Free reading
Spotting a Fake Testimonial
Eight checks that work on any sales page — then the same eight applied to the image set supplied to this website.
Written and reviewed by the SodaSlim Editorial Desk · Last updated September 24, 2026
Quick answer
Eight checks, none needing anything but your browser. Look for the same face twice. Check the lighting on the selfies and the angle of the product. Read the filenames. Check the name format, then the sentences for repeats. Weight the star split against the printed average. And ask what citation sits behind each journal logo.
The method
Why the gallery is the part of a sales page to check first
The panel on the back of a bottle is regulated. The photographs are not.
On a supplement sales page the panel is regulated and the gallery is not. Nobody has to substantiate a photograph, prove that a smiling person in a kitchen bought anything, or hold evidence that a sentence in quotation marks was ever typed by the person whose first name sits under it. That asymmetry is why the gallery is usually the most persuasive thing on the page and almost always the least accountable. What follows is a method rather than a verdict: eight checks, in the order they are quickest to run, each answerable from a page’s own material in minutes. At the end the same eight are run on the images supplied to this website.
Why this matters more here than it looks
FDA makes the connection itself. On weight-loss products found to contain hidden drug ingredients: “These products are sometimes falsely advertised as dietary supplements, food and/or all-natural treatments. They often have multiple positive reviews and/or are being shared through social media.” NCCIH puts the reader’s side of it: “Ask yourself if a product sounds too good to be true. Be cautious if the claims for the product seem exaggerated or unrealistic and use phrases like ‘quick and effective’ or ‘totally safe.’”
What a testimonial can and cannot tell you
Even a wholly genuine testimonial is one person’s recollection of an uncontrolled experiment they ran on themselves, with no comparison group and no way to separate the capsule from everything else that changed that month. That is not a criticism of the person; it is a description of the evidence type. A real testimonial is weak evidence and a staged one is none — and a page willing to manufacture its gallery has told you something about the rest of its claims.
The checklist
The eight checks, in order
| # | The check | What it tells you |
|---|---|---|
| 1 | Look for the same face twice | Models are reused. The same jawline in two outfits is the commonest tell there is. |
| 2 | Check the lighting on the selfies | A phone in a kitchen gives uneven light and a wandering camera height. A studio does not. |
| 3 | Check the angle of the product | Every bottle held at the same height, label square to the lens, means a director. |
| 4 | Read the filenames | Inspect the image, or open it in a tab and read the URL. Filenames are rarely tidied. |
| 5 | Look at the name format | A first name, an initial and a city reads as specific and identifies nobody. |
| 6 | Read the sentences for repeats | Copy a distinctive phrase from one and look for it in the others. |
| 7 | Reconcile the star split with the average | Weight the breakdown yourself. Two numbers that disagree were not both measured. |
| 8 | Ask what sits behind each logo | A wordmark with no authors, title, year or link is decoration, not a citation. |
The first four are the pictures, the next two the words, the last two the numbers and the logos.
Checks 1 to 3: the pictures themselves
Scroll the gallery once without reading any of it, looking only at faces. Stock photography is licensed in sets, so a page that bought one shoot has a limited cast and will reuse it; look at jawlines, ears and hairlines rather than clothes and hair, because clothes and hair are what changes between frames of one shoot. Then look at the light: a phone held in someone’s own kitchen gives hard shadows, an uneven background and a wandering camera height, while studio light is even and identical in every frame. Third, look at the product rather than the person. In a snapshot the bottle is wherever the hand happened to be; in a photoshoot it faces the lens at the same height every time.
Checks 4 and 5: the filenames and the names
The fourth check settles arguments and almost nobody runs it. Right-click an image and inspect it, or open it in a tab and read the address. What comes back is the filename the page’s own builders gave it, and filenames are working notes: written for whoever assembles the page, never tidied up because nobody expects a shopper to look.
The fifth is the format of the attributions. A first name, an initial and a city reads as a real, specific person and identifies nobody. A page genuinely publishing feedback usually has something checkable attached — a purchase date, an order reference, a verified-buyer marker from a platform that is not the seller.
Check 6: read the sentences
Take the most distinctive phrase in one testimonial and look for it in the others. Copywriters under deadline repeat themselves; different people do not. Watch for a uniform rhythm too, and the same mild reservation at the end to make the enthusiasm land. Real feedback is ragged.
Checks 7 and 8: the numbers and the logos
The seventh is arithmetic and takes thirty seconds. If a page prints a star breakdown and an average above it, multiply each star value by its share and add them up; if the answer does not match, one of the two numbers was typed rather than calculated. The eighth applies the same principle to authority: a journal wordmark reads as an endorsement whatever the small print says, so look for the citation — authors, title, year, a link.
What the checks cannot do
None of the eight tells you whether the capsules work. They tell you how much weight to put on everything else the page says that you cannot verify.
The worked example
The method run on SodaSlim’s own supplied image set
Each withheld image is described in words in its place.
The method, run on the material supplied to this website. None of the images below is published anywhere on this site, so each is described in words with a marked placeholder in its place.
