Once a week, look at every product on the same growth chain, find the one whose weakest step moved the most, read why, and decide one change. Split it into three five-minute blocks: scan all products, open one, decide and write it down. The point is not to see everything; it is to leave with one decision you can check next week.
Why weekly
Daily numbers for a small product are mostly noise: a launch post, a bot spike or a quiet Sunday swings them more than anything you changed. Monthly is too slow to notice that a sign-up page broke three weeks ago. A week is long enough to smooth the noise and short enough to act on.
Compare the last full week with the week before, both complete. A week still in progress always looks like a decline.
Before the first review: give every product the same shape
The review only takes fifteen minutes if every product reads the same way. Put each one on one growth chain: Reach (impressions, mentions), Visitors, Engaged (active users, sessions), Converted (sign-ups, trials, orders) and Revenue. Then give each product one objective, the number that means progress for it right now: sign-ups for a new app, trial-to-paid for a subscription product, orders for a shop.
Without this, you compare a newsletter's open rate with a shop's revenue and learn nothing.
Minutes 0 to 5: scan every product
Read across, not down. For each product you want one answer: did its objective move, and which step moved it?
| Look at | The question | Ignore when |
|---|---|---|
| The objective | Up, flat or down against last week? | The change is a handful of people |
| The weakest step | Which step got clearly worse? | It moved within its usual range |
| Revenue | Did collected revenue change, across all payment accounts? | Only part of the accounts reported |
| Guardrails | Did speed, errors or refunds get worse? | Nothing crossed a threshold |
Pick one product to open. Usually it is the one whose objective moved most, in either direction: a jump you do not understand is as worth reading as a drop.
Minutes 5 to 10: open one product
Go one level down, to the step that moved. If visitors dropped, which source and which landing page? If sign-ups dropped while visitors held, which step of sign-up? If revenue moved, new customers or churn?
Stop at the first explanation the evidence supports, and write down what you saw, not what you assume. "Organic search visits to the pricing page fell 40%" is a finding; "Google hates us" is a guess.
Minutes 10 to 15: decide one change and write it down
Choose one thing to change this week and record four facts: the product, the change, the date it goes live, and which number should move by next Monday. Next week's review starts by reading that note.
One change at a time is not caution; it is the only way to know what worked. Two changes in the same week on the same product leave you guessing which one moved the number.
Four traps in multi-product numbers
- Averaging growth rates. "Our products grew 12% on average" mixes a product with 50 visitors and one with 50,000. Read each product on its own, or add up absolute numbers.
- Counting shared accounts twice. If one Stripe account serves two products, its revenue must be split, not shown under both.
- Adding different people. Active users from one tool and sign-ups from another are not the same population and cannot be summed.
- Reacting to every red arrow. A small product moves 30% on three sign-ups. Look at the absolute numbers before the percentage.
How the review looks in TractionScout
The portfolio of six sample products in the TractionScout demo workspace, with simulated live visits and payments.
TractionScout's portfolio puts every product on the same growth chain, so the scan is one screen. Each project card leads with its own objective and names its weakest link; guardrails stay quiet until speed, errors or refunds clearly get worse. Revenue comes from Stripe, Paddle, RevenueCat, Gumroad, WooCommerce and Wix, with each account counted once and a shared Stripe account split per project. Scout's picks say what drives the portfolio now and where attention should go next, and a live feed shows sign-ups and payments as they arrive.
Click any cell in the matrix and the product's Performance page opens on that step, which is the five-minute dig. See Portfolio and Smart KPIs.
Questions
How many products can one person review this way?
The scan stays at five minutes for up to about ten products if they share one growth chain. Beyond that, sort by attention and read the top five each week.
Which number should I compare across products?
Compare the direction of each product's own objective and the step that moved, not raw numbers. When you do compare trends side by side, index them to the same starting point so a small product is not flattened by a large one.
What if a product has no revenue yet?
Give it the objective that comes before revenue, such as activated sign-ups, and leave the Revenue step empty rather than showing zero. An empty step says "not measured yet"; a zero says "measured, and nothing happened".
See every product you run on one growth chain
Traffic, sign-ups, trials and revenue for each product, with the weakest step named.