Over-ear headphones and a smartwatch on an oak desk in low side light.

The spec sheet is complete, and the shopper still cannot choose

Which model, and what it actually works with

Where the specification stops being the answer

Four situations this category produces constantly, and more specification does not resolve any of them.

Every difference is listed and none of them is ranked. A range is built so that each step up matters to somebody, and the page publishes the full matrix without saying which row matters to this shopper. The buyer is not short of information — they are short of somebody who will say which two of the eleven differences apply to what they actually do.

“Will it work with what I already have?” Ports, standards, the cable, the accessory bought two years ago, the thing it has to talk to. It is the most common question in the category and the one a product page answers worst, because the answer depends on what is already in the shopper's house — which is information only they have and only a conversation gets at.

The price is not the price. Trade-in, instalments, warranty, the bundle, what the previous device is worth. Every one of those is a step the shopper has to work out alone on a page designed to show one number, and every one of them is a place where a basket that was ready goes quiet.

The worry is the first hour, not the purchase. Setting it up, moving the data across, whether the old one keeps working in the meantime. Nobody abandons a basket saying so, but the hesitation is real, and it is settled by a person in ninety seconds and by a help centre article almost never.

A consumer electronics storefront with the AI sales agent offering to show two sets of headphones side by side.
The agent recommending earbuds and a USB-C dock, both shown with prices inside the conversation.
The bag inside the storefront widget with the earbuds and an adapter, and the total.

The comparison is two products and one question

A shopper weighing two sets of headphones is not short of specification. The AI sales agent opens with the part that decides it — which of the two works with the watch already on their wrist — and puts both where the question was asked, at whatever hour it arrives.

“Will it work with what I have?” is answerable

The shopper names the watch and the laptop they already own, and the answer comes back out of your catalogue rather than out of a forum: these buds pair without a phone, that dock returns the ports the laptop is missing. Two products, both explained by what the shopper said.

The basket closes where it used to stall

One item, chosen for a reason the shopper can repeat, added while they watch. The adapter was in the bag before the conversation started — the agent finished an order rather than assembling one, which is what the checkout looks like when the compatibility question has already been answered.

iStore is on Immerss. The question that stalls a ready basket is the one a person answers in a sentence.

On selling a considered purchase

All insights →

Eleven differences, and none of them ranked

The matrix is complete, accurate and, for a shopper who already knows the category, the fastest thing you could hand them. Its limit is structural rather than a shortcoming: it compares your products with each other, and the question that decides the order is about something that is not in it.

A comparison matrix
Immerss for electronics
What it compares
Your products against each other
Your products against what the shopper already owns
Who ranks the rows
Nobody — all eleven weigh the same
The two that apply to what they actually do
“Will it work with my…”
Not a column, because the answer is not about you
The first thing it answers, out of your own catalogue
The parts of the price that are not the price
Trade-in, instalments and warranty, on three other pages
In the same conversation as the product
The first hour after it arrives
Out of scope
Setup and migration, asked before the order rather than after

FAQs