Checks 1 to 4: the photographs, and what their filenames said
The supplied asset pack contains a set of customer photographs, and the pack named them itself. Three portraits arrived with the filenames same-person-different-clothes-1, same-person-different-clothes-2 and same-person-different-clothes. Two further images, styled as a strip of polaroids captioned as real buyers, show the same three people again — different clothes, a different room, each holding the bottle up to the camera as though they had photographed themselves at home. A fourth headshot is filed under a person’s full name on a plain white background. None of these images appears anywhere on this site. Publishing them would manufacture customers who do not exist, and the filenames make it impossible to pretend otherwise.
Three studio portraits supplied as customer photographs, whose own filenames state they are one set of models in different clothes.
Two strips styled as polaroids of real buyers, showing the same three people again in a different room, each holding the bottle up as though self-photographed.
A fourth studio headshot supplied under a person’s full name. No such customer was ever identified to this desk.
Notice how little work those four checks needed. The first found three faces across five images; the second, studio lighting on supposed self-portraits; the third, the bottle at one height. The fourth stated the conclusion outright.
Checks 5 and 6: the names and the sentences
The supplied pack describes its own review block, in its own words, as “layout copy only”. It is right to. Four of the eight named buyers carry the same sentence word for word — “Fair at a month. Wanted more in week one.” — and three more share a second. Set that beside the photographs described above, whose own filenames say they are one set of models in two sets of clothes, and there is nothing here that survives reading.
Four of eight named buyers sharing a sentence word for word is one person writing eight testimonials and running out of ideas at the fourth.
Check 7: the arithmetic on the star split
There is a second reason. The star split printed in the pack is 66% five star, 21% four, 8% three, 3% two and 2% one. Weight that out and it averages about 4.46. The figure printed above it is 4.7. Neither number is ours and we would publish neither, but a set of figures that does not reconcile with itself is worth saying out loud.
Neither figure is one this site would publish. A mismatch that size means two numbers printed above each other were produced independently of one another.
Check 8: the journal and newspaper logos
One supplied image is a row of five wordmarks: the New England Journal of Medicine, Nature, the International Journal of Obesity, The Boston Globe and the American Physiological Society. There is no caption on it and no citation with it. No study of SodaSlim appears in any of those publications, and none of them has said anything about this product. A logo row reads as an endorsement whatever the small print says, and there is nothing behind this one, so it is not published here.
A row of five journal and newspaper wordmarks supplied with no caption and no citation attached to any of them.
Two further supplied pictures did not survive the same reasoning: a stock photograph of an unnamed man in a white coat, and one of two fitness models. The testing and quality page carries the reasoning.
The takeaway
The two rules, and what to read instead
Two general rules fall out of that exercise, and they are worth more than the findings, because you can carry them to the next page you read.
On photographs
Customer photographs are the cheapest thing on a sales page to fake and the most persuasive thing on it to a reader. If the same face appears twice in two different outfits, if everyone is holding the bottle at the same angle, or if the lighting on a “selfie” is studio lighting, you are looking at a photoshoot. It says nothing about whether the capsule works — but it tells you exactly how much weight to put on everything else the page says about buyers.
On logos
A media or journal logo bar with no article behind it is one of the most common devices in this category. The test is simple and it takes thirty seconds: if a page shows you a journal’s logo, look for the citation — authors, title, year, a link to the paper. If there is no citation, the logo is decoration.
Where to put the weight instead
If the gallery is the least accountable part of a sales page, the Supplement Facts panel is the most. ODS describes where the responsibility sits: “Supplement companies are responsible for having evidence that their products are safe, and the label claims are truthful and not misleading.” That obligation attaches to the panel, not the pictures. So count the rows, look for a proprietary blend, write the stimulant figure down, and look each ingredient up on a federal fact sheet.
Questions
SodaSlim FAQ: spotting a fake testimonial questions
Why don’t you publish the customer photographs?
Because three of them arrived with filenames reading same-person-different-clothes, and the two polaroid strips show those same three people again in different clothes and a different room. Publishing them would manufacture customers who do not exist.
The star split does not add up. Did you check?
We did. The printed split weights to about 4.46 and the figure printed above it is 4.7. Neither number is ours and we would publish neither, but a set of figures that does not reconcile with itself is worth saying out loud.
What about those journal logos?
There is no article behind them. One supplied image places the wordmarks of five publications in a row with no caption and no citation. No study of this product appears in any of them, and we have not published the image.
References
Sources for this page
- National Center for Complementary and Integrative Health, NIH. 6 Things To Know About Dietary Supplements Marketed for Weight Loss. www.nccih.nih.gov. Accessed September 24, 2026.
- National Center for Complementary and Integrative Health, NIH. 5 Tips: What Consumers Need To Know About Dietary Supplements. www.nccih.nih.gov. Accessed September 24, 2026.
- National Center for Complementary and Integrative Health, NIH. Using Dietary Supplements Wisely. www.nccih.nih.gov. Accessed September 24, 2026.
- Office of Dietary Supplements, National Institutes of Health. Dietary Supplements: What You Need to Know. ods.od.nih.gov. Accessed September 24, 2026.
- U.S. Food and Drug Administration. Tainted Weight Loss Products. www.fda.gov. Accessed September 24, 2026.
Next
Keep reading
Read the SodaSlim panel instead of the gallery
Five ingredients, five published amounts, no proprietary blend and three printed standardisations — the part of this product that can be audited. 30 capsules a bottle, one capsule daily. Price is shown at checkout.
Price shown at checkout · 60-day window from the date of purchase · ships from inside the US · the label’s own support line is +1 (302) 200-3480