Sales analysis is a breakdown of what drives changes in the number of purchases, revenue and the profitability of deals. To find losses, compare equivalent periods, split sales by source and product, calculate the contribution of each change and identify the causes in individual deals. The outcome of the analysis is a decision: what work to fix now, where to put money and what result to hold the head of sales responsible for.
Enquiry numbers are the same, but purchases have fallen noticeably. The first suggestion at the meeting is to raise salespeople’s bonuses: let them sell more actively. Marketing has done its part; the flow has stayed the same. The owner has a clear explanation and an expense already lined up.
But the same number of enquiries tells you nothing about who made them. If yesterday partners brought in half the buyers, while today advertising generated most of the enquiries, the sales team now has a different task. The previous overall conversion rate can no longer be the only measure of salespeople’s work.
I believe that before changing bonuses, the head of sales must separate the effect of the mix of enquiries from the effect of how they are handled. Otherwise, the owner pays the team to fix the part of the shortfall it has no control over. Let’s work through the numbers and then turn the calculation into a task for this week.
Sales analysis: the reports to start with
You need three sets of data for the first analysis. That does not necessarily mean three new files: the relevant fields in your existing records are enough.
- Sales and money. Paid orders, their amounts, products, discounts and refunds. This shows what changed: the number of purchases, the value of a purchase or the mix of what was sold.
- Enquiries and their outcomes. The enquiry date, source, salesperson, whether a purchase was made and when. This shows which enquiries led to a result.
- Costs associated with orders. Product or service costs, delivery, variable commission paid to the salesperson and the cost of attracting customers. You need this set when deciding which sales are worth increasing.
The head of sales gathers information on enquiries and the team’s work, and checks the amounts and costs with whoever keeps the financial records. The owner only needs to set the question for the analysis: for example, whether it is worth increasing the advertising budget with the current flow of enquiries. Without a question, analysis can easily grow into a study of the whole business.
First, make the comparison fair
Choose periods of the same length and define the event you count as a sale. In the calculation below, this is an order paid in full. Partial payments, refunds and sales of different products need separate rules; you cannot mix them with purchases paid in full.
Calculate conversion using the same set of enquiries: how many came in and how many of those exact enquiries led to a purchase. The groups you compare must have the same observation period. For example, the outcome of each enquiry on the thirtieth day after it arrived. This is a condition of the calculation, not a standard for how long a deal should take.
Keep seasonality separate too. If there was a sale in one period and normal prices applied in the other, the difference in conversion includes a change in the offer. The head of sales must note such a circumstance alongside the calculation so the owner does not mistake the effect of a promotion for a new skill in the team.
Of the 32 missing purchases, 20 have a different cause
The following is an illustrative example, not a RENTROP case study. All the numbers and actions in it have been defined for this exercise. The company sells one service for 50 thousand roubles, every order is paid in full in one payment, and there are no refunds. We compare two groups of 400 enquiries received over two periods of equal length; for each group, we count purchases made within the first 30 days of an enquiry.
In the first group:
- Partners brought in 200 enquiries. 60 customers paid: a conversion rate of 30%.
- Advertising brought in 200 enquiries. 20 customers paid: a conversion rate of 10%.
- A total of 80 purchases worth 4 million roubles. Overall conversion was 20%.
In the second group:
- 100 enquiries came from partners. 30 customers paid: the same 30%.
- 300 enquiries came from advertising. 18 customers paid: a conversion rate of 6%.
- A total of 48 purchases worth 2.4 million roubles. Overall conversion was 12%.
The shortfall compared with the first group is 32 purchases, or 1.6 million roubles. The flow stayed the same, but two changes combined: the share of partner enquiries fell, and advertising enquiries were less likely to lead to payment.
Now let’s keep the new mix of enquiries and apply the previous conversion rate for each source:
- 100 partner enquiries × 30% = 30 purchases.
- 300 advertising enquiries × 10% = 30 purchases.
- A total of 60 purchases instead of the previous 80.
Even with the previous conversion rate within both sources, there would have been 20 fewer purchases. That is a difference of 1 million roubles due to the mix of enquiries. The remaining 12 purchases, or 600 thousand roubles, are the difference between the calculated 60 and the actual 48, linked to the fall in conversion of advertising enquiries.
1.6 million = 1 million due to the mix of enquiries + 600 thousand due to conversion within the source. The calculation follows a set order here: first we change the mix of enquiries, then conversion. The two contributions together explain the entire shortfall.
Asking salespeople to recover all 1.6 million by doing the same things as before means ignoring the first million. A bonus for salespeople cannot restore the previous share of partner enquiries. But the 600 thousand needs a separate analysis: what happened to the advertising enquiries within the department and before they reached it.
This 600 thousand is a starting point for finding the cause. The company did not issue invoices for this amount, and the buyers do not owe it. The calculation shows the difference from the previous result, but does not by itself establish who caused it or how much money can be brought in.
Business sales analysis: count what remains from an order
Now the owner has to choose: invest in partners, continue advertising or first change how enquiries are handled. A high conversion rate does not by itself make a source the best investment. There are costs of attracting customers, a limit on volume and costs of fulfilling orders.
