How long customers need to decide: calculate it from your data instead of guessing

Almost every team has a ready answer that it has never verified. “Our customers decide immediately,” “in our industry it takes two weeks,” “nobody buys anything in summer anyway.” These statements then determine when reminders are sent, how long reports run and whether a website change is judged effective. Yet the time between the first visit and the inquiry is already in your data—you only need to calculate it.

Why “our customers decide immediately” is almost always untrue

This belief does not come from bad faith, but from what people see each day. A salesperson sees the inquiry and the visit immediately preceding it. The report shows the session in which contact occurred. Everything suggests the decision took a dozen minutes because those are the only minutes recorded in the visible place.

• You see the last visitThe inquiry is attached to the session in which it occurred, not the one that first sparked interest.
• Memory selects extremesThe team remembers the same-day customer and the one who returned after a year, not what is typical.
• Silence looks like lossThree weeks without contact is read as abandonment, even though it is often a normal stage.
• Industry assumptions“In our industry, decisions take X” is usually a borrowed claim, not a measurement.

The consequences are highly practical. If you assume decisions take a day, after a week of silence you write the opportunity off and stop working on it. If you assume they take a quarter, you send the reminder after the customer has long since bought elsewhere. Both mistakes cost the same amount, just in different places.

This article deals only with the length of the decision window and what follows from it. We do not cover connecting visits into a coherent history or attributing sources—that is a separate topic covered in our guide to the customer journey on your website. Here we care about one quantity: the time between the first contact with the website and the inquiry.

How to determine your decision window step by step

You do not need a model or a spreadsheet with ten tabs. You need a list of inquiries from recent months and the ability to trace each one back to the same person's first recorded visit.

1. Choose one type of end event.A chat conversation, call request or inquiry. Do not combine them in one calculation because each has a different threshold.
2. Take the last few dozen cases.Thirty to fifty is enough to reveal the shape of the distribution. Fewer than twenty is still anecdotal.
3. Define the starting point.The first recorded visit from the same browser—not the first visit that month or the first arrival from a campaign.
4. Record the interval in days.One column, one number per case. Record the same day as zero, not one.
5. Exclude cases with no history.If an inquiry has only one session because no earlier visit history exists, exclude it instead of entering zero.
6. Calculate the median, not the mean.Sort the figures and take the middle one. A single customer returning after eight months will distort the mean.
7. Also record the nine-out-of-ten value.The number below which ninety percent of cases fall is the other half of the answer.

Using the median instead of the mean is not an academic nicety; it is essential to a meaningful result. Time-to-decision distributions are always skewed: many cases cluster near the beginning, followed by a long tail of isolated returns months later. The mean of such a set points to a value at which almost no customer actually converted—and that is precisely what the team then enters in its reminder schedule.

Two figures instead of one: typical time and the tail

One figure cannot describe this phenomenon. The median tells you when an inquiry is most likely, but not when it stops being possible. Operational decisions need both: the first sets the rhythm of work, while the second defines how long to remain patient and how much data to retain.

↔ Scroll the table sideways to see all columns.

To illustrate with invented figures: the median falls on day four, nine out of ten inquiries arrive within thirty-one days, and one in twenty arrives after more than three months. A team looking only at the median scheduled reminders for day three and closed cases after a week—losing everything between days seven and thirty-one, which in this illustration is a quarter of all inquiries.

The tail has another consequence that few people consider when choosing a plan. If your typical decision window is longer than your data-retention period, some of the history will simply be unavailable when you need it. In Meetlead, message retention is configurable and depends on the plan, so compare this setting with your own figure rather than a default.

What this figure means in day-to-day work

The figure has no value until it changes four specific decisions. Most companies currently set each of them by intuition, even though all four can be derived from this number.

↔ Scroll the table sideways to see all columns.

The penultimate row is usually the most costly. A website change introduced on Monday cannot show its full effect by Friday if the typical decision takes two weeks—on Friday you see only customers who began their journey before the change. An evaluation after a few days measures noise, not impact. In that situation, base the earlier signal on intermediate steps, which we cover in our article on website micro-conversions.

The second conclusion concerns comparisons between periods. If the decision window is three weeks, comparing two consecutive weeks says nothing about effectiveness because most inquiries in the second week come from people who arrived before the first one. Variations in demand over time add a separate layer, discussed in our article on seasonality in traffic and inquiries.

Measurement pitfalls that make the figure too low

Every common error in this calculation works in the same direction: it shortens the result. The starting point moves closer to the inquiry, making the decision window look shorter than it really is. Keep this in mind before concluding that your customers decide faster than expected.

↔ Scroll the table sideways to see all columns.

It is worth stating this explicitly within the team as well: a session identifier belongs to a browser, not a human being. The same person on two devices creates two histories, while two people sharing one device create one. Treat the result as an order-of-magnitude estimate, not a measurement accurate to the hour. We have collected a broader catalogue of what standard traffic reports cannot show in our article on what Google Analytics does not show.

A specific person's visit history is personal data even if you do not know their name. Calculating intervals for your own analytics and using that knowledge in sales contact are two different purposes, each requiring separate justification, disclosure in the privacy policy and a limited retention period. We break down the obligations surrounding chat and website data in our article on live chat and the GDPR. This material is not a substitute for individual legal advice.

How to use the result without becoming intrusive

Knowing the decision window can easily tempt you to approach everyone who returns for a third visit. That is the fastest way to look like a company that watches customers and makes sure they know it. The difference between helpfulness and intrusion is not frequency, but exactly what you say and what you refer to.

• Make contact within the decision window, not every time someone returns to the website.
• Refer to what the customer said, not what you can see in their visit history.
• One reminder within the window is enough; a second in the same week usually has the opposite effect.
• Instead of asking for a decision, add information that was missing from the proposal.
• Set a final point for closing the case and stick to it instead of returning every month.
• Recalculate your figure after a quarter because a change to the offering can genuinely shorten it.

A separate question is what to check immediately before sending a reminder and how to phrase it so that it does not sound like a demand. We break this down in our article on following up after a proposal. The rule here is enough: the decision window tells you when, not what to write.

The simplest conclusion comes last. If your decision window is longer than a few days, the most important element of service is not how quickly you respond to an inquiry, but whether useful contact is available during all the visits that have not yet produced one. A customer returning for the third time who cannot ask a quick question will usually not return for a fourth.