The 68 CRO Principles Behind CROAssist’s Test Suggestions

Every test idea CROAssist generates names the psychological principle it is built on. You will see it on the idea card as a label like “Loss Aversion” or “Information Scent”. These are not made-up marketing terms: each one is a documented effect from psychology, behavioural economics, or human-computer interaction research, and knowing what they mean helps you judge an idea before you run it — and write better hypotheses of your own.

This page is the full reference. Sixty-eight principles, grouped the way CROAssist uses them, each with what it means for your pages and where it comes from.

Persuasion

The seven principles from Robert Cialdini’s research on why people say yes.

• Reciprocity — People feel obliged to give back. Offer real value first (a free tool, guide, or sample) before asking for the email or the sale. (Cialdini, Influence, 1984)
• Commitment & Consistency — Once someone takes a small step, they tend to follow through to stay consistent with it. Get a small yes before the big ask. (Cialdini, 1984)
• Social Proof — When unsure, people copy others. User counts, client logos, reviews and testimonials all signal “people like you chose this”. (Cialdini, 1984)
• Authority — Credentials, certifications and press mentions make claims believable. (Cialdini, 1984)
• Liking — We buy from people we like. Relatable faces, shared identity and a warm tone outperform corporate distance. (Cialdini, 1984)
• Scarcity — Limited stock, seats or time makes an offer more desirable. Use only when the limit is real. (Cialdini, 1984; Worchel, 1975)
• Unity — A shared “we” identity with your audience (“built for agencies, by an agency”) persuades beyond mere liking. (Cialdini, Pre-Suasion, 2016)

Behavioural economics

How people actually weigh value, price and risk — mostly from the work of Daniel Kahneman, Amos Tversky and Richard Thaler.

• Loss Aversion — Losses hurt roughly twice as much as equivalent gains feel good. “Don’t lose your progress” can outpull “keep your progress”. (Kahneman & Tversky, 1979)
• Prospect Theory — Value judgments are relative to a reference point, not absolute. Control the comparison your visitor makes. (Kahneman & Tversky, 1979)
• Anchoring Effect — The first number seen skews every later judgment. Showing the premium plan first makes the mid plan feel reasonable. (Tversky & Kahneman, 1974)
• Framing Effect — The same fact lands differently depending on wording: “90% success rate” beats “10% failure rate”. (Tversky & Kahneman, 1981)
• Peak-End Rule — Experiences are remembered by their most intense moment and how they end. Polish the last step of checkout and onboarding. (Kahneman et al., 1993)
• Endowment Effect — People value what they feel they already own. Free trials and personalisation create ownership before purchase. (Thaler, 1980)
• Sunk Cost Fallacy — Effort already invested pulls people forward. “You’re 80% done with setup” discourages abandonment. (Arkes & Blumer, 1985)
• Decoy Effect — A deliberately inferior third option steers choice toward the plan you want chosen. (Huber, Payne & Puto, 1982)
• Default Effect — Whatever is preselected tends to win. Default the annual plan, the recommended tier, the opt-in. (Johnson & Goldstein, 2003)
• Status Quo Bias — Change feels risky even when it is better. Reduce perceived switching pain when selling against an incumbent. (Samuelson & Zeckhauser, 1988)
• Hyperbolic Discounting — Rewards now are worth far more than rewards later. Lead with what the visitor gets today. (Ainslie; Laibson, 1997)
• Zero-Price Effect — “Free” is disproportionately attractive, beyond its economic value. A free tier or free shipping changes behaviour more than the numbers justify. (Shampanier & Ariely, 2007)
• Left-Digit Effect — $29 reads meaningfully cheaper than $30 because we anchor on the left-most digit. (Thomas & Morwitz, 2005)
• Choice Overload — Too many options paralyse the decision. Trimming plans, menu items or form options often lifts conversions. Evidence is mixed on when exactly it bites, but the effect is mainstream. (Iyengar & Lepper, 2000; Schwartz)
• Mental Accounting — People sort money into mental buckets. “Less than a coffee a day” moves a subscription from the “big commitment” bucket to the “trivial” one. (Thaler, 1985)
• Present Bias — “Start today” beats a list of future benefits. Immediate outcomes convert. (O’Donoghue & Rabin, 1999)

UX laws

Classic findings from human-computer interaction and Gestalt psychology about how people use interfaces.

