In this essay 13 sections
A sales conversation should leave the customer better able to make a decision. They should understand what is being offered, whether it suits them, what remains uncertain and what would happen next. If the conversation produces enthusiasm without understanding, the work is unfinished.
I take sales seriously. A business needs paying customers, and somebody has to ask for the order. But asking for a decision is different from making it difficult for somebody to say no. The first makes the conversation clear. The second can conceal a weak offer behind a strong personality.
Williams Corporation’s June 2021 history book, pages 14–16, records three practical choices behind that view. In 2018 I began building an in-house sales team: Stephen Pike joined in April and Matt Withington in August. In 2019 we were running property seminars in Auckland, Wellington and Christchurch. That year, Blair and I also planned our Auckland office, with Stephen moving north as Sales Manager. These are historical company accounts, not evidence of a particular conversion improvement.
Those choices put responsibility for explaining our homes inside the business and closer to the people considering them. That is the connection to this essay: sales needs people who understand the offer, a way to explain it and someone accountable for the next answer. Hiring a team or holding an event is not proof of good service. The quality still has to show up in the individual conversation.
Start with the decision, not the presentation
It is tempting to begin with everything you know about the product. You have spent time developing it. You understand its features. You know which details distinguish it from alternatives. That knowledge matters, but it does not tell you which part matters to the person in front of you. The businesses I am involved in offer different products and services; a useful explanation has to start with the particular decision, not a presentation that treats every enquiry alike.
Someone enquiring about a home might be comparing locations, trying to understand the buying process or deciding whether a particular layout fits their life. The enquiry does not establish that they are ready to purchase. It establishes that they have a question. Answering the wrong question in great detail is still poor service.
Begin by finding out what they are trying to decide. What prompted the enquiry? What are they comparing? Which requirements are fixed, and which are preferences? What information would let them make progress? These questions give the explanation a purpose. They also allow the customer to correct the seller’s assumptions before those assumptions take over the conversation.
There is no need to turn this into an interrogation. Ask for information relevant to the decision, explain why it is useful and let the person set appropriate limits. A buyer can tell you that timing matters without giving you the private details behind it. A business should be capable of understanding a requirement without collecting everything it can about a person.
Then repeat the problem back in ordinary language. You are looking for a particular result, these constraints matter, and this is the question you have not yet resolved. If that account is wrong, correct it. If it is right, the product discussion can become much shorter and more useful. You know what the explanation needs to accomplish.
This does not mean a customer must arrive with a perfectly formed brief. Sometimes the conversation helps them discover what matters. Seeing a layout, comparing two options or learning how a process works can change the question. The important discipline is to notice the change rather than continuing through a presentation built for the original assumption.
Revenue is feedback, not a complete explanation
I have written about sales as a test of whether people value what a business has made. That is a reason to pay attention to the market, not a reason to pretend the first result explains everything. A weak response tells you to investigate. It does not identify the cause on its own.
Perhaps the offer does not solve an important problem. Perhaps it reaches the wrong people. Perhaps the explanation is unclear, the price is out of reach or the timing is unsuitable. Those are different hypotheses. Each suggests a different action. Cutting the price before understanding the problem can leave the original weakness in place while reducing the money available to deliver the product.
A sale needs interpretation too. A customer might choose a product for a reason the team barely mentioned. They might accept a limitation because another feature matters much more. If the business records only that the deal was won, it loses information about the value it actually created.
The answer is to separate observations from explanations. A request for a different completion date is an observation, if that is what the customer said. Describing the customer as not serious is an interpretation. The first gives the team something specific to investigate. The second may simply protect the salesperson from having to ask another question.
There will also be unknowns. Some people do not reply. Some prefer not to explain their decision. Record that honestly. A database filled with invented reasons may look more complete than one containing unknowns, but it is a worse foundation for changing the business.
Listening to the market does not require changing direction after every conversation. One person’s preference is not a market-wide instruction. Look for patterns, check the context and consider the cost of responding. The discipline is to let evidence challenge the offer without allowing every isolated comment to redesign it.
