Presell / DTC generally
DTC: the page has not changed, and nobody can say whether a new one would help
Under the vertical-specific problems there is one shared pattern. Creative gets tested constantly, acquisition gets more expensive, and the destination page stays where it was, because writing a better one is slow and proving it was better is harder.
Both halves of the problem, and they are different problems
The first half is production. A page per angle is obviously better than one page for every ad, and no team does it, because each page is a brief, a writer, a designer and a build, and the angle that justified it has usually been retired by the time it ships.
The second half is proof, and it is the one that never gets solved. A new page with its own analytics reports what the new page did. It cannot tell you whether it beat the page you were already running, because the two are counted by different instruments with different attribution, and comparing them compares two definitions of a session rather than two pages.
Solving production without proof gets you more pages and the same argument. That is why the measurement here is not an add-on: the tag comes in two halves, one for the new page and one for the page it is replacing, and both sides carry the same instrument.
What is free here, and why that is a design decision
Measured traffic, unmetered
Sessions, tag events and experiments carry no charge and no ceiling, however much you spend on ads. If tagging the page you already had cost anything, nobody would tag it, and without the control there is no comparison to have.
Stored pages, unlimited
A stored page is a row in a database and costs nothing whether or not it is hosted. Charging rent on pages sitting in a database would be metering the wrong thing, so we do not.
Exports and version history
Pure renders and rows. Every version of every page is kept and any earlier one can be restored, and none of that is metered either.
What does cost
Generation, rewriting, images and the research behind a brand, because those spend real money on models and scraping. Meter real cost, never invent scarcity.
On conversion multiples
This page does not tell you a presell page converts some number of times better than a product page. It is a popular claim in this category and we are not going to repeat it, because it is not our number, it arrives with no methodology, and every statistic we publish carries a source and a date.
What we built instead is the thing that would make an honest version possible: control and variant carrying the same instrument, with sample size, duration and a confidence interval attached. When customers have run real tests, we will publish real figures with all of that shown. Until then the argument is the mechanism, not a multiple.
Where to start
If you are already running paid traffic, start with what you have. The message match grader tells you whether the page continues the promise the ad makes, and the claim risk scanner tells you which of the promises the page does make it cannot back up. Both are free, neither needs an account, and between them they will tell you whether you need a new page or a fixed one.
Start with your product URL
We read your site, your reviews and your competitors before writing anything, and every factual claim on the page carries the source it came from. Presell is included with Compete.