How we rate the evidence.
Every principle in the database carries one of four ratings, and this page sets out what each one means, how we arrive at it and, just as usefully, what it doesn’t tell you.
What each rating means, and how to treat it.
Replicated many times, or backed by a meta-analysis.
Safe to build on, though still worth testing on your own site, because a reliable effect isn’t the same as a large one.
For example, Social proof. See all 31
Solid studies, but dependent on context or tested mainly in the lab.
Worth trying, with a test: the effect is real in the studies, but it may not carry over to your products or your customers.
For example, Goal-gradient effect. See all 58
Failed replications, or an effect that is weak or conditional.
Treat with caution, and read the entry to see when it holds and when it doesn’t before building anything on it.
For example, Urgency (time limits). See all 7
A useful way of thinking, not a single measured effect.
Use it to look at a page differently, and don’t expect a measured lift from it.
For example, Information scent (information foraging). See all 8
Has it been tested on real shoppers?
A rating says how well the effect has held up, which isn’t the same as saying it has been shown to work in a shop. Some of the most reliable findings in psychology have never been tested on anyone buying anything, so every entry also carries one of three tags.
At least one study measured what real customers did: sales, purchases, payments or behaviour on a live shop.
51 of 104 principles, for example Anchoring. See all 51
Tested in simulated shops, on stated intentions or in usability studies, or on real customers in a neighbouring setting such as services or web search.
42 of 104 principles, for example Curiosity gap. See all 42
The effect comes from general psychology, and no study in a shopping setting was found. Using it on a site is an inference.
11 of 104 principles, for example Peak-end rule. See all 11
| Rating | Real shoppers | Shopping-like settings | Not tested | All |
|---|---|---|---|---|
| Strong | 15 | 11 | 5 | 31 |
| Moderate | 28 | 29 | 1 | 58 |
| Contested | 4 | 1 | 2 | 7 |
| Framework | 4 | 1 | 3 | 8 |
The tag doesn’t change the rating. A Strong principle that has never been tested in shopping is still Strong; it is simply one to test before relying on.
A famous first study counts for less than what happened when others repeated it.
Five steps, the same for every principle.
- 01Start from the original researchWe find the studies the principle rests on and confirm each exists as cited: authors, year, journal and title.
- 02Look for what came afterWe search for meta-analyses, replications and failed replications, because a famous first study counts for less than what happened when others repeated it.
- 03Look for evidence from real shoppingWe look for studies of real buying behaviour, online where possible. Where a principle has only been tested in a lab, or outside retail, the entry says so.
- 04Check and link every sourceEach citation is checked against the publisher’s page, a university repository or a research index, and linked from the entry.
- 05Rate it and write it upWe give the principle one of the four ratings and summarise what the research found, including where it stops applying.
Three well-known ideas that looked different once we read past the first study.
What a rating does not tell you.
A Strong rating says the effect is reliable across studies; it doesn’t say how big it will be for your products, your prices or your customers. Test it.
A principle can be well supported while a popular way of applying it isn’t, which is why each entry lists the tactics the research supports and any it suggests avoiding.
The ratings are editorial: we read the research and decide, and a careful reader could reasonably land one step either side.
We check each source and read its summary and findings, though not always the full paper, and our search for replications is thorough without being exhaustive.
Where an entry notes UK consumer law, it’s a pointer to check, not a ruling on what you can do.