Why People Don't Buy: Stop Guessing Why

You close the chat after another “thanks, I’ll think about it.” An unpleasant feeling remains: something went wrong again. Maybe it is too expensive. Maybe the person was simply not serious. Maybe you need to redo the whole website, price list, and your entire life by Monday.

But after a few days, you can no longer remember all those chats. One client asked about timing, another disappeared after the price, a third wanted examples, a fourth kept clarifying details and eventually chose someone else. When all of this lives in messengers and in your memory, the same obstacle can easily disguise itself as coincidence. And the other way around: several random refusals can look like one big problem.

An upsetting conversation can stand out in your memory more than an ordinary one. Afterwards, you may find yourself treating it as if all clients think the same way. That is a reason to check your impression against notes, not to blame your intuition. When you compare dozens of similar conversations, a record lets you return to the details instead of relying only on what comes to mind.

You do not need to solve every person immediately. You need to stop filling in explanations where there are none. For that, it is enough to keep a short record of lost leads - enquiries that could have become an order but did not. It will not promise sales or give you a magically precise answer for every chat. But it will help you see what actually repeats and what you merely remember well because it was upsetting.

This is not a separate complex system and not a reason to postpone work until you find the “right” service. You are not collecting dossiers on people, but working memory for yourself. Its task is simple: after another refusal, leave a trace you can return to without having to retell the story to yourself.

When a refusal sounds like “it is too expensive for me,” your hand reaches to lower the price. That is understandable: the phrase is short, painful, and seems to leave no room for another conclusion. But it does not always mean your price list is too high.

The person may have been comparing you with a different service format. They may not have understood what exactly is included in the offer. They may not have had the needed amount right now. They may have been looking for the lowest price, while you sell a different scope of work. Or they may have been ending the conversation politely without explaining the real reason. All these situations look similar from the outside, but you need to act differently in each one.

Imagine a hypothetical example: Olena bakes custom cakes. After several messages about the cost, she decides to make everything cheaper. But when she starts recording conversations, she sees that in many of them people asked before the price whether morning delivery was available and whether they could see the fillings. These questions do not establish why people declined. They give her something specific to check before rushing to cut prices: are the filling options, delivery terms, and examples of completed orders easy to find and understand?

In this situation, you do not need to convince yourself that price does not matter at all. A more honest conclusion is: “People mention the price, and before that they often ask about delivery and fillings. I do not yet know whether the information is missing, unclear, or simply needs confirmation.” If a review of the offer reveals a gap, the next action becomes specific: add or clarify that information where people will see it before asking about the amount, and keep recording new conversations the same way.

If people directly compare your offer with another one, do not reduce this immediately to a fight against a lower number. Check whether you name the scope of work, timing, service contents, and the result the client will receive in the same way. Usually, comparison is easier when the offer can be explained in one clear sentence instead of being assembled from pieces of correspondence.

A spreadsheet is not meant to turn your work into correspondence accounting. It should take out of your head what your head stores poorly: many similar conversations in a row. A regular spreadsheet, a note with consistent lines, or a CRM - a program where contacts and progress on leads are stored - will do. If you do not have a CRM, that is not a reason to postpone tracking.

The five fields below and the weekly review later in this article are GrandMa’s editorial starting points, not universal requirements from the cited publishers. Adapt them to your sales cycle. First distinguish a closed enquiry from one still in progress: an agreed conversation next month is pending, not yet lost. For silence without an agreed next step, choose a consistent closure rule for your workflow and record it; that rule does not reveal why the person stopped replying.

After every important conversation that you close without an order, record five things:

• the date;
• the service or product they asked about;
• the lead source - the place the person came from, such as a referral, website, or social media;
• the reason for the loss;
• the client’s brief words or important context.

These fields answer different questions. The date stops old and new conversations from blending into one overall impression. The service shows whether the problem concerns the whole business or only one offer. The lead source helps you notice that people coming from different places have different expectations. The reason and comment leave a trace of the conversation itself, not only your reaction to it.

Record it so you can return without translating it

You do not need to record the entire conversation, judge the person’s character, or invent complex categories in advance. If you sell different services, name them as you name them in your work. If the lead came “from Instagram,” do not turn that into a mysterious code. The spreadsheet should make sense to you when you are busy, not only when you have set it up beautifully.

The reason for the loss can be short too: “chose someone else,” “timing does not work,” “price” - when the client actually says so. “Did not reply after the offer” describes what happened, while the reason remains “unknown.” “Asked to return later” belongs with pending enquiries if you agreed on a next step. Keep these distinctions in the comment, or add a status field if you track open and closed enquiries together. Do not change the name of a reason merely because another one sounds neater. If later you cannot explain how “no budget” differs from “too expensive,” those labels will not help you compare records.

A hypothetical five-field entry for Olena could look like this: date — 18 August 2026; product — custom cake; source — Instagram; reason — unknown; comment — “Client: ‘Thank you, I will not order.’ Before the quote, asked about morning delivery. My hypothesis: delivery may have mattered; the client did not confirm this.” The weekly review can group this entry with other delivery questions without counting it as a confirmed delivery-related refusal.

As a rule, a short recurring list of reasons is more useful than a detailed document you stop opening after a few entries. If a field does not help you understand what happened in the conversation, remove it. If without it you have to guess, keep it.

One trap starts not in the spreadsheet, but in your head. A client says: “I’ll think about it.” You record: “too expensive.” After a while, it seems you have accumulated refusals because of price, although not everyone directly mentioned price.