What helpful service can look like
Williams Corporation publishes customer testimonials that give useful examples of what people notice in a sales process. In one, Anton describes Stephen listening, allowing him time and contacting him when a preferred option became available. Brad describes being able to consider his options without feeling pressured. These are individual accounts selected for the company’s testimonial page, not an independent survey or proof that every customer has the same experience.
What matters to this argument is the kind of help being described. The seller is present. Information arrives when it is relevant. The buyer retains room to decide. That is not the absence of selling. It is selling that makes the customer’s task easier.
Pressure and urgency are not the same thing. A real deadline or a genuine change in availability may matter to a decision. Explain it accurately, including what is known and what is not. Do not manufacture a deadline because the conversation is taking longer than you would like. A customer needs the facts about timing, not a performance designed to make reflection uncomfortable.
There is also a difference between respecting someone’s pace and disappearing. If a customer asks for a document, send the document. If you agree to check something, return with the answer or explain why it remains unresolved. Giving someone space should not become an excuse for leaving them to chase work you accepted.
The relevant standard is not a particular number of follow-up calls. It is whether the contact has a purpose the customer can recognise. An answer to their question, a requested comparison or an update they asked to receive can be useful. Repeating the same demand for a decision after a clear refusal is not the same service.
A worked comparison: two homes, different answers
Consider a fictional buyer comparing two homes. This is an illustration, not a Williams Corporation customer account. One home has an extra room but a longer journey to the buyer’s usual activities. The other has less space but is closer to the places they visit most often. Both could be good homes. That does not mean they are equally useful to this buyer.
The weak sales conversation begins with whichever property the salesperson most wants to sell. The additional room becomes essential when discussing the larger home. Location becomes everything when discussing the smaller one. The buyer hears two confident pitches without receiving a consistent way to compare them.
A better conversation establishes the decision criteria first. How often will the extra room be used? What needs to fit inside it? Which journeys happen regularly? What does the buyer want to inspect before deciding? The salesperson does not have to choose the buyer’s priorities. They have to help connect those priorities to the actual properties.
Suppose the buyer says the extra room is intended for occasional visitors, while the daily journey matters throughout the week. That is useful information, not a verdict. They may still value hosting enough to choose the larger home. The comparison becomes clearer because both parties understand what the additional space would do and what is being traded for it.
Now suppose there is an unresolved question about whether a desk and storage will fit comfortably in the smaller home. An enthusiastic assurance is not the best next step. Check the relevant dimensions, identify any limits in the information and give the buyer a practical way to assess the space. If a visit is possible, that may help. If it is a planned home, distinguish drawings and representations from something they have inspected.
The seller should also be willing to reach a conclusion that does not produce an immediate sale. Neither option may meet an essential requirement. That does not make the conversation worthless. It can prevent a poor match, clarify what the buyer should look for and reveal a gap in the available offer. Those are useful outcomes, even though they are not revenue.
The illustration is not a rule that location always outweighs space, or that smaller homes are automatically better value. It shows why a feature needs to be connected to a use. A feature list says what is there. A good comparison explains what the difference means for the person making the decision.
Ask for the decision without making it for them
I have also argued that business owners need to give sales serious attention and ask for the order. There is no contradiction between that position and a low-pressure conversation. A clear question gives the customer an opportunity to choose. Avoiding the question can leave both parties guessing.
The question should match the stage reached. If the buyer still needs essential information, ask what must be resolved next. If the explanation is complete and they have had the opportunity to assess it, ask whether they want to proceed. If they say no, accept that answer. Where they are willing to explain, listen without turning every explanation into another attempt to wear them down.
Be precise about what proceeding means. Requesting information, arranging a viewing, making an offer and entering an agreement are not interchangeable steps. A customer should not discover afterwards that the seller interpreted an expression of interest as a commitment. The language should make the next action understandable before it is taken.
The buying process described on Williams Corporation’s website separates the initial discussion from later agreement, due diligence, deposit and settlement stages. That distinction is useful here: progress through a process is not one undifferentiated yes. The terms of an actual property transaction belong in its agreement and the buyer’s professional advice, not in a general sales essay.
Finish the conversation with an agreed next step, or with a clear ending. Who is doing what? What information is needed? Has a follow-up been requested? If nothing further is wanted, record that too. A sales process should make it possible to close a conversation honestly, rather than keeping every name in an optimistic category indefinitely.
