A store dashboard is not a report. It is a pile of true facts arranged by what was easy to display, and it will happily let you spend an hour on a chart that cannot change a decision.
So we read it in a fixed order, every time, looking for one thing: which of the four numbers behind revenue moved, and therefore which room to go into. Sessions, conversion rate, average order value and returning customer rate. Everything else on the screen is either an input to one of those or a distraction.
The order matters because the numbers multiply. Revenue can fall while every individual chart looks acceptable, if two of the four slipped slightly at once. Reading them as a sequence catches that, where reading the revenue headline never will.
What follows is the sequence, what each step is actually telling you, and the traps in the default views that make stores draw the wrong conclusion.
Start with sessions, split by where they came from
The total session count is the least useful number on the page, because it blends sources that behave nothing alike. Read it split.
Paid sessions should track your spend. If spend held and paid sessions fell, something changed in the auction or in your feed, and that is where to look next.
Organic search sessions move slowly, so a sharp change is usually technical rather than editorial. A sudden drop is worth checking against your product page count and any recent theme or app change.
Direct and email sessions are your existing audience. A rising share here is a good sign for the business even when the total is flat, because those visitors cost you nothing.
Social sessions browse differently, so a month where social grew as a share of the total will drag your blended conversion rate down while nothing has actually got worse. This is the single most common false alarm we get asked about, and the split is what prevents it.
Compare against the same month last year as well as against last month, because retail has a shape and your dashboard does not know it.
A dashboard is a pile of true facts arranged by what was easy to display.
Then the conversion funnel, one step at a time
The conversion rate summary in most dashboards shows the funnel as three or four stages, and the value is entirely in which stage moved, not in the final figure.
Sessions to product views tells you whether people are finding products. If this is the weak step, the problem is in navigation, collection pages, internal search or the landing pages your ads use, and no product page work will fix it.
Product views to add-to-cart is the product page's own score. This is where unanswered questions show up, and it is the step we look at first for most stores.
Add-to-cart to checkout started is where shipping cost, delivery timing and unpleasant surprises live.
Checkout started to order placed is payment, form length, forced accounts and trust.
Read each step for mobile and desktop separately. A blended funnel can show a healthy average while the device most of your traffic uses is failing at a specific step, and that failure is invisible until you split it.
Average order value needs its context
Average order value on its own is easy to misread, because it moves for at least three different reasons and the dashboard does not say which.
It rises when people buy more items, which is the outcome you want. It also rises when your product mix shifts toward expensive items, which is not the same thing and may reverse next month. And it rises when a discount ends, which is not a result at all.
So read it alongside items per order. If both moved together, behaviour changed. If the average moved and items per order did not, price or mix moved.
Read it by channel too. Traffic sources bring different sized baskets, so a change in the traffic mix moves the store average without anyone changing anything about the store.
And where your reporting allows it, look at the distribution rather than the average. A handful of large orders will pull the mean somewhere that describes nobody. The value most of your customers actually land on is the number that should be setting your free shipping line and your bundle prices.
Returning customer rate is the slow number
The fourth number changes slowly and is worth more than it looks, because a customer who buys again costs nothing to acquire.
Read it as a share of orders and as a share of customers, since the two answer different questions. Share of orders tells you how much of this month's revenue came from people you already had. Share of customers tells you whether the base is compounding.
The trap here is that the rate moves when acquisition moves, in the opposite direction. A strong month of new customer acquisition pushes the returning share down while the business gets healthier. Read it next to new customer count or you will draw exactly the wrong conclusion.
Also watch time between first and second order, because it sets the timing for every message you send afterwards. If most second orders arrive around week six, a replenishment email in week nine is late.
And give it time. This number responds to changes over quarters, not weeks, which is why we report it as a trend line rather than as a monthly figure with an arrow next to it.
The traps in the default views
Four habits prevent most bad decisions made from this screen.
Mind the date comparison. Comparing this month with last month compares a different number of weekends, a different holiday position and often a different promotional calendar. Same period last year, plus the previous period, gives you a sense of both.
Do not trust a percentage change on a small base. A move from four orders to six is not a fifty per cent improvement in anything, and dashboards will present it as though it is.
Expect the analytics figure and the finance figure to differ. Orders, refunds, taxes, shipping and cancelled orders are treated differently in different views, and neither number is wrong. Decide which one you make decisions from and use it consistently.
And when a platform report disagrees with your store's own order data, believe the store. Ad platforms and email tools each claim conversions on their own terms, and the sum of their claims usually exceeds what you actually sold.
Then write down what you concluded, in one line per number. A month later that line is worth more than the dashboard, because it tells you what you expected to happen and whether it did.