Is it the copy, the page, or the traffic?
Three different failures produce an identical low conversion rate. Weak copy means the visitor belonged there but was not convinced. A broken page means they were convinced but could not act. Wrong traffic means they were never going to buy. Two checks separate them: conversion rate split by traffic source, and mobile against desktop.
Why the dashboard cannot tell you
A conversion rate is a single number describing the outcome of a chain of quite different things. It falls when the message fails, when the interface fails, and when the audience was wrong — and it falls by the same amount and looks the same in every case.
That matters commercially because the three fixes cost very different amounts and are performed by different people. Rewriting a product page is a week of copy work. Fixing a mobile add-to-cart button is an afternoon of development. Changing who the ads reach is a campaign decision. Picking wrong costs the full price of the work and moves nothing.
Weak copy, a broken page and the wrong audience produce an identical low conversion rate. Both checks below read from analytics you already have, in minutes.
They belonged here. They were not convinced.
They were convinced. They could not act.
They were never going to buy.
One source far below the others, same page? The audience is the variable, not the copy. All sources uniformly low? Traffic is ruled out.
Same words, different interface. A mobile rate at a fraction of desktop is an interface failure. Devices agreeing rules the page out.
Both checks cost minutes and require no changes to the site. Running them first eliminates the two cheapest explanations before any expensive work is commissioned — and rewriting a page that was never the problem costs the same as rewriting one that was.
Check one — split conversion by traffic source
Compare conversion rate by traffic source before anything else. If one source converts at a fraction of the others while the page stays the same, the page is not the variable — the audience is. A page cannot be well written for one source and badly written for another; the copy is identical for both.
This is the fastest check available and it eliminates the most expensive mistake, which is rewriting a page that converts perfectly well for the visitors who should be on it.
Two patterns worth recognising:
One paid source far below the others. Targeting or creative is bringing people who were never in market. That is a campaign problem, not a copy problem.
All sources uniformly low. The audience is not the variable. Move to check two.3. Begrijp het verschil tussen "merk" en "marketing"
Check two— compare mobile against desktop
Compare conversion rate on mobile against desktop for the same page and traffic mix. A large divergence points at the page rather than the message, because the words are identical on both and only the interface changed. Some gap is normal; a mobile rate at a fraction of desktop is an interface failure, not a persuasion failure.
Where it usually is: an add-to-cart pushed below the fold, a variant selector that does not register a tap, a form that triggers the wrong keyboard, a page that loads slowly enough that visitors leave before the argument starts.
Check traffic source first, then device. Both are readable from existing analytics in minutes and require no changes to the site. Running them in that order eliminates the two cheapest explanations before any expensive work is commissioned.
When copy is the right lever
Copy is the right lever when visitors arrive from a source that converts well elsewhere, stay on the page long enough to have read it, convert similarly on mobile and desktop, and still do not act. That pattern says the audience was right and the interface worked — what failed was the argument, the proof, or an unanswered question.
The most common version of this failure is not bad writing. It is a page that never answers the questions a buyer needs answered before paying: what it costs to return, how long delivery takes, whether it fits, what happens if it goes wrong. Those are risk questions, and a page that omits them loses buyers who liked the product.
Copy is the wrong lever when the delivery promise, the price, or the product itself is what fails the comparison against alternatives. Better writing cannot close a gap created by a slower delivery time or a higher price against an equivalent competitor. It can make an honest case; it cannot make a worse offer win.
What this does not tell you
The two checks narrow the cause; they do not size the opportunity. A page problem worth €2.000 a year and one worth €30.000 look identical in this diagnostic. Price the gap before deciding how much work it justifies.
They also assume enough volume to read. Below a few hundred conversions in the window, differences between sources or devices are frequently noise rather than signal.