Measure the journey you actually have
In another public post about sales measurement, I pointed to enquiries, response times, objections and lost deals. Those are useful starting points because they describe different parts of the work. They should not be collapsed into one number that makes the pipeline look busy.
Define the stages before comparing them. An enquiry may come from a new person or an existing contact asking another question. A conversation may establish genuine interest or reveal that the offer is unsuitable. An offer may have conditions. A completed transaction is different again. The definitions need to reflect the actual business, not whichever categories happen to be supplied by its software.
Keep people, opportunities and transactions distinct. One person can enquire more than once. One opportunity can involve several decision-makers. One customer can make more than one purchase. Without that distinction, a team can report an apparent improvement that is simply a change in how it counts the same activity.
Then decide what each measure is supposed to help you do. A response-time measure can reveal an unattended queue. A record of unresolved questions can show where the offer needs a clearer explanation. A stage with unusually old opportunities can prompt a review of whether those opportunities remain active. Each measure should lead to a question somebody can answer.
A report should make uncertainty visible rather than tidy it away. If the next step is unconfirmed, say so. If the reason for a lost deal is unknown, keep it unknown. If the data excludes a particular channel or period, show the boundary. A smaller set of reliable observations is more useful than an impressive total built from incompatible definitions.
Follow one group all the way through
Here is a hypothetical example. These are invented numbers to explain a measurement problem, not Williams Corporation results. A business receives enquiries from 100 different people during April. For this illustration, each person represents one possible purchase, and repeat messages have already been combined. The business follows that same group until the end of June.
By the review date, the team has held a two-way conversation with 80 people. Of those 80, 40 have moved to a defined comparison stage. Of those 40, 16 have made an offer. Of those 16, eight have completed a purchase. Assume that each completed purchase passed through these stages, and that the definitions stayed unchanged throughout the example.
Eight purchases from the original 100 people gives an observed completion rate of 8% at that date. Eight from the 16 who made offers gives an offer-to-completion rate of 50%. Both calculations are correct. They answer different questions. Presenting the second number as though half of all enquirers purchased would be wrong.
The intermediate figures matter for the same reason. Two-way conversations reached 80% of the original group. Half of those conversations reached the comparison stage. Forty per cent of that comparison group made offers. Those proportions describe movement between defined stages; they do not explain why people moved, stopped or remained undecided.
Now imagine the business also completed 12 purchases during April, all from people who first enquired before April. Dividing those 12 purchases by April’s 100 new enquirers produces 12%, but it is not the conversion rate of the April enquiry group. The numerator and denominator refer to different groups. The business can report both monthly volumes. It cannot turn their convenient proximity in a spreadsheet into a customer journey.
There is another limit. At the end of June, suppose 10 of the original 100 opportunities remain active, 82 have been closed without a purchase, and eight have completed. That accounts for all 100. It does not mean the final completion rate is known. Some of the 10 active opportunities may complete later; others may close without buying. The report needs its review date, not just its percentage.
The stage totals and the current status totals are different views of the same group. Forty people reached comparison at some point. That does not mean 40 are still comparing on the review date. Confusing a milestone reached with a present status can make a pipeline appear much larger than it is.
None of these numbers establishes that a salesperson did good or bad work. Perhaps the initial enquiries included many people the product could never suit. Perhaps the offer stage was defined too loosely. Perhaps a genuine timing constraint affected several buyers. The report identifies where to investigate. Conversation records, the actual offer and the customer’s explanation are needed to interpret it.
Comparison over time needs the same care. An April group observed for three months is not directly comparable with a July group observed for two weeks. Nor is a group attracted by a broad campaign necessarily comparable with people who specifically requested an available product. Keep the observation window and the source of demand visible before celebrating a change in the percentage.
Speed matters when there is an answer behind it
Response time is useful, but it is easy to define badly. An automatic acknowledgement shows that a message reached the system. It does not show that somebody understood the question. A fast message asking the customer to repeat information already supplied may improve a dashboard while making the experience worse.
Separate acknowledgement from a useful first response. That response might answer the question, identify the person responsible for answering it or explain what needs checking and when an update can be expected. It need not pretend that a complicated question has an immediate answer. It should show that the enquiry has entered a real piece of work.
