Start by defining the unit you are counting
Before discussing sales statuses, decide whether each row represents a person, a company, an enquiry or a deal under negotiation. These units can be related without being equivalent. Identify them separately and document how they enter the report, so several contacts at one company do not automatically become several customers.
In a B2B service business, two people may participate in buying the same project. The same person may also complete the form twice. If you add all those entries as different potential customers, you increase apparent volume without necessarily having more purchase decisions to follow up.
Preserve the history of interactions, but avoid confusing it with the number of entities. A new call adds activity to the process; it does not create a new company. An enquiry about another service may justify a separate opportunity, keeping its connection to the customer and the contacts you already know.
I would ask the team to explain an actual row in the report before interpreting the total. Who is this entity, what did it request and which negotiation does it belong to? If the answer changes according to who presents the numbers, the definition needs resolving before comparing performance.
To prepare your next step, explore Lead generation for service businesses.
Define when a contact becomes a lead
A lead should correspond to an identifiable entry into your sales process under a shared definition. Explain whether you are counting contact requests, registrations with commercial interest or prospecting contacts. Keep the origin visible and distinguish an observed interaction from a need that the person has not yet confirmed.
A newsletter registration shows interest in receiving content; it does not demonstrate an intention to buy a service by itself. If the business treats that registration as a lead, it should explain what kind of lead it is. This avoids comparing content registrations with proposal requests as though they meant the same thing.
A contact found through company research is not equivalent to someone who requested help either. It can enter a prospecting list with its origin and context, without receiving confirmed opportunity status. The classification should explain what you know, not anticipate what you hope will happen.
When building a funnel the sales team can work with, define these entry points from the beginning. Record when contacts arrived and who follows them up. If contacts are duplicated, invalid or outside the service scope, show those cases separately instead of silently removing them from the interpretation of results.
Use concrete evidence to open an opportunity
An opportunity should represent a possible deal whose need and conditions for progressing are sufficiently clear to justify its own follow-up. Define criteria appropriate to your service and record the next step. A booked meeting or a high score can help qualification, but does not automatically confirm all those elements.
For example, you might require that the problem is understood, there is a plausible fit with the offer and a step has been agreed to assess the solution. This is an illustrative operational definition, not a universal rule. What matters is being able to verify the criteria and apply them consistently across people and periods.
Sending a proposal does not guarantee that the opportunity is well qualified either. If nobody understood the need and the document was sent simply to answer a generic price request, the status may hide a gap. Record the activity performed without using it as a substitute for the information required.
Assign an owner to each opportunity and keep its status current. If the person postpones the decision or no longer wants to proceed, record what is known and the date. An opportunity that has been open for a long time with no next step should not continue supporting a sales forecast as though it were about to close.
Choose a clear confirmation for counting a customer
Count a new customer when the commercial event defined for recognising the first purchase has occurred, with evidence and a date. That event may depend on your business model. Keep agreement, invoicing, payment and delivery separate, because they can happen at different times and do not answer the same management question.
In a hypothetical service example, the sales report might recognise a sale after formal acceptance of the proposal. The cash report only recognises the money when it arrives. This operational choice does not replace accounting criteria; it helps the team understand what each number represents.
An additional purchase by an existing customer increases sales, but not necessarily new customer numbers. A company returning after a period without purchases may require a reactivation category. Define that treatment before comparing acquisition, retention and changes in your customer portfolio.
You also need a visible rule for cancellations, refunds and deal changes. Preserve the original event and the later adjustment without invisibly rewriting the past. The aim is to reconcile reports and explain differences, not force every total to appear identical regardless of what actually happened.
Calculate conversions using comparable groups and costs
A conversion rate is only interpretable when you know the starting cases, the outcome counted and how much time they had to reach it. Use the same entry group when tracking transitions between statuses. For costs, identify the expenses included too, avoiding the label total acquisition cost for a calculation containing advertising alone.
In an entirely hypothetical example, a campaign costs 600 euros and generates 100 unique leads. Of those same leads, 20 become opportunities and five produce new customers within the chosen follow-up window. Advertising cost per lead is 6 euros; per opportunity, 30 euros; and per new customer, 120 euros.
Lead-to-opportunity conversion is 20%, opportunity-to-customer conversion is 25%, and lead-to-customer conversion is 5%. The three percentages use different denominators and describe different transitions. Do not add these rates together or use cost per lead as though it were cost per customer.
If you combine this month's sales with this month's leads without confirming their relationship, you may be mixing groups. Some sales may come from older entries, while some recent leads have not had enough time to decide. This distinction helps you choose indicators showing whether the business is improving without attributing a change to the funnel that the measurement does not demonstrate.
Make definitions part of everyday work
Definitions need to be available to the people entering records and those reading the report. Document criteria, dates, owners and common exceptions in a short reference. Review real cases with the team and identify differences in interpretation before automating transitions that could multiply one mistake throughout the entire contact database.
An automatic transition can be useful when the event is verifiable, such as receiving valid confirmation. If the system merely infers interest from an interaction, keep that inference labelled. Automation should respect the chosen meaning of the status, with clear conditions for correcting an inaccurate record.
When asking how much you can pay for a lead, the answer depends on that lead's definition and observed conversion into a customer. If you change the form or qualification criteria, flag the change. Comparing series with different definitions without that context can lead to wrong investment decisions.
I would start by following one complete case from entry to purchase, with evidence at every transition. Then I would check whether the aggregate report counts that journey consistently. When lead, opportunity and customer have clear meanings, the team can discuss where to act without losing the conversation trying to discover what the numbers mean.