To prevent this, keep three different states in your record. The first is the client’s direct answer: what the person actually said or wrote. The second is your assumption: a version of why the conversation stopped. The third is an unknown reason: when there is no answer or it is too vague.

The GOV.UK guide to analysing a research session makes the same distinction for user research: record exactly what was seen or heard, then group similar observations, formulate findings separately, and decide what to do next. Here we adapt that principle to sales notes. A direct quote establishes what the client said; it does not necessarily establish their complete motive.

In the comment, you can write: “Client: ‘I’ll come back after the season.’ My assumption: there is no budget right now.” Record the stated postponement and any agreed next step. If the enquiry is later closed without an explanation, the reason is “unknown”; you cannot tell from that phrase alone whether it was a polite way to end the chat. This separation may seem minor, but it helps you avoid decisions built on your own mood.

You can ask about the reason gently if the conversation is open to it. Start with an open question: “Thank you for letting me know. If you are comfortable sharing, what was decisive in your decision not to proceed?” The GOV.UK guide to in-depth interviews recommends open, neutral questions, real examples, and clarification of unclear answers. Applied here, that means letting the client name the issue before suggesting budget or timing. If they want to continue and say something unclear, you can ask what they mean. The person is not obliged to answer. Do not interrogate them, argue, or repeat the request after silence or a refusal to explain. No answer is information too, just not about a specific reason.

Sometimes a client gives a very general reason: “Not for me,” “I changed my mind,” “Not relevant for now.” Do not force these words to mean more than they mean. Record the phrase and next to it only the context that was actually in the conversation. If later such answers repeat at one point in the process, you will have a reason to look at what exactly the person sees before refusing. Until there is a pattern, an honest “unclear” is more useful than a neat theory.

For the most part, this approach is useful for the team too. When one person records “the client is unreasonable” and another records the exact phrase from the conversation, it is impossible to discuss the problem. When there are facts and separate versions, you can argue about the next action rather than memory. Agree that everyone calls the same situations by the same words, otherwise one barrier will fall apart into several pseudo-reasons.

Once a week, review all losses together. Not to put yourself or the salesperson on trial. Look at which labels come up again: for which service, after which stage of the conversation, from which lead source, and with which client words.

You do not have to count everything down to the smallest detail. Look for the same answers and questions, and read similar records side by side: this is how you see a pattern rather than a random label.

Sometimes the signal will be obvious: several people do not understand what exactly they will receive. Sometimes it will be contradictory: some write about price, others about timing, and others simply disappear. Do not force the records to give an unambiguous answer before they can. If there are few reasons or they differ, your conclusion can be simple: “I do not yet know what matters most. I will keep recording consistently.”

Test one change at a time

When you can see a pattern, choose one change to test. Do not change the website, ads, script, price list, and packaging at the same time: otherwise it becomes harder to understand what had an effect. Even one change does not make a simple before-and-after comparison proof of cause and effect.

For advertising experiments, Google Ads Help recommends a clear hypothesis, one variable at a time, one or two success metrics chosen before the test, and records of past experiments. These are useful planning principles for our editorial workflow too, but successive sales conversations are not equivalent to a controlled advertising experiment. Lead sources, season, availability, and the people enquiring can change. Decide beforehand what you will observe and compare similar enquiries; treat any apparent improvement as a signal to investigate, not a guaranteed effect of your change.

In another hypothetical example, Artem renovates apartments. Several times, his records repeat “the work start date does not work.” At the same time, these people came through referrals and hardly negotiated. Instead of a discount, Artem can test one simple change: before the estimate, state the nearest possible start date and explain how the queue is formed. If in new conversations there are fewer questions about timing or people disappear less often right after this point, there is a reason to continue the test. Fewer questions alone do not mean more orders, so he should also track whether comparable enquiries reach an agreed next step or an order. If nothing changes, that is not failure: these observations have not confirmed the assumption, and he can look further. When there are too many changes, the process especially loves to ceremoniously announce five renovations and then hide which one was needed.

Before the next review, note next to the change what exactly you changed and why. Not for a report to anyone, but so that after a few conversations you do not forget what you were testing. This also protects you from the temptation to declare success or failure based on one vivid case.

For regular lead tracking, a analytics system may be useful, but its purpose is the same: not to collect data for data’s sake, but to connect it to the next decision.

The first records may feel awkward. Some reasons will be unknown. Several refusals may have nothing in common. You may see “price” and “too expensive” in different wording but not know whether they are one signal. This is a normal state at the start.

Instead of telling yourself “clients do not buy because of price,” say it more precisely: “In several conversations, people mentioned price; I do not yet know whether this is about the amount itself, a comparison, or an unclear offer.” This wording is not weaker. It leaves you options to test and does not force you to punish the price list for someone else’s vague answer.

If there are few records, do not draw a big conclusion. If there are many reasons, do not try to fix all of them. If the team names the same situation differently, agree on simple shared words in the list of reasons. If the problem seems obvious to you, find a direct client answer in the records, not only your own impression. The reverse happens too: a person seems to have named a reason, but it does not explain what to change. “Chose someone else” is a fact, but not a ready instruction to copy a competitor. First check whether that answer repeats for a particular service or after a particular moment in the conversation. If not, leave it on the list without a big reaction. Not every record has to end in a change.

Over time, the spreadsheet will become not an archive of failures but working memory. It will not replace a conversation with a client or make every offer successful. But it usually removes the most exhausting part: the need to start from an assumption every time.

The first step is simple: after the next important conversation that you close without an order, record the date, service, source, reason, and the client’s exact phrase - or an honest “unknown.”