Be clear about the clock. Does the measure count evenings and weekends? Does it begin when a form is submitted or when an enquiry reaches a particular team? If a message sits in the wrong inbox before being forwarded, excluding that delay may make an internal measure look cleaner. It does not remove the time the customer experienced.
I would rather see the distribution of waiting times than a reassuring average on its own. A small number of people waiting much longer can disappear inside an average dominated by quick replies. Look at the oldest unanswered enquiries as well. Those records tell you where someone may still be waiting, which is more actionable than discussing a number in isolation.
Do not solve this by promising a response standard the business cannot sustain. Establish what can actually be delivered, explain any relevant limits and make missed commitments visible. If a specialist answer takes longer, the person waiting should receive an honest update. Silence does not become better service because the final answer is technically excellent.
A lost deal needs evidence, not a convenient label
Price is an attractive explanation because it sounds concrete. Sometimes it is the reason. But there is a difference between a customer saying an option is beyond their budget, choosing a lower-priced alternative and declining without explanation. Recording all three as price loses the distinction between evidence and assumption.
Take another fictional record. A buyer says they will not proceed because the available timing does not work for them. The staff member suspects that price also mattered. A useful note keeps those statements separate: timing was the reason given; price is an unconfirmed interpretation. The next person reading it should not have to guess which part came from the customer.
Keep an unknown category. If the customer does not answer a reasonable request for feedback, the business has an unanswered question. It does not have permission to supply the answer itself. Requiring every closed opportunity to contain a confident reason can encourage staff to create certainty for the software rather than learn anything about the market.
Reasons can also overlap. A customer may accept a longer journey at one price but not another. A timing problem may disappear if a different option becomes available. A single category helps with counting, but a short factual explanation often carries the information needed for a useful decision. Do not confuse a reporting convenience with the whole account.
The next question is what the business could sensibly change. Some mismatches point towards a clearer advertisement, better product information or a different offer. Others are outside the available range. If someone needs something the business does not sell, describing that enquiry as a sales failure may encourage the wrong response. Honest disqualification can be evidence that the team understood the requirement.
Do not turn feedback collection into a second sales campaign. Someone who has declined may be willing to explain, but they do not owe the business a debrief. Ask briefly, accept their limits and avoid making the answer a condition of ending contact. The purpose is learning, not finding another opening after the person has closed the conversation.
When reviewing a pattern, return to some of the underlying records. Ten entries marked timing might describe ten different situations. They might also reveal the same missing piece of information appearing repeatedly. The category tells you where to look. The evidence tells you whether a proposed response has any connection to the problem.
The strongest objection: this can become robotic
There is a fair objection to everything I have said about stages and records. A sales process can become an exercise in filling boxes. The person making an enquiry receives scripted questions, repeats their circumstances to several people and feels processed rather than understood. Meanwhile, the team spends more time maintaining the system than helping customers.
That is a real failure, not resistance to discipline. A framework has to earn the time it takes. If a field does not improve the next conversation, support a real decision or meet a genuine operational need, question why it is there. Collecting more information is not automatically a more intelligent way to work.
The same applies to scripts. A prompt can remind somebody to explain an important distinction. It should not require them to ignore an answer because the next line has already been written. A capable person needs enough room to recognise that the customer has asked a different question, or that the proposed product is simply unsuitable.
Targets introduce another risk. If the only rewarded measure is the number of conversations, staff can increase contact without increasing usefulness. If it is offers, an expression of interest can be promoted into an offer before the customer intended it. If it is conversion, difficult enquiries can be excluded from the denominator. The reported improvement may be a change in classification rather than better selling.
The response is not to abandon measurement. Keep the definitions stable, inspect exceptions and look beyond the rewarded number. A conversation count can sit beside unanswered questions. Completed purchases can be reviewed alongside promises that were not met. A manager should be willing to recognise a well-handled refusal, even though it adds nothing to that month’s revenue.
Nor should a record become a judgement about a customer’s character. Busy, uncertain and unsuitable are not the same as dishonest or unserious. Use language another colleague can act on without inheriting an unnecessary prejudice. Describe the decision and the evidence, not a personality assessment assembled from a few messages.
