bails you out of an expensive mistake. If you're growing on your own revenue, every decision counts double: a lost month is cash that doesn't come back.
Measuring does three things for you, whether you raise or not:
-
Decide on facts: what works, what doesn't and where to put the next euro.
-
Spot problems early: almost everything that sinks a startup shows up in its numbers months earlier.
-
Prove you can execute: a history of well-kept numbers is the best introduction to a bank, a partner, a buyer or an investor the day you need one.
What I've seen Many founders work flat out and still don't see coming what's holding them back. It isn't a lack of effort: the day-to-day of a startup takes up all the attention and leaves no room to look at the whole. Measuring every month, with the same numbers, is the cheapest way to look from the outside.
Which figures make a good dashboard for an early-stage startup?
The eight-figure dashboard
- 1 Cash collected
- 2 Customers who come back
- 3 Cost to win a customer
- 4 Gross margin
- 5 Net burn
- 6 Runway
- 7 Progress toward your milestones
- 8 Your model metric
| Figure | What it tells you | How to calculate it |
|---|---|---|
| 1. Cash collected | Whether the market pays | What you collected in the month, not what you invoiced |
| 2. Customers who come back | Whether customers stay | Customers from three months ago who still buy or use it, out of the total back then |
| 3. Cost to win a customer | Whether growth is sustainable | Marketing and sales spend for the month ÷ new customers that month |
| 4. Gross margin | Whether each sale makes money | (Revenue − direct cost of serving) ÷ revenue |
| 5. Net burn | How much you lose each month | Total spend − cash collected |
| 6. Runway | How much you have left | Available cash ÷ net burn (Runway calculator) |
| 7. Progress toward your milestones | Whether you're heading where you said | Milestones achieved versus those planned this quarter |
| 8. Your model metric | What drives your particular business | Depends on the model (see below) |
Tip Eight well-measured numbers beat forty badly measured ones. If you can't calculate one every month in under an hour, simplify it.
Which metric should you add for your business model?
The eighth figure changes with how you make money. These are the most common:
| Model | The metric that matters most | Why |
|---|---|---|
| Subscription (software, recurring services) | Monthly recurring revenue and churn | Growth only counts if customers don't leave |
| Marketplace | Transaction volume and repeat buyers and sellers | You have to move both sides at once |
| E-commerce | Margin per order after shipping and returns | Selling a lot at a negative margin speeds up the losses |
| Professional services | Team utilization and margin per project | The limit is people's time |
| Hardware | Unit cost and manufacturing lead time | A cost mistake gets multiplied by every unit |
| Health, food or other regulated sectors | Progress on authorizations and certifications | Without them you can't sell, even if there's demand |
A word of caution Unit economics matter more than they seem. In CB Insights' March 2026 analysis of venture-backed startups that shut down since 2023, 19% of those with an identified cause said the numbers per customer didn't work. If each customer costs more than they bring in, growing only speeds up the problem.
How often should you measure?
| Frequency | What to look at | What for |
|---|---|---|
| Weekly | Sales, usage and available cash | Reacting to what's urgent |
| Monthly | The full eight-figure dashboard | Seeing the trend and deciding |
| Quarterly | Milestone review and the five questions | Correcting course |
| Yearly | A full diagnosis, from the outside | Seeing the whole startup |
The week is for acting, not for drawing conclusions. Conclusions come from several months of data.
How do you read the numbers without fooling yourself?
The most common mistake isn't measuring badly: it's reading badly. A founder can swing from euphoria to doubt in a matter of weeks, and one week's numbers feed both.
New customers per week (fictional example)
How to read it. If you'd decided in week 1, you'd have over-hired. If you'd decided in week 3, you'd have over-cut. The average over several weeks and the trend over several months are what should decide.
"A good week isn't a trend. Neither is a bad one."
Three habits for reading well:
- Use three-month averages for revenue, spend and customers. They smooth out the spikes without hiding the trend.
- Look at cohorts: group customers by the month they joined and track how many are left each month afterwards. That shows whether new customers retain better or worse than older ones.
- Compare against your own plan, not just last month. Growing 5% can be good news or a disaster, depending on what you needed.
What does a full monthly dashboard look like? An example
Here's the dashboard for a subscription startup after six months (fictional figures; starting cash of €240,000):
| Figure | Month 1 | Month 2 | Month 3 | Month 4 | Month 5 | Month 6 | Average of last 3 months |
|---|---|---|---|---|---|---|---|
| Cash collected | €8,000 | €9,500 | €9,000 | €11,000 | €12,500 | €12,000 | €11,833 |
| Customers who come back (at 3 months) | — | — | — | 78% | 80% | 74% | 77% |
| Cost to win a customer | €310 | €290 | €330 | €300 | €280 | €360 | €313 |
| Gross margin | 62% | 63% | 61% | 64% | 65% | 64% | 64% |
| Net burn | €24,000 | €23,000 | €24,500 | €22,000 | €21,500 | €23,000 | €22,167 |
| Runway at month end (months) | 9.0 | 8.4 | 6.9 | 6.7 | 5.8 | 4.4 | — |
| Quarter milestones | — | — | 1 of 2 | — | — | 2 of 3 | — |
How to read it. Revenue is growing on average, margin is holding and burn is edging down: the business is moving forward. But there are two warning signs. In month 6, repeat customers and cost per customer both get worse at the same time, and runway dropped below six months in month 5 and now stands at 4.4. The decision this month isn't to celebrate revenue: it's to find out what happened with the customers who joined in month 3, and to start working on cash or the round. (Runway and burn rate)
What should you do with what you measure?
