Startup Metrics Dashboard: The Numbers to Review Weekly
A startup metrics dashboard should answer one question every Monday: are we moving in the right direction, and if not, where is it stuck? This guide shows you which numbers deserve a weekly look, how to lay them out on one page, and how to run a short review that turns numbers into decisions instead of anxiety.
Key takeaways
- Track 8 to 12 numbers weekly, grouped into acquisition, activation, revenue, retention and cash.
- Pair every number with an owner, a target and a four-week trend so you see direction, not just a snapshot.
- Pick one "north star" number that best reflects the value customers get, and put it at the top.
- Run a 30-minute weekly review with a fixed agenda and end it with no more than three actions.
Most early-stage founders have the opposite problem from the one they expect. It is not that they lack data. It is that every tool they use, from the payment processor to the website analytics to the ad account, offers dozens of charts, and none of them agree on what matters. You check ten tabs, feel worried, and change nothing.
A weekly dashboard fixes that by forcing choices. You decide in advance which numbers you will look at, what "good" looks like for each, and who is responsible for moving them. Then you look at the same page, at the same time, every week. After a month or two, patterns jump out that were invisible when you checked numbers at random.
Why weekly is the right rhythm for an early startup
Daily numbers are noisy. A quiet Tuesday or a busy Friday tells you very little, and reacting to every swing burns attention. Monthly numbers are steadier, but at an early stage a month is a long time to wait before noticing that sign-ups have stalled or a key customer has gone quiet.
A weekly rhythm sits in the middle. It is frequent enough that you can still change course within a quarter, and slow enough that you see real movement rather than random noise. It also matches how most small teams already plan their work, so the dashboard can feed straight into the week's priorities.
Weekly reviews also connect well with goals. Many startups set OKRs, Objectives and Key Results, on a quarterly cycle. A weekly dashboard is how you check, thirteen or so times a quarter, whether the key results are on track before the quarter is already over.
The numbers worth reviewing every week
There is no universal list, because a subscription app, a marketplace and a hardware brand all work differently. Still, most early startups can build a strong weekly view from five groups. Pick one to three numbers from each group, and stop at about twelve in total.
1. Acquisition: are new people finding you?
- New visitors or leads. How many new people reached your site, app store page or sales inbox this week.
- Sign-ups or trials started. How many took the first concrete step.
- Top source of sign-ups. Which channel brought the most, so you know where to double down.
2. Activation: do new users reach the first moment of value?
Define one clear "activated" action, the point where a new user first gets real value. For a scheduling tool it might be booking a first meeting. For a design app it might be exporting a first file. Then track the share of new sign-ups who reach it within a set number of days. This is often the most useful early number of all, because it tells you whether your product promise holds up in practice.
3. Revenue: is value turning into money?
- New revenue booked this week. New customers, upgrades and one-off sales.
- Recurring revenue run rate. If you sell subscriptions, your current monthly recurring revenue, updated weekly.
- Pipeline value. For sales-led businesses, the value of qualified deals in progress.
4. Retention: do people stay and come back?
- Active users. Weekly active users or accounts, defined by a meaningful action, not just a login.
- Cancellations or churned accounts. Count them and, just as important, note the reasons.
- Repeat purchase or usage rate. The share of last month's customers who came back.
5. Cash: how much runway do you have?
Keep this group short: cash in the bank at week end and the number of months that cash would last at your current spending. You probably already model this in more depth elsewhere, so the weekly dashboard only needs the headline figure and its trend.
Pro tip
Choose one north star number and put it in the top left of the dashboard. It should reflect value delivered to customers, such as "accounts that completed a weekly report" or "orders delivered on time," rather than something you can inflate with spending, like page views.
How to build your dashboard step by step
You do not need special software. A shared spreadsheet works well for the first year, and it has one big advantage: you understand every formula in it. Follow these steps.
- Write down your current goal. One sentence about what matters most this quarter, for example "prove that small agencies will pay for the product." Every number you add should relate to this goal.
- List candidate metrics. Brainstorm everything you could track, then cut ruthlessly to 8 to 12 using the five groups above.
- Define each metric precisely. Write how it is calculated, where the data comes from and what time window it covers. "Active user" means nothing until you define the action.
- Assign an owner. One person per number. The owner updates it and explains movement during the review. In a two-person team, split them evenly.
- Set a target or a range. Even a rough target is better than none. It turns a number into a question: are we above or below where we expected?
- Add a four-week trend. Show the last four weeks next to this week. Direction matters more than any single value.
- Set a fixed update time. For example, numbers are filled in by Monday at 10 a.m. and the review happens at 11 a.m.
Common mistake
Changing metric definitions every few weeks. If "active user" meant one thing in January and another in March, your trend line becomes meaningless. When you must change a definition, note the date in the sheet and, if possible, recalculate the old weeks.