• Hick’s Law — Decision time grows with the number of choices. Fewer menu items and fewer competing CTAs mean faster action. (Hick & Hyman, 1952)
• Fitts’s Law — Bigger, closer targets are faster to hit. Size and position your CTA accordingly, especially on mobile. (Fitts, 1954)
• Miller’s Law — Working memory holds about seven items. Chunk pricing tables, feature lists and forms. (Miller, 1956)
• Jakob’s Law — Visitors spend most of their time on other sites, so they expect yours to work the same way. Familiar patterns beat clever ones. (Nielsen, 2000)
• Aesthetic-Usability Effect — Attractive design is perceived as more usable and more trustworthy, before a single click. (Kurosu & Kashimura, 1995)
• Tesler’s Law — Complexity cannot be destroyed, only moved. Absorb it on your side (smart defaults, auto-detection) so the user doesn’t carry it. (Larry Tesler, Xerox PARC)
• Doherty Threshold — Responses under ~400ms keep users in flow; slower ones break engagement. (Doherty & Thadani, IBM, 1982)
• Serial Position Effect — First and last items in a list are remembered best. Put your strongest points at the ends. (Ebbinghaus; Murdock, 1962)
• Von Restorff Effect — The one thing that looks different gets noticed and remembered. Isolate your primary CTA visually from everything around it. (von Restorff, 1933)
• Zeigarnik Effect — Incomplete tasks nag at the mind. Open loops — an unfinished quiz, a started checkout — pull people through to completion. (Zeigarnik, 1927)
• Ovsiankina Effect — Given the chance, people resume interrupted tasks. The engine behind cart-recovery and “pick up where you left off”. (Ovsiankina, 1928)
• Goal-Gradient Effect — Effort accelerates as the goal gets closer. Progress bars and step counters push completion. (Hull, 1932; Kivetz, 2006)
• Law of Proximity — Elements close together are perceived as related. Group form labels with their fields, prices with their plans. (Gestalt; Wertheimer, 1923)
• Law of Similarity — Elements that look alike are perceived as related. Style all secondary actions consistently so the primary one stands apart. (Gestalt; Wertheimer, 1923)
• Law of Common Region — Borders and card backgrounds group content into single units. (Palmer, 1992)
• Cognitive Load Theory — Mental capacity is finite; every ounce spent decoding the page is unavailable for the buying decision. Cut extraneous effort. (Sweller, 1988)
• Paradox of the Active User — Users act first and read manuals never. Design for doers: make the first action obvious rather than explaining everything up front. (Carroll & Rosson, 1987)

Attention & perception

What people actually see, notice and remember on a page.

• Picture Superiority Effect — Images are remembered far better than words. A real product screenshot can outsell a paragraph of description. (Paivio; Nelson, 1976)
• Mere Exposure Effect — Repeated exposure builds preference. Consistent branding and retargeting work partly on this alone. (Zajonc, 1968)
• Processing Fluency — Copy that is easy to read feels more true and more likable. Simpler sentences convert. (Reber, Schwarz & Winkielman)
• Halo Effect — One strong quality colours the whole judgment. Lead with your single best proof point and it lifts everything else. (Thorndike, 1920)
• Bandwagon Effect — Popularity itself drives adoption: “fastest-growing”, “most popular plan”. (Leibenstein, 1950)
• Curiosity Gap — A teased gap between what we know and what we want to know demands closing. Headlines that withhold the answer earn the click. (Loewenstein, 1994)
• Inattentional Blindness — Users can look straight past anything outside their focus. Key information in the periphery may as well not exist. (Simons & Chabris, 1999)
• Banner Blindness — Anything styled like an ad gets ignored, including your own content. Don’t make your guarantee look like a banner. (Benway & Lane, 1998)
• F-Pattern Reading — Users scan pages in an F shape: across the top, then down the left. Front-load your value there. (Nielsen / NN/g, 2006)
• False Bottom — A visual break can make the page look finished when it isn’t, so visitors never see what’s below. Add continuation cues to keep them scrolling. (Tognazzini, 1998; NN/g)
• Signifiers & Affordances — Clickable things must look clickable. Flat text that is secretly a button converts like flat text. (Gibson; Norman, 1988)

