Suppose that, out of 30 proposals, 12 have led to payment: sales team analytics separates the remaining 18 into refusals, agreed waiting periods and overdue actions by the team. Without this distinction, the business owner risks paying for negotiation training when buyers simply never received a call back. A single conversion rate hides the difference between a customer who decided not to buy and a salesperson who stopped working on the deal.
In a report like this, I am interested above all in where the refusal came from. Did the buyer choose another supplier? Postpone the purchase? Or did the salesperson send a quote, set a reminder, move it several times and close the record? All three events can easily end up with the same label in the report. But what you will need to spend money on afterwards differs.
So the business owner asks for more detailed analytics. The wish is understandable: they want to finally see where enquiries are going. But the new chart will draw its reasons from the same records. Until the head of sales has worked out what staff record as a refusal, a polished report will use the salespeople’s own decisions to give a convincing explanation of buyer behaviour.
Sales funnel analysis
Sales funnel analysis shows how one group of customers moves between stages and what happens to those who stop. Each transition needs three results: how many moved on, how many declined and how many are still in the sales process; for the last group, it also needs to show who is holding up the next action.
Let us continue the calculation. Suppose 100 new enquiries came in during October. By the end of the month, the team had spoken to 60 buyers about what they needed, 30 had received a proposal and 12 had paid. All the figures refer to the same hundred enquiries, not to different customers the team worked with in October. Here, every buyer who paid had a conversation and received a proposal.
Conversion is calculated by dividing the number who reached the next stage by the number who reached the previous one:
- From enquiry to conversation: 60 / 100 × 100% = 60%.
- From conversation to proposal: 30 / 60 × 100% = 50%.
- From proposal to payment: 12 / 30 × 100% = 40%.
Of all the enquiries, 12% led to payment. If the owner is shown only the final 40%, they get an answer about work on proposals that have already been prepared. That figure says nothing at all about what happened to the other enquiries. The measure only makes sense in relation to the people included in it.
Now let us break down the 18 proposals that have not led to payment. Suppose eight are marked as declined in the CRM system. Of the ten open deals, in four the buyer has confirmed a date for making a decision, while in six the team has missed the promised reply deadline. This gives us a short breakdown you can save as a template:
- Proposals: 30.
- Paid: 12.
- Recorded as declined in CRM: 8.
- Buyer confirmed a future decision date: 4.
- Team missed its action deadline: 6.
- Total outcomes: 12 + 8 + 4 + 6 = 30.
The grounds for the eight recorded refusals need checking. Work on the four deals with agreed dates will continue at the agreed time. The six overdue actions require the head of sales to intervene now. Conversion from proposal to payment stays the same — 40%. The work behind it is completely different.
Look: if you remove the six forgotten deals from the calculation, you get 12 / 24 × 100% = 50%. No one else has bought anything. The team has simply stopped counting part of its work. So excluding an enquiry from the report is a decision with consequences for the budget: the next advertising calculation will be based on a percentage improved in this way.
Who exactly is waiting for a reply?
An entry saying “waiting for the customer” is useful for managing work when it shows what the buyer needs to do and what has been agreed with them. A reminder date that a salesperson set for themselves says nothing about the buyer’s timescale.
If a customer has asked for a new configuration and the salesperson has passed the question to a technical specialist, the buyer is waiting for a quote from our company. Until the reply arrives, they have nothing to approve. Yet the record remains at the proposal approval stage. If the head of sales attributes the delay to the customer taking a long time to choose, they will start rushing the wrong person.
The owner can usefully ask one question of the report here: how many buyers with open deals are waiting for a reply from us? The answer changes the discussion about additional advertising. While the team is not sending the quotes already requested, a new stream of enquiries increases the amount of unfinished work. The head of sales needs to resolve the delay with the salesperson or reach an agreement with a colleague in another department, then check that the buyer has received the reply.
A buyer declining and the team stopping work are different events
A customer has declined when they have communicated their decision. The team has stopped work when it has decided not to spend any more time on the deal. Both are acceptable, but they have different reasons, and the report must preserve that distinction.
If a buyer has failed to respond several times, the head of sales is entitled to stop trying to contact them. Calling endlessly costs money too. But there is no reason to record that customer as having declined because of the price: they never discussed it. An entry saying contact attempts have stopped explains why the team stopped and leaves the reason for the buyer’s silence unknown.
A reasonable objection comes up here: does the head of sales have to investigate every closed record? Manually compiling all the reasons and assessing the grounds for closure involve different amounts of work. The software puts the summary together. The head of sales examines the way deals are repeatedly closed, especially when there are plans to change sales terms or spending on that basis.
For example, of the eight refusals in our group, five are marked “too expensive”. Before lowering the price for all future buyers, the head of sales finds out what happened in those five conversations. If the customer compared offers for the same configuration and chose a cheaper proposal, there is a basis for working on price. If the salesperson sent the amount and never got back in touch, the company has no answer about price. It has an unfinished conversation.