The test of the system is practical: does it help a capable person respond more accurately with less avoidable repetition? If not, simplify it. Removing an unused field can be an improvement. So can replacing a compulsory script with a short set of questions that only need asking when the answers are not already known.
Carry the actual agreement forward
A sales conversation does not stand apart from delivery. What the customer was told can shape what they expect afterwards. If an important detail stays only in a salesperson’s memory, the next person may begin with an incomplete picture. That is where a clear record has a direct purpose.
Distinguish the customer’s request, the business’s confirmed commitment and a question still being investigated. A request is not automatically an agreement. An internal discussion is not necessarily an approved change. Recording those distinctions protects the customer from assuming something has been settled when it has not.
For a property transaction, the relevant documents and professional advice matter. A general sales note should not be presented as a substitute for the agreement or as a way to resolve a disputed term. Where the conversation and the formal information appear inconsistent, identify the issue and have it clarified through the proper channel rather than letting the ambiguity travel downstream.
This is not a case for copying every conversation to everybody. Pass on what the next person needs to perform the agreed work, with appropriate care for private information. A useful handover is specific enough to prevent the customer having to reconstruct the decision, and limited enough that the important details remain visible.
Then use later questions to improve the earlier explanation. If customers repeatedly ask what a particular stage means after they reach it, review how that stage was introduced. The answer may be a clearer explanation at the right moment, rather than another document added to a collection nobody knows how to navigate. I explore the longer-term effect of keeping those promises in A Brand Is the Memory of Repeated Behaviour. Here, the immediate task is simpler: make sure the next person knows what was actually agreed.
Review a decision, not just a total
A practical review can begin with a small selection of opportunities: one that completed, one that closed without a purchase and one that is still active. This is a proposed review method, not a claim about a particular WC meeting. The purpose is to see whether the records explain the work behind the reported totals.
For each one, ask what the customer was trying to decide. Can somebody other than the original salesperson explain the essential requirement? Is the available offer described accurately? Are the unresolved questions identifiable? If the account consists mainly of contact attempts and optimistic adjectives, it may contain very little decision-useful information.
Next, ask what is known and how it is known. Which statements came from the customer? Which were checked against product information? Which remain assumptions? A record does not need to be long to make that distinction clear. A short, precise account can be more useful than pages of messages without an explanation of their significance.
Then check the next action. Is it a commitment somebody accepted, or a task the seller assigned to themselves? Sending requested information is different from deciding to call again. Both may appear as future activity, but only one necessarily reflects an agreement with the customer. The review should make that difference visible.
For the completed purchase, ask whether anything learned should change the explanation for future customers. For the closed opportunity, check whether the recorded reason is supported. For the active one, ask why it remains active and whether the next step still makes sense. Do not keep it open merely because closing it would make the pipeline smaller.
Finally, choose a specific improvement and a way to assess it. If a comparison was unclear, revise the comparison. If an enquiry waited in the wrong place, fix its route. If staff repeatedly guessed an answer, provide a dependable source or a clear escalation path. Avoid turning a review of a precise problem into a vague instruction to try harder.
A small review does not establish the experience of every customer. It can expose a problem worth checking more widely. That distinction matters: use examples to generate and test questions, not to declare a universal result from whichever three records happened to be convenient.
Earn a clear answer
I want a business to sell. Making something useful and then refusing to explain it or ask for a decision is not a virtue. The commercial work matters. But it should be possible to take that work seriously without treating every hesitation as resistance to overcome.
A good sales conversation connects the product to a real requirement. It makes a limitation visible while there is still time to consider it. It answers what can be answered, identifies what cannot and establishes what happens next. Sometimes the result is a purchase. Sometimes it is a better-defined question. Sometimes it is a clear no.
Those outcomes are not commercially identical. They should not be counted as though they are. But an honest account of each gives the business something useful: revenue it can deliver against, a specific piece of work to complete or evidence about where the offer does not fit.
That is the standard I am arguing for. Ask for the order. Know what the numbers mean. Listen carefully enough to discover when the answer should be different. The customer should leave understanding the decision more clearly than when they arrived.