Let’s add to the same illustrative example. Of the service price of 50 thousand roubles, 20 thousand remains after the service is delivered and the salesperson’s variable commission is paid. The cost of attracting customers and the company’s fixed costs are not yet included.
For the actual 48 purchases, the amount left is 960 thousand roubles: 48 × 20 thousand. For the calculated 60, it is 1.2 million. The difference is 240 thousand. This amount is before the costs of fixing the problem, not a promise of profit from taking action.
This shows the scale of the decision. If the proposal is to spend 300 thousand just to secure these 12 additional purchases, their contribution in the selected group would not cover that cost. If the spending benefits future sales, it needs a separate calculation with a stated time frame. You cannot hide such a bet on the future inside a promise to recover the current 600 thousand.
When there are several products, the head of sales calculates the number of orders and the actual price for each. If the overall average order value has fallen because the share of a cheaper service has grown, the team has not necessarily started giving way on price more often. You can see discounting by comparing identical offers, and changes in the product mix by looking at each product’s share.
Figures that look attractive have another limitation: a source can bring in several buyers who depend on a single project. This is a separate issue, discussed in the article “Which figures should you rely on? Many customers, but one construction project”. It does not replace conversion in our calculation, but it affects the owner’s decision on how much money to put into that source.
What lies behind the 600 thousand
After the calculation, the head of sales needs an answer on why conversion fell from 10% to 6%. Starting with a demand for more frequent calls means once again choosing an action before analysing the problem.
The head of sales compares advertising enquiries from the two periods: what people wanted to buy, what terms the advertising offered, and who replied and how. If the promised scope of the service changed, the salesperson faced a new objection. If the terms stayed the same but the number of late replies increased, there is specific work to do within the department.
Suppose that, in the illustrative analysis, the head of sales took 20 enquiries from the second group that did not result in a purchase and reconstructed their history:
- In eight, salespeople sent the promised quote after the agreed deadline. The dates are visible in the correspondence.
- In seven, buyers had given a later decision date in advance, and it had not yet arrived at the cut-off date.
- In five, buyers said they had chosen another provider; the reason for their choice had not been established.
Eight delays give grounds for fixing quote preparation. Seven future decisions require work on the agreed dates. For the five rejections, the reason needs to be established. You cannot lump everything together as “not pushing hard enough”: more insistent reminders will not make a late proposal timely.
The sample of twenty does not explain the entire 600 thousand. It is the start of an analysis that has found an error the head of sales is already obliged to fix. To assess how widespread it is, the head of sales counts delays across the entire group, then compares them with the old group. The owner does not personally open hundreds of records.
The article “Sales funnel: stages, an example and where money gets stuck” shows in detail where to look for a hold-up between an enquiry and payment. Here we have a different question: which part deserves analysis first, and what amount can reasonably be linked to it.
This week’s task: two causes, two decisions
The owner makes a limited decision: leave the whole team’s bonuses unchanged for now, and allocate no additional advertising budget until the analysis is complete. The department fulfils its existing agreements with customers. You can set the head of sales’ task using this ready-made wording:
By Wednesday, 16:00, break down the shortfall into the effects of the mix of enquiries and conversion within each source. For advertising enquiries, show the number of late quotes in both groups and where the delay occurred: with the salesperson, in preparing the proposal or in agreeing the terms. With the head of marketing, establish why partner enquiries fell. Bring one proposal for allocating the budget, with a calculation of the costs and expected contribution from orders. State separately which causes remain unknown.
The next part of the task concerns carrying out the work:
By Friday, 16:00, show the proposals sent for the eight delays found, or changes agreed with the buyer. For all new quotes due on or before Friday, show how many were promised, how many were sent on time, where a deadline was missed and what you did. If preparing a proposal is held up by another department head, bring me the specific dispute over resources and the deadline.
When checking the work, the owner looks at three things: whether the arithmetic explaining the shortfall adds up, whether the point of delay has been established, and whether promises on new enquiries are being kept. The number of payments over these few days is no substitute for completing the task: in the example, outcomes are compared on the thirtieth day.
What counts as a poor result is also defined here. If new quotes still go out after the agreed deadlines and the head of sales has merely reminded the team again, the way the team works has not changed. If deadlines are met, monitoring of purchases in this group begins. This way, the head of sales is responsible for their action, while the owner decides whether it is worth paying for the work to continue.
When finding the source of losses takes manual work
If the link between the source, salesperson and money has to be reconstructed from different data exports every time, regular analysis becomes an extra burden. RENTROP has AI ROP for this task — sales team analytics: figures, the sales funnel, salespeople and money by source.
Once a discrepancy has been found, a different kind of detail is needed: what happened in an individual sale. This is the focus of the review of deals, calls and correspondence with a next step in AI ROP. Choosing an action and making sure it is carried out is the head of sales’ job.
The proposal to raise bonuses now has a simple test: the head of sales can say which action by a salesperson the bonus will change and which part of the shortfall it will affect. Without that answer, a new bonus is an expense with a known amount and an unknown purpose.