Motivation & behaviour change

What gets people to actually act, not just agree.

• Fogg Behavior Model — Behaviour happens when Motivation, Ability and a Prompt line up at the same moment. A conversion failing usually means one of the three is missing. (BJ Fogg, Stanford, 2009)
• Foot-in-the-Door — A small request first (an email address) makes the bigger request (a purchase) more likely to succeed later. (Freedman & Fraser, 1966)
• Door-in-the-Face — A large ask that gets refused makes the real, smaller ask feel reasonable by contrast. (Cialdini, 1975)
• IKEA Effect — What users build themselves, they value more. Investment during onboarding (setup, customisation) increases retention. (Norton, Mochon & Ariely, 2011)
• Endowed Progress Effect — Pre-filled progress (“step 2 of 10 already done”) measurably boosts completion versus starting from zero. (Nunes & Drèze, 2006)
• Labor Illusion — Showing work being done (“searching 400 providers…”) increases perceived value, even when the wait is artificial. (Buell & Norton, 2011)
• Effort Justification — The effort someone invests raises how much they value the outcome. (Aronson & Mills, 1959)
• Implementation Intentions — Concrete when/where plans (“book your Tuesday slot”) dramatically outperform vague intentions. (Gollwitzer, 1999)
• Fresh Start Effect — Temporal landmarks — Mondays, new months, new years — motivate people to start new behaviour. Time launches and copy around them. (Dai, Milkman & Riis, 2014)
• Reactance — Pushy demands trigger resistance. Preserving the feeling of free choice (“no pressure — cancel anytime”) disarms it. (Brehm, 1966)

Findability & trust

Whether visitors can find what they came for, and whether they believe you when they do.

• Information Scent — Users follow links the way animals follow scent: every link and label must “smell” of its destination. Weak scent means abandoned trails. (Pirolli & Card, PARC, 1999)
• Message Match — The landing page must mirror the promise of the ad or link that brought the visitor. A mismatch reads as the wrong page — or a scam. (MarketingExperiments / Unbounce)
• Progressive Disclosure — Show the essentials first and reveal complexity on demand. Collapsed sections and step-by-step forms both apply it. (Nielsen; Krug)
• Risk Reversal — Guarantees shift the risk from buyer to seller. “60-day money-back guarantee” and “cancel anytime” remove the fear that blocks the click. (Jay Abraham, direct-response marketing)
• Ambiguity Aversion — People avoid options with unknown odds. Being explicit about what happens next (“no credit card required”, “takes 2 minutes”) converts better than leaving it vague. (Ellsberg, 1961)
• Visibility of System Status — Confirm every action instantly. Silence after a click erodes trust at exactly the wrong moment. (Nielsen heuristics, 1994)

How CROAssist uses these

When CROAssist analyses your site, every test suggestion is tagged with the principle it exploits, drawn from this list. That does two things:

• It tells you why the idea should work, not just what to change — so you can judge whether the reasoning applies to your audience before spending traffic on it.
• It keeps the AI honest. CROAssist only cites documented, replicated effects. You will never see an invented label or a discredited pop-psychology claim (no “reptilian brain”, no priming folklore) attached to a suggestion.

If a suggestion’s principle doesn’t seem to fit your visitors, that is useful information too: ignore the idea and CROAssist will never propose it again.

---

Last updated: July 2026