I am against a discount based on this kind of report. It will reduce revenue from the next sale while preserving the habit of leaving the buyer without further contact after sending a quote. First, the head of sales gets the salesperson back into the conversation and finds out exactly what is holding up the decision. The result of this work shows in the buyer’s response: they state a condition for buying, agree to continue or decline and give a reason.
The old entries are kept. If the review leads to more confirmed refusals and fewer open deals, the data is being handled better, even though the report looks worse. Closing stalled records does not create new sales in itself. It gives a clear view of how much work the team will continue and why it stopped working on the rest.
Forty payments in October do not mean forty buyers out of a hundred
Conversion for October enquiries is calculated from purchases by those particular customers. Money received in October is counted by the date of payment. Mixing these two reports attributes money from earlier enquiries to new advertising.
Let us return to our hundred. Suppose a total of 40 buyers paid in October: 12 from new enquiries and 28 from previous months. Dividing 40 by 100 gives 40%, but the new hundred produced 12 buyers. If 40 purchases are built into the calculation for the next advertising spend, the company will expect a result this stream of enquiries has not yet shown.
The money from old enquiries is real. It just has a different history.
To compare sources, give groups the same amount of time to buy: for example, 60 days from each customer’s first enquiry. A comparison of the first 30 days is kept separate from the result for 60. If purchase approval takes longer in your business, extend the observation period; deals that are still open remain visible. This way, a new source does not lose to an old one simply because its buyers are still discussing terms.
The history of stage changes matters more here than the current stage. Moving a record back does not create a new customer, a second payment on an order does not create a second sale, and handing a deal to a colleague does not double the team’s money. Permanent customer and deal numbers allow these events to be joined into one chain. The accounting system confirms money received; CRM and conversations with the buyer confirm actions and agreements.
Expected payments are kept separate from payments received. The grounds for expecting money are examined in the article “Sales work is under way. But which figure can you trust?”. Here, we are establishing what has already happened to the stream of enquiries we paid for and what work the team has not yet finished.
Search brought in more money but left less
A source with a higher sales total should not always get a larger budget. To make this decision, compare the money from one group of buyers with the costs of attracting them and fulfilling their orders.
Suppose our hundred enquiries came from search and the 12 buyers paid 600 thousand roubles. All paid orders have been fulfilled, and there are no returns. Fulfilment cost 360 thousand roubles, and attracting the hundred enquiries cost 120 thousand roubles. After these costs, 600 − 360 − 120 = 120 thousand roubles remained.
Now take partner referrals with the same time from enquiry to the report date: 40 enquiries, 10 buyers and 500 thousand roubles in payments. These paid orders have also been fulfilled, and there are no returns. Fulfilment cost 300 thousand roubles, and attracting the enquiries cost 50 thousand roubles. Money left: 500 − 300 − 50 = 150 thousand roubles. Salaries, rent and the company’s other general expenses are paid out of these remaining sums.
Search brought in 100 thousand roubles more but left 30 thousand roubles less. Its cost per new buyer is 120 thousand / 12 = 10 thousand roubles; for referrals, it is 50 thousand / 10 = 5 thousand roubles. The owner has grounds to examine the stream of partner referrals. Before increasing spending, they still need to find out how many extra referrals the partner can provide on the same terms.
But search still has six overdue actions by the team. The two final sums do not yet show how much of the difference came from the source and how much from the work done with it. That is a reason to first look into how the work was handled. When the team replies to the buyers, it will become clear which deals are continuing and who really declined. Once the chosen period has ended, the money comparison can be updated.
Related purchases also need attention. Ten buyers who depend on one shared project do not provide ten independent sources of money. This risk is examined separately: “Which figures can you rely on? There are many customers, but one construction project”.
Who is responsible for what the figure means?
The analyst is responsible for making sure the records add up and the measure is calculated consistently. The head of sales is responsible for the grounds on which the team labelled a customer as lost, a deal as pending and work as complete. If the owner has to establish this part every time, the person preparing the report has been given a task, while the head of sales has yet to do theirs.
To start, one area and one completed period are enough. The person preparing the report links enquiries, deals and payments, and shows returns and unknown sources separately. The head of sales examines the reasons for refusals and delays that affect the spending in question. The owner gets an explanation of the discrepancy and a proposed action. They do not have to reread records personally to understand why the company is about to pay for attracting customers again.
When these rules are in place, manually putting summaries and conversations together becomes a separate demand on time. At RENTROP, AI ROP helps with this task: sales team analytics shows performance figures, the funnel, salespeople and money by source, while reviewing calls and messages alongside the record helps show what happened with the buyer. The service reads the data; the head of sales decides whether to continue the work, change it or close the deal.
I do not require the head of sales to get every buyer to agree. But a record of a refusal must mean that the buyer declined. If it conceals our decision to stop working on the deal, that is exactly what the owner needs to see — before paying to replace the lost customer.