Your sales data already knows the answer — 6 questions nobody asks
Once a month you open the sales report, look at the total, and close it. Higher than last month means a good month. Lower means "the market's quiet" and you move on. That habit can run for years, because nothing visibly breaks: the site works, orders come in, money lands in the account.
But that same file — with no new column added, just the existing rows grouped a different way — already holds the answer to at least six questions nobody formally asks. They go unasked because the report comes in "monthly total" shape, not "who, what, when" shape.
The six questions below fall into three groups: who buys, what they buy, when they buy. None needs an analytics tool — an ordinary spreadsheet of your order history is enough, once you know what to ask it.
Who buys: the customer side
1. Where does your most expensive customer come from?
Group every customer by the channel that brought them in — paid ads, referral, Instagram, a direct call. For each group, work out the average order value and how often that customer comes back. The picture that comes out is usually uneven: one channel brings a lot of customers who buy once and vanish; another brings fewer people who stay for years and buy again. If most of the budget goes to the first channel while the second gets no attention, money is flowing the wrong way.
You don't need complicated math to see it — add up total customer spend per channel and set it next to the ad spend that channel cost you. When the two numbers converge, that channel isn't making you money, it's just moving turnover around.
2. How many customers is half your revenue riding on?
Sort customers by total spend over the last 12 months, highest first, and add the revenue up as you go down the list. Which customer do you reach when you hit half the total — the 5th, or the 50th? A small number is a hidden risk: if one of that handful leaves for a competitor, half the revenue can disappear in a day. The large customer isn't the problem — not knowing how much you depend on them is.
To check this fast, delete the top 10 customers from the list and look at what's left. If the number drops sharply, that's the real count of customers your business is standing on — and the relationship with them can't be left to chance.
What they buy: the product side
3. Which product sells another one?
Look at which product pairs show up together in the same order almost every time. If one item is nearly always bought alongside a specific other item, that's not chance — seeing the first reminds the customer of the second. Once you know the pair, placing them next to each other, mentioning it at checkout, or bundling them as an offer lifts the average order without spending anything extra on ads.
Finding the pair doesn't need a clever algorithm — scanning the last 100 orders by eye usually surfaces the pattern on its own, because buying behaviour repeats.
4. Which way is your average order size moving?
Compare months on average order value — not total revenue, but revenue divided by number of orders. If total revenue is rising while the average order keeps falling, a rising customer count is masking the opposite trend: each individual buyer is spending less over time. The usual cause is a discount habit — a salesperson gets used to shaving a little off every order, it builds up over months, and nobody notices because the total still looks fine.
To separate the two effects, keep discounts in their own column instead of folding them into the sale price — otherwise the discount disappears from view entirely.
When they buy: the timing side
5. Which month "dies" — and why?
Line up sales from the last two or three years, month against month. If the same month comes out weak every single year, that isn't chance, it's a season — and once you know it's coming, you can plan a specific campaign, offer, or product launch for it. The real problem is treating that weak month as a fresh surprise every time, when last year's numbers already showed it coming.
Knowing the season matters beyond marketing — it feeds inventory, staff scheduling and cash-flow planning. A business that sees the weak month coming meets it by trimming costs, not by being caught off guard.
6. Does a customer buy once and leave, or come back?
For every customer, work out the gap between their first order and their most recent one. If the vast majority show up exactly once, the business is running on a permanent hunt for new customers — the most expensive and most tiring growth model there is. Even if only a small share comes back, understanding why they stay — delivery speed, service quality, or a personal relationship — is the key to keeping the rest too.
To see the return rate clearly, split customers into groups by the month of their first purchase, then check what share of each group bought again after 3, 6 and 12 months. That's the simple method known as cohort analysis, and it can be built by hand in an ordinary spreadsheet.
How to actually ask these questions
None of the six above needs new software. Three steps cover it:
- Export your order history — with date, customer, channel, product and amount as columns. If there's no channel column, add one by hand for at least the last three months; it's a one-time job.
- Build a separate sort or pivot table for each question: by channel, by customer, by month, by product pair.
- Write the result down and keep it. The answer only takes one pass to find, but it needs rechecking every quarter — channel behaviour, seasonality and customer loyalty all shift over time.
The hard part isn't the arithmetic, it's writing the question down in the first place, because usually nobody has.
The total at the end of the month tells you what happened; the same rows sorted differently tell you why.
You can work through all six yourself. If you don't have the time, or can't tell which column shows what, see how we find the answers to these six questions inside your own data — delivered as a one-page report with the reasoning spelled out.
