The Monday Brief Every week: what your customers asked for, what they couldn't find, and the one fix worth making first. See a brief →
Problem How it works Safe to go live Intelligence Free audit Pricing Blog Docs
All posts
AnalyticseCommerceGuides

Silent Demand: The Sales Your Analytics Will Never Show You

October 14, 20267 min read
AskVio Unmet demand report listing products customers asked for that the store does not carry

Somewhere on your site today, a customer typed "hand cream 250 ml", got nothing useful, and left. Tomorrow another one will. By the end of the month there could be dozens of them, all asking for the same thing in their own words, and no report you own will show a single one. That is silent demand: what customers ask your website for and don't get. It stays hidden because analytics was never built to record it, not because it is rare.

What silent demand is

Silent demand is every need a customer states on your site that goes unmet and unrecorded. It comes in three forms, and each one belongs to a different team.

Product gap · merchandising

You don't sell it

A size, format, variant or product you don't carry. The customer wanted to buy and there was nothing to buy.

"Do you have the hand cream in 250 ml?"
Content gap · ecommerce and content

You never wrote it down

The answer exists inside your company, in a supplier sheet or someone's head, but not on your site.

"Is this serum safe to use with retinol?"
Policy gap · operations

Your answer is unclear

A rule exists, but the customer can't tell what it means for their order.

"Do you deliver to Ticino on Saturdays?"

The split matters. A list of "unanswered questions" with no owner gets read once and forgotten. The same list split by owner turns into three sets of tasks.

If you have read our post on Failed Intent Rate, product and content gaps are its catalog and content components. Failed Intent Rate tells you how often your site fails. Silent demand is the list of what people actually wanted.

Why your analytics will never show it

GA4, heatmaps and session replay all record behaviour. They tell you a visitor viewed three product pages, scrolled 60% of the way down and left. They can't tell you why, because the reason was never an event.

The closest thing GA4 has is the view_search_results event, which stores the term a visitor typed into your site search when enhanced measurement is on. It is useful, and almost nobody reads it as demand. It holds keywords rather than the need behind them, it says nothing about whether the results helped, and it only covers the search bar.

Chat is worse. The richest statements of intent your customers make, full sentences with a size, a budget or a deadline in them, sit in Intercom, Zendesk or Tidio. Support reads them one at a time to close tickets. Nobody reads them together to decide what to stock or what to write.

The test: open your last three range or content meetings. Was any decision backed by the number of customers who asked for the thing? If not, the signal exists in your stack and isn't reaching the people who decide.

Where it already lives in your stack

You don't need a new tool to start. Silent demand is already being collected in at least three places.

SourceWhat it holdsHow to get it out
Site searchTerms typed, including the ones that returned nothingGA4: Reports › Engagement › Events › view_search_results, add the Search term dimension, download CSV. Algolia: Analytics › No results.
Chat and help deskFull questions in the customer's words, and what the bot or agent repliedConversation exports from Intercom, Zendesk, Chatbase, Tidio, LiveChat, Gorgias or Crisp
Support inboxThe same ten questions asked every weekYour help desk's ticket export

Thirty to ninety days from any one of these is enough to see the pattern.

What it looks like when you read it

Here is one week from Roast & Origin, the demo coffee store we use to test AskVio. These are demo numbers, not a customer's, but the shape is what you will see on a real site.

202
questions visitors asked in one week
9
asks for coffee capsules, which the store doesn't sell
6
conversations where the subscription plan went unanswered

Read as exits, this is noise. Read as demand, it is two decisions: whether to stock capsules, and whether to publish a subscription page. Neither appears anywhere in GA4. Both were stated in plain words by customers who then left.

AskVio Unmet demand report on a demo coffee store, with product gaps such as subscription plans and single-serve coffee ranked by asks

Product gaps on the demo store, ranked by how often they were asked.

Three questions for every gap

1

What kind of gap is it?

Product, content or policy. This decides who owns the fix.

2

How often, and is it growing?

One ask is an anecdote. Ten in a week, rising, is a trend. Count needs, not phrasings: "250 ml", "a bigger size" and "the large one" are the same request.

3

What is it worth?

A rough figure moves a range meeting further than a long list. The next section shows how to get one.

Then pick one. The value of the exercise is the fix you make on Monday, not the length of the list.

Putting a number on it

You don't need attribution to estimate silent demand. A conservative calculation is enough to rank gaps against each other.

Monthly value of a gap
asks per month × share who would have bought × average order value
An estimate for ranking, not a forecast.

An illustration with invented inputs: 67 asks a month for a 250 ml size, one in five of those visitors would have bought, and an average order of €35. That gives 67 × 0.2 × €35, about €470 a month or €5,600 a year, from one missing size.

Use your own inputs. For the share who would have bought, start from the conversion rate of sessions that used search, which you can read in GA4 by comparing sessions with and without a view_search_results event. It is a better proxy than your site average.

There is a better input than any assumption: ask. When a visitor requests something you don't sell, offer to notify them if you add it. Sign-ups per gap are demand with an email address attached, and they turn the estimate into a list of buyers.

Measure your silent demand this week

01

Export one source

30 to 90 days of site search terms or chat transcripts. One source is enough to start.

02

Remove personal data

Names, email addresses, phone and order numbers.

03

Group by need

Merge every phrasing that needs the same answer into one row.

04

Tag and count

Mark each group as product, content or policy, and count the asks.

05

Rank and fix one

Order by count and estimated value. Make the top fix and write down the date.

06

Count again in four weeks

If the fix worked, asks for a content or policy gap fall, and a product gap turns into sales.

A spreadsheet and an afternoon will handle a few hundred lines. Beyond that, grouping by hand gets slow and inconsistent. That is the part we automated.

Or send us the file

The Silent Demand Audit does steps 2 to 5 for you. Upload an export from Intercom, Zendesk, Chatbase, Tidio, LiveChat, Gorgias or Crisp, or your GA4 or Algolia search terms. You get back the top gaps with counts, split into products you don't carry and information your site doesn't give, plus the one fix worth making first, in the same format as our weekly Monday Brief.

Emails and phone numbers are removed before analysis, and the file is deleted once the report is ready.

Find your silent demand

Free audit. No install, one export file, a ranked list of what your customers asked for and didn't get.