What it is
Anchoring is the tendency for a number you see first to pull later judgments towards it. In the classic study, people estimated quantities after seeing a number they knew was random, and their estimates still leaned towards it (Tversky and Kahneman, 1974).
On a website, the “first number” is often a price: a higher plan, a previous price, the cost of the alternative.
How strong is the evidence?
Strong. Anchoring is one of the most robust findings in judgment research and appears in many domains, including pricing and negotiation (Furnham and Boo, 2011). The size of the effect varies a lot: implausible anchors, expertise and how much people already know about prices can weaken it.
The CRO step, from “anchors affect estimates” to “this anchor will raise conversion on my page”, is an inference. Test it.
How it works on a page
People rarely know what something “should” cost. They look for a reference point. Whatever the page shows first becomes that reference, and the price is judged as a distance from it.
Examples
Ecommerce
- Honest: a sale price shown next to the real price it sold at recently.
- Honest: “A set of three costs €45; bought separately they’d be €54.”
- Manipulative: a “was €99” that the product never actually sold at.
SaaS
- Honest: a pricing table that shows the plans in a consistent order, with the enterprise plan as a real option.
- Honest: comparing the subscription with the real cost of the work it replaces, if you can back the number.
- Manipulative: a “premium” plan nobody can buy, added only to make the middle plan look cheap.
Ethics
Anchoring sits close to deceptive design. An invented reference price misleads people about the deal they’re getting, and many consumer protection rules require “was” prices to be genuine. For the wider family of manipulative patterns, see the catalogue at Deceptive.design. Keep anchors real, current and easy to verify. See ethical CRO.
How to test it
- Hypothesis: “Because pricing-page recordings show visitors comparing plans back and forth, we believe showing plans from highest to lowest will increase paid sign-ups, because the higher plan sets the reference point.”
- Primary metric: paid sign-ups or revenue per visitor, not clicks on a plan.
- Guardrails: refunds, downgrades and support tickets about pricing.
- Watch the mix: an anchor might move people between plans without changing total revenue. Measure revenue per visitor to see that.
Read more about how people choose between options in choice overload.