Measuring without deciding is just bookkeeping. The dashboard only earns its keep if every month ends in a decision. Three simple rules:
- A figure that worsens two months in a row needs a written explanation: what happened and what you're going to do.
- A figure that improves three months in a row needs a question: what's driving it, and how can I reinforce it?
- Every important decision cites a figure from the dashboard. If you can't point to the number that justifies a hire, a campaign or a price change, the decision isn't ready.
Tip Share the dashboard with your team every month. The same numbers, seen by everyone, align people better than any goal-setting meeting.
How do you set up a dashboard in an afternoon?
- Choose your eight figures, including the one for your model.
- Write down how each one is calculated, so next month it's calculated the same way.
- Gather the data for the last six months, even if it's approximate: you need a starting line.
- Build a simple spreadsheet with one column per month and one row per figure, and add the three-month average.
- Pick a fixed day each month to update it, say the 5th.
- Write three lines underneath every month: what improved, what got worse and what you're deciding.
Common mistake Changing the definition of a figure when you don't like the result. If you change how you count returning customers, redo the previous months too, or you lose the comparison.
Which questions should you ask yourself every quarter?
The quarterly review in five questions
A guide, not a rulebook. You can print it.
If the answer to the first question is "no" two quarters running, it isn't a bad patch: it's a signal that needs a proper look. (Startup health check)
Which signals tell you something is wrong?
| Signal | What it may mean | What to look at |
|---|---|---|
| Customers are growing but revenue isn't | Low prices or discounts | Average revenue per customer |
| Revenue is growing but cash is falling | You're collecting late, or margin is low | Payment terms and margin |
| Cost per customer is rising | The channel is saturating or getting pricier | Cost by channel |
| Fewer customers are coming back | Customers aren't sticking | Cohorts and reasons for leaving |
| Runway is falling faster than planned | Actual spend is above plan | Spend by line versus the plan (Runway and burn rate) |
| Milestones slip every quarter | The plan is unrealistic, or the team is stretched | Team capacity and priorities |
How does measuring progress help if you do raise one day?
An early-stage investor looks at the trend over several months, not your best month. If you've been tracking the same figures every month for a year, you'll arrive at your round with something few founders have: a consistent history that shows you can execute and that you know your business.
Your monthly figures will also be the basis for your data room and for your answers to investors' questions. (Is your startup ready to raise?)
Do it with a method Measuring yourself every month shows how you're evolving. An outside evaluation shows where you stand against what's expected at your stage. The Startup Index Score evaluation gives you your level from 1 to 5 and your first three priorities, free, in 30–40 minutes; and you can measure yourself again to check whether you've moved forward. (Start the evaluation)
What are the common mistakes when measuring progress?
| Mistake | Why it misleads | What to do instead |
|---|---|---|
| Tracking vanity metrics | Visits and followers don't pay the bills | Measure cash collected and repeat customers |
| Using invoiced instead of collected | Cash comes in when you're paid | Revenue collected |
| Deciding on the last week | A week isn't a trend | Three-month averages |
| Changing the definitions | You lose the comparison | Write them down and keep them |
| Measuring without deciding | Numbers are useless if they don't change anything | Three lines of decisions every month |
| Not measuring the team | The ability to execute can be measured too | Milestones hit and deadlines met |
In short
-
Measuring progress matters whether you raise or not: to decide, spot problems early and prove you can execute.
-
An eight-figure dashboard is enough, with one metric specific to your model.
-
Measure collected, not invoiced, and use three-month averages.
-
Read trends and cohorts, not single weeks.
-
Review every quarter with five questions.
-
A consistent history is your best introduction to an investor, a bank or a buyer.
Do it with a method Find out where your startup stands today, and measure yourself again to see whether you're moving forward. Evaluate your startup with Startup Index Score: free, in 30–40 minutes. (Start the evaluation)
Frequently asked questions
Which metrics should an early-stage startup track? About eight is enough: cash collected, customers who come back, cost to win a customer, gross margin, net burn, runway, progress toward milestones and one metric specific to the business model. What matters is measuring them the same way every month and reading the trend over several months.
How often should I review my startup's metrics? The urgent ones weekly; the full dashboard monthly; milestones and direction quarterly; and a full outside diagnosis at least once a year. Important decisions should rest on several months of data, not on a single week. A good week can hide a bad trend, and the other way around.
How do I know if my startup is doing well? If most of your key figures are improving on the three-month average, you're hitting the milestones you set and your runway covers your plan, you're on track. If a figure gets worse two quarters in a row, that's not a bad patch: it's a signal worth investigating properly.
What is a customer cohort? It's a group of customers who started in the same period, for example the same month. Tracking how many from each cohort keep buying or using the product over time shows whether you're retaining better or worse than before, something the total customer count hides.
Is it worth measuring progress if I'm not going to raise? Yes, even more so. Without a round, every mistake costs your own cash and nobody forces you to look. Measuring every month helps you decide where to put the next euro and leaves you a track record that will help with a bank, a partner or a buyer.
Keep reading
Sources
-
CB Insights, "The top reasons startups fail," March 2026: www.cbinsights.com/research/report/startup-failure-reasons-top/
-
The rest of this guide reflects common practice in managing early-stage startups and Jordi Gil's experience. The examples with figures are fictional.
About the author
Jordi Gil is the creator of Startup Index Score. He has spent almost 30 years developing projects and working with startups across the Americas, Europe and Asia. His view: a new business needs results soon, revenue with margin, more than theory. (Meet Jordi Gil)