The difference between a useful metric and a flattering one
Sort store data by whether it changes a decision, rather than whether it makes progress look impressive.
You open your dashboard on Monday and see that visits are up, your follower count has grown and one post reached more people than usual. It looks like progress. Then you check sales and find that nothing has changed. The awkward question is not which number is higher. It is whether any of those numbers should alter what you do this week.
Start with the decision, not the dashboard
Creators often collect numbers in the order a platform presents them. Visits come first, followed by views, likes, followers, clicks and sales. That order is convenient for software, but it is not necessarily useful for running your shop. Start at the other end: name the decision you are trying to make, then find the smallest number that can inform it.
| Decision | More useful metric | What it can tell you | What it cannot tell you |
|---|---|---|---|
| Whether to improve a product page | Product-page visits and purchases for that product | Whether the page is turning interested visitors into buyers | Why a visitor decided not to buy |
| Whether to repeat an email campaign | Email clicks and sales attributed to the campaign | Whether the message sent people towards an offer | Whether the same result will happen with a different audience |
| Whether to promote a product again | Sales and refunds for that product over a defined period | Whether demand produced worthwhile orders | Whether the next promotion will reach the same people |
| Whether a new traffic source deserves time | Visits, purchases and revenue from that source | Whether attention became commercial activity | How much unpaid effort the source required |
A number is useful only when it changes an action
Take a metric and complete this sentence: “If this number is higher than expected, I will…” Then complete the opposite: “If it is lower than expected, I will…” If both endings are the same, the metric is probably reporting rather than guiding. You may still enjoy seeing it, but it should not take priority in your weekly review.
- 1Write the decision in plain language: “Should I send another email to this segment?”
- 2Choose one primary measure and, if needed, one diagnostic measure. For the email example, sales could be primary and clicks could help explain the result.
- 3Set a comparison before looking at the result. This could be the previous campaign, a different product or a simple target you chose in advance.
- 4Record the action beside the number. “Keep the subject line, change the offer” is more valuable than “click rate: 3.4%”.
- 5Review the action later. A metric becomes more useful when you check whether the change produced the intended result.
Separate attention from buying intent
Attention metrics describe what people noticed or did before a purchase. They can help you find a question, but they rarely answer the commercial question on their own. A post with many views may have reached people who were curious but not ready to buy. A product with fewer visits may produce more orders because its visitors have a clearer need.
Use ratios carefully
Ratios can make unlike-sized activities easier to compare. For example, conversion rate is purchases divided by visits, multiplied by 100. If a page receives 200 visits and makes 6 sales, its conversion rate is 3%. If another receives 20 visits and makes 2 sales, its rate is 10%. The second rate is higher, but it is based on fewer visits and two purchases. Both facts matter.
The same problem appears with average order value, revenue per visitor and email click rates. A ratio removes some of the effect of audience size, but it does not remove uncertainty or explain cause. Pair it with the underlying counts and the time period. In your store analytics, compare like with like: the same product, traffic source or campaign type where possible.
Choose a small weekly set
A one-person business does not need to inspect every available report every week. Choose a small set that covers the path from attention to order. One measure can describe qualified visits, one can describe buying behaviour and one can describe the result that pays you. Add a fourth only when it explains a decision you cannot make from the first three.
For example, you might review visits to a particular product, purchases from those visits and revenue after refunds or other relevant adjustments. If you are testing email marketing, add clicks and campaign sales for that email rather than replacing your whole dashboard with newsletter statistics. If the question is customer experience, orders needing support or refund requests may be more useful than another acquisition number.
Make the review practical
Give each number a job and a boundary. “Sales this month” is broad. “Sales of the new template in the first seven days after launch” is easier to interpret. “Followers” is broad. “Clicks from the profile link after changing the product description” is closer to a decision. The more specific the question, the less tempting it is to fill the review with numbers that merely look active.
A storefront can help by keeping products, payments and customer activity closer together. With store.fan, money goes straight to your own Stripe account and store.fan takes 0% of sales, while Stripe's own processing fee still applies. If you use the iOS app, instant sale alerts can tell you that an order happened; the alert is useful for awareness, but the decision still needs the product and period around it.
No. They can show that a message travelled or that a topic attracted attention. Treat them as supporting evidence. Give them more weight only when you have a clear mechanism linking them to the decision you are making, such as a click through to a product page.
Use a rhythm that matches the decision. Weekly reviews suit ongoing activity, while a launch or email campaign may need a review after the campaign ends. Checking more often does not create more information; it can simply make normal fluctuations feel like instructions.
Keep the disagreement visible. More visits with fewer purchases may mean the new audience is less ready to buy, the offer is unclear or the traffic count includes a different mix. Segment the result by product, source or campaign, then choose one explanation to test rather than changing everything at once.
Set up a storefront where your product and sales information are easier to review.
Try store.fanTurn your knowledge into income
Launch your Store.Fan in minutes — sell digital products, courses, and calls straight from your bio. Free to start.
