ProstaFlow Reviews: Rating and Buyer Comments

Those four numbers come from the distributor and none of them has been audited by anybody outside it. They are printed here because withholding a seller's own figures would be a strange sort of transparency, and they are printed with their provenance attached because that is the only honest way to print them.

One piece of arithmetic is worth doing in public. 747 written reviews against 9,707 orders is a response rate of about one in thirteen, which is high for a supplement and ordinary for a seller that asks for a review by email. A rate that high usually means the reviews were solicited rather than volunteered, and solicited reviews skew positive for reasons that have nothing to do with the product.

Eight rows and three comment texts. The right-hand column names which of the three each buyer was given, so that the repetition is visible rather than hidden behind eight sets of quotation marks.

The fifth column is the interesting one. Every entry on this roster was filed between day 18 and day 52, which means the whole of it sits inside the refund window. Nobody on this list is reporting from month four, and nobody is reporting from week one either.

Five of the eight bought six bottles, two bought two, and one bought three. That distribution does not match the seller's own statement that three bottles is its best-selling tier, which is a reminder that a roster of eight is a sample somebody chose rather than a cross-section of the order book.

Printed at their real length. Three sentences is what the distributor supplied, and stretching three into eight is the thing this page exists not to do.

Read as three statements rather than eight, the roster says something quite specific. The majority note is about a habit holding and a reorder being placed early, which is a statement about the routine rather than about the formula. The second is a mixed verdict at a month with impatience in week one attached to it. The third is somebody who wanted a faster change, telephoned the desk and stopped at two bottles.

None of those three is a claim about what the panel does, and none of them is contradicted by the panel either. A seven-row amino acid formula at these amounts would not be expected to produce a week-one change, and the one complaint on the roster is exactly that it did not.

The three short lines the distributor supplies separately, verbatim: “Daily capsule.” (Carl, 57); “Wait past week two.” (Derek, 65); “Desk answered.” (Luis, 62). The third of those describes somebody answering the telephone at this desk rather than describing the capsule, and it is worth separating from the other two.

Sixty-six per cent at five stars and eighty-seven per cent at four or better is a distribution shaped like most supplement review sets: heavily top-loaded, with a thin tail. Five per cent sits at one or two stars.

What a distribution of this shape cannot tell you is who is missing from it. People who stop taking a product in week two are the least likely group to write anything at all, and they are also the group whose experience would be most informative. That is true of every review set in every category and it is worth remembering before reading anything into the 4.6.

Set against the split, the seller's own figure of 90% recommending is consistent: recommendation usually tracks the four-and-five share, and here it does to within three points.

Five rows, none of which is evidence that the product works. A roster matching what a panel predicts is a sign that the roster has not been dressed up, which is a different and smaller claim.

The comparison is worth making because the commonest failure of a review roster is not fabrication. It is a set of comments describing effects the formula could not plausibly produce on the timescale claimed, and this roster does not do that.

• Whether the capsule worked. Nobody on this roster was compared against anybody taking a placebo, and in an uncontrolled setting improvement and the passage of time look identical.
• Whether the people who stopped were satisfied. A roster is drawn from people who chose to write, and the ones who quietly stopped in week two are absent from every review set ever published.
• Whether eight is representative of 747. It is one per cent of the written reviews and under a tenth of a per cent of the orders.
• Whether the underlying transcripts exist. This desk received summary copy rather than a review file, and says so rather than implying an archive it has not seen.
• What any of it means for a reader in particular. Individual variation in this category is large, which is why a written baseline is worth more to one buyer than a roster of eight.

Three things, and any of them would move the review date at the foot of this page forward.

A different roster from the distributor, which would be reproduced the same way: as a table, with the repeated texts counted. A verified review file, which would let this page say something it currently cannot. Or a correction from a reader who bought the product and finds a figure here wrong.

Send any of those to [email protected]. A correction is made on the page that carried the error rather than noted elsewhere, and the about page sets out the four steps between a message and a changed sentence.