A weekly metrics dashboard template
Copy this layout into a spreadsheet. One row per metric, one column per week. The numbers below are made up for illustration only.
| Group | Metric (definition) | Owner | Target (example) | 4 weeks ago | Last week | This week | Status |
|---|---|---|---|---|---|---|---|
| Activation | New sign-ups who created a first project within 7 days | Product lead | 40% | 28% | 33% | 36% | Improving |
| Revenue | Monthly recurring revenue at week end | CEO | $12,000 | $9,400 | $10,100 | $10,250 | Behind |
| Retention | Accounts cancelled this week (with reasons logged) | Support lead | 2 or fewer | 1 | 3 | 4 | Needs review |
Add a final column called "Notes" for one line of context, such as "pricing page changed Tuesday" or "two cancellations cite missing integration." Those notes become gold when you look back at the quarter.
How to run a 30-minute weekly review
A dashboard only helps if someone looks at it and acts. Keep the review short, regular and structured so it never turns into a long debate. Here is an agenda that works for teams of two to fifteen people.
- North star first (3 minutes). Read the top number aloud and say whether it went up, down or stayed flat.
- Walk the reds and yellows (12 minutes). Skip anything on target. For each metric that is behind, the owner gives a one-sentence explanation and a likely cause.
- Look for links (5 minutes). Ask whether movements are connected. A drop in activation and a rise in cancellations may share a cause.
- Agree on actions (7 minutes). Choose no more than three actions for the week, each with an owner and a due date.
- Note the learning (3 minutes). Write one sentence in the notes column about what you learned this week.
Protect the time. If the meeting regularly runs over, you are tracking too many numbers or debating solutions that need their own session.
Hypothetical example
Imagine Priya and Tom, co-founders of a small invoicing app for freelance designers. Their February review shows sign-ups steady but activation stuck at around 30%. Using the made-up figures in their sheet, cancellations also ticked up for two weeks, and the support notes mention people struggling to import client lists. They agree on one action: Tom builds a simple import guide and a sample file by Friday. Three weeks later, activation reaches 37% and cancellations fall back. The dashboard did not solve the problem, but it pointed straight at it.
Getting the data in without wasting hours
The biggest threat to a weekly dashboard is the time it takes to update. If filling it in takes half a day, it will quietly stop happening. Aim for 20 to 30 minutes of updating per week, split across owners.
- Start manual, then automate. Typing in numbers by hand for the first few weeks helps you understand them. Once the definitions settle, connect the sources that are easy to export or link.
- Pull from the source of truth. Revenue should come from your billing or accounting system, not from memory. Users should come from your product database or analytics, using the same definition every week.
- Keep a data dictionary tab. One row per metric, with the definition, source and owner. New team members will thank you.
Be careful with customer data as you connect tools. Only pull what you need, keep the dashboard access limited to the team, and follow the privacy rules that apply where you and your customers are based. In the UK and EU, GDPR applies to personal data, so aggregated counts are usually safer than lists of named users.
Vanity metrics to leave off the weekly page
Total registered users ever, social followers and page views can keep rising while the business stalls. Glance at them monthly if you like, but keep them off the weekly page.
Adjusting the dashboard as you grow
Your dashboard should change when your goal changes, not every week. A good moment to review the list itself is at the start of each quarter, when you set new goals. Mid-February is a natural checkpoint: the first quarter is half over, and you can see whether the numbers you chose in January are actually guiding decisions.
Signs it is time to update the list include a metric that has been green for months and no longer needs weekly attention, a new product line or sales channel that is not represented, or a team that keeps asking for a number that is not there. When you do update, archive the old version rather than deleting it, so you can still compare.
The weekly page also makes investor updates easier, since the trend lines are already there.
A quick note: this guide is general information about running a business, not financial, legal or investment advice. For questions about financial reporting or how to present numbers to investors, talk to a qualified accountant or adviser.
Your weekly dashboard checklist
- Write one sentence describing this quarter's main goal
- Choose one north star number that reflects customer value
- Pick 8 to 12 metrics across acquisition, activation, revenue, retention and cash
- Write a precise definition and data source for each metric
- Assign one owner and a target or range to every number
- Show this week plus the previous four weeks for each metric
- Set a fixed update deadline and a 30-minute review slot
- End each review with no more than three actions and owners
- Review the metric list itself once per quarter
Frequently asked questions
How many metrics should a startup dashboard have?
For a weekly review, 8 to 12 is a practical range. Fewer than that and you may miss an important signal. More than that and the review becomes a reading exercise rather than a discussion about what to do next.
Do I need dashboard software or is a spreadsheet enough?
A spreadsheet is enough for most early startups. It is flexible, cheap and transparent. Consider a dedicated tool later, when updating by hand takes too long or many people need live access. Whatever you choose, check current features, pricing and privacy policies first.
What if a number looks bad several weeks in a row?
Treat it as a signal to investigate rather than a reason to panic. Talk to a few affected customers, look at what changed around the time the trend began, and test one fix at a time so you can tell what worked.
Should solo founders bother with a weekly review?
Yes, and it may matter even more. Without a team to challenge you, a fixed weekly look at the same numbers keeps you honest. Block 30 minutes in your calendar and write your three actions down.
Related reading
The bottom line
A startup metrics dashboard is not about collecting more data. It is about choosing a small set of numbers, defining them clearly, giving each an owner and looking at them together every week. Once that rhythm is in place, problems surface earlier and wins are easier to repeat.
Open a blank spreadsheet this week, list your quarter's goal and pick your first eight numbers. Fill in four weeks of history if you can, then hold your first 30-minute review next Monday.
