When an ad says most users felt better, ask who was surveyed, how many, compared with what, measured how, and paid by whom. Without clear answers, a percentage is just decoration. Look for named studies, honest limits, and caution words, and be wary of results-guaranteed language.
You have probably seen the banner: 96% of users felt better. It looks scientific. It has a number, a percent sign, and maybe a clipboard icon. What it rarely has is a sentence explaining where the number came from.
This post is not about any single product. It is a toolkit of five questions you can ask of any statistic in any ad, plus a few patterns that should make you slow down. The goal is not cynicism, just a habit of asking before you buy.
Question one: who was counted?
A survey of ten friends and a survey of ten thousand strangers can produce the same percentage. Ask how many people were included, who they were, and how they were chosen.
If the group is only customers who stuck with the product, the unhappy ones may have already left. Selection quietly bends results toward a pleasant answer, and the ad will not mention it.
A useful trick is to rewrite the claim in plain words. 96% felt better becomes some number of people who answered a question said yes. Seen that way, the missing details are obvious.
Question two: compared with what?
Felt better than when? Better than a sugar pill, better than doing nothing, or better than last Tuesday? Good studies usually include a comparison group so that natural ups and downs and the placebo effect can be separated from the product itself.
Mood and energy fluctuate on their own. If people start a product during a rough patch, feeling better a few weeks later can happen with or without it.
Be aware that even honest comparison groups can mislead if the people knew what they were taking. Expectation alone can shift how someone rates their own energy, which is why careful studies try to keep participants unaware.
Question three: how was better measured?
Feeling better is vague. A reliable measure is defined in advance, such as a questionnaire used consistently or a timed task. An ad that simply asks Did you feel better? invites a yes from anyone who likes the idea of the product.
Also check whether the claim is about the final product or about an ingredient in a lab. Research on an ingredient, even if real, does not automatically transfer to a gummy with unknown amounts of it.
A worked example of reading a claim
Imagine an ad that says most customers reported feeling more energetic. Run it through the questions. How many customers, and were they all asked? Compared with how they felt before, or with people who took something else? Measured how, and by whom? Did the company pay for the survey?
If the page answers none of these, you have learned that the claim cannot be checked. That does not prove the product is useless. It means the statistic gives you nothing to stand on, so you should weigh other factors such as label transparency, safety information and the return policy.
Compare that with a cautious sentence such as an ingredient is studied for tiredness, though the evidence is mixed. It is less exciting, but you can follow it to PubMed and judge for yourself.
Watch for borrowed credibility as well. An ad may mention a university, a journal or a famous broadcaster without saying their work supports this product. Proximity is not endorsement, so click through and see what the source actually says.
Question four: who paid and who is talking?
Testimonials are stories, not evidence, and the FTC has guidance requiring that endorsements be honest and that material connections, like payment or free product, be disclosed. A glowing review with no disclosure deserves doubt, including on affiliate pages like this one, which is independent but may earn a commission if you buy.
Be alert to before and after photos, doctor-sounding names without credentials you can verify, and results that sound the same across many reviews.
Finally, give yourself a cooling-off period. Close the tab, sleep on it, and come back the next day with the five questions written down. Ads are built to produce a quick feeling, and a day of distance often turns a compelling pitch into an ordinary one.
- Round, impressive numbers with no footnote
- Phrases like clinically proven with no study named
- Urgent countdown timers and only a few left
- Guarantees of results rather than of refunds
Question five: what would change my mind?
Read the ad as if you were trying to find the weak point. Does it admit limits? Honest pages use words like may, studied for, and evidence is mixed. Pages that promise certainty on a supplement are overreaching.
A strong sign of respect is a clear refund policy. The brand behind Soda Steel, for example, advertises a 60-day satisfaction guarantee, which is a business term rather than evidence that the product works. Compare the policy, not the promise.
Key takeaways
- A percentage means little without sample size, comparison and method.
- Testimonials are stories; disclosures of payment matter under FTC guidance.
- Ingredient research does not automatically apply to a specific product.
- Prefer pages with cautious language and a clear refund policy.
Frequently asked questions
Sources and further reading
- FTC: Endorsements, Influencers, and Reviews
- NCCIH: Using Dietary Supplements Wisely
- PubMed: L-citrulline
Learn more about Soda Steel
Explore our independent guides: What is Soda Steel?, Soda Steel ingredients, how it works and pricing and refund terms, or return to the blog.