The problem isn’t finding metrics, it’s cutting them
If you run a one-person business, you probably have access to more numbers than you know what to do with. Your payment processor has a dashboard. Your email tool has open rates and click rates. Your analytics platform has traffic, bounce rate, session duration, and a dozen other things it calculates automatically. None of that is hard to find.
The hard part is deciding what goes on the one page you actually look at every week.
I’ve tried building weekly reports that tried to cover everything. They didn’t get read. Not because I was lazy, but because a report with twenty numbers on it doesn’t tell you anything. It just sits there. A report with five or six numbers, chosen on purpose, tells you what happened and usually points at what to do next. That’s the whole difference.
Start from the decision, not the data
The question I ask before adding anything to a weekly report isn’t “can I track this?” It’s “what decision would change if this number moved?”
If a metric going up or down this week wouldn’t change what you do next week, it doesn’t belong on the weekly report. It might still be worth tracking somewhere, monthly or quarterly, but a weekly report is for numbers you’re prepared to act on weekly.
A concrete example: say you run a small subscription service. Your weekly report might include new trials, trial-to-paid conversion rate, and churned customers. Those three numbers, together, tell you whether you have an acquisition problem, a conversion problem, or a retention problem, and each of those points at a different fix. Something like average session length on your app, on the other hand, rarely changes what you do this week even if it moves. It’s interesting, but it’s not decision-driving on a weekly cadence.
The metrics that tend to earn a spot
I’m not going to hand you a universal list, because the right metrics depend on your business model. But there’s a pattern in what actually survives the cut for most solo operators.
Revenue-adjacent numbers. If you run a subscription business, MRR (monthly recurring revenue) and its week-over-week change usually make the cut, because it’s the closest thing to a single number that summarizes the health of the business. Churn, the percentage of customers who cancel in a given period, sits right next to it, because MRR going up while churn quietly climbs is a warning sign that a top-line number alone will hide.
Something at the top of the funnel. New leads, new signups, new trials, whatever the entry point is for your business. This tells you whether the pipeline feeding everything else is still flowing.
Something at the point of conversion. Trial-to-paid rate, quote-to-close rate, cart-to-purchase rate. This is where you find out if the problem is that not enough people are showing up, or that the people showing up aren’t converting. Those are different problems with different fixes, and a report that only shows top-of-funnel numbers can’t tell them apart.
One quality-of-customer number, if you have it. Customer lifetime value (CLV) is the estimate of how much revenue a customer generates before they leave. It’s built on assumptions, average purchase value, purchase frequency, average customer lifespan, so treat it as a directional estimate, not a precise fact. Still, watching whether it’s trending up or down over time tells you whether the customers you’re acquiring are getting more or less valuable, which a raw revenue number won’t show you on its own.
That’s usually four to six numbers. That’s a report.
What doesn’t belong, even if it’s interesting
A lot of numbers are genuinely interesting and still don’t belong on a weekly report.
Vanity metrics that don’t connect to revenue or retention are the most common offender. Social media followers, total pageviews, email list size without a conversion number attached to it. These can go up steadily while the business underneath them is flat or shrinking. I’m not saying they’re worthless, they can matter for brand-building over a longer horizon, but they don’t belong competing for space on a page meant to drive this week’s decisions.
Metrics with too much noise at a weekly resolution are another problem. Some numbers only mean something over a longer window. A single week of traffic data can swing wildly because of a holiday, a single viral post, or a payment processor hiccup, and reacting to that noise as if it were signal will send you chasing ghosts. If a number needs four or eight weeks of data before a trend is visible, it belongs on a monthly report, not a weekly one.
Metrics you can’t act on this week also don’t belong here. If you’re not going to look at a number and consider a specific next step, it’s not weekly report material even if it’s technically important. Put it on a quarterly review instead.
A quick gut check before you add anything
When I’m deciding whether a metric earns a spot, I run it through three questions:
Would this number moving change what I do next week? If not, it’s not weekly material.
Do I understand what’s actually driving this number? A metric you can’t explain is a metric you can’t act on. If MRR drops and you don’t know whether that’s churn, discounts, or a data problem, you need to look one level deeper before the number is useful, not just watch it move.
Is this number stable enough at a weekly resolution to mean something? If it’s mostly noise until you zoom out, it belongs on a longer-cycle report.
If a metric passes all three, it’s a strong candidate. If it fails even one, I leave it off, or I move it to a less frequent report where it makes more sense.
Format matters almost as much as selection
Once you’ve picked your five or six numbers, resist the urge to also list this week’s number, last week’s number, the percentage change, a four-week average, and a chart, for every single metric. Pick one or two ways to show change, consistently, across every metric on the page. I like showing the current week’s number next to the prior week’s number, with a simple up or down arrow. It’s enough to spot a trend without turning the page into a spreadsheet.
Also resist adding commentary for every line. If a number didn’t move meaningfully, you don’t need to write anything about it. Save the notes for the two or three things that actually changed. A report where every line has a paragraph of explanation is a report nobody reads twice.
The report is a habit, not a dashboard
The point of a one-page weekly report isn’t to replace your dashboards or your accounting software. It’s a habit, a five-minute checkpoint you do every week that forces you to notice trends before they become problems. A dashboard with fifty metrics doesn’t build that habit. A page with five numbers you understand does.
If you’re not sure where to start, pick the three numbers that would worry you most if they moved in the wrong direction, and build from there. You can always add a metric later once you’ve proven you’ll actually look at it.
Want more on this? I break down metrics like MRR, churn, and CLV, plus how to build automated tracking for them, over on the Data Research Analysis Collection YouTube channel and site.
Get new guides and videos first — join the Telegram channel.