Thomas McKinlay’s Post

🎓 New research: people were 12% to 18% more likely to say they’d buy when an AI recommendation had an explanation (vs not) 👇 🔬 Scientists ran 4 experiments and found that when given an explanation (vs no explanation) of how the AI algorithm worked, people were: - 12.5% to 17.6% more likely to say they would buy a pair of sneakers - 54.6% more likely to click "See more details" for recommended t-shirts 🧠 Why? - We don’t always understand how AI works - This makes us skeptical of its choices - When the recommendation is explained we trust it more - And become more willing to follow it 📈 So always try to explain your algorithm or AI-based recommendations (e.g. “Popular with other people living in your area”). Full 3min summary of the study (with limitations): https://lnkd.in/dUuuaE5e 🎓 Want more science-based practical insights? - Follow me - Hit the bell icon 🔔 - Subscribe to Ariyh, my newsletter

🎓 New scientific research: Plural brand names work better than singular ones. 🔬 Scientists ran 4 experiments and analyzed 12,885 brand and app names and found that: - Plural names for large brands (e.g. Snickers) had 12.3% better brand attitudes - People were 13.4% more likely to prefer the name ‘Fine Secrets’ (plural) compared to ‘Fine Secret’ (singular) - 6.2% more likely to prefer the name Urban Elves vs Urban Elf for a delivery app 🧠 Why? - Plural words give the perception of a coherent, larger group - This group feels like it’s more than just the sum of its parts - So we form a better opinion of it 📈 So if you’re naming a brand, use a plural form People will prefer it and have better attitudes ✋ Be careful: the effect works best for mainstream brands, but disappears for premium brands 3min practical summary (with limitations) of this research: https://lnkd.in/dQ-GFrD2 🎓 Want more science-based practical insights? - Follow me - Hit the bell icon 🔔 - Subscribe to Ariyh, my newsletter

🎓 People hate to wait. Refund them instantly. It made them 12% more likely to intend to repurchase 👇 📈 Give an immediate refund as soon as customers ship a return. Customers will be happier and more likely to buy again. And return rates stay the same (they don't worsen). 🔬 In online experiments, researchers found that instant refunds: – Felt 135% faster VS the traditional 7-14 day processing cycle. – Led to a 12.2% higher intention to purchase again. – Increased product return satisfaction by 12.4%. 🧠 Why it works – The delay between a return and a refund makes us anxious – Instant refunds alleviate this anxiety – And feel more fair and efficient – So they improve the overall customer experience 👉 So avoid making customers wait, unless they’re a high fraud risk. 🎓 Want more science-based practical insights? - Follow me - Hit the bell icon 🔔 - Subscribe to Ariyh, my newsletter

🎓 New scientific research (8 experiments) Rounded CTA buttons boost conversions 👇 🔬 Scientists found that when using rounded (vs sharp-edged) CTA buttons: -> A landing page for an event planning tool had a 55.5% higher CTR -> A restaurant ad received a 24.6% higher CTR -> People were 16.8% more likely to click to order food online ->Students were 46% more likely to click a button to take an extra survey. 🧠 Why? - We associate roundness with friendliness - But sharp angles with threats and toughness - So we prefer round shapes - And are more likely to approach them 📈 So use curved edges (instead of straight angles) if you want to increase CTA clicks. ✋ Note: if the CTA calls for a negative action (e.g. unsubscribe), the difference between button shapes disappears, they both perform the same. 🎓 Want more science-based practical insights? - Follow me - Hit the bell icon 🔔 - Subscribe to Ariyh, my newsletter Full 3min practical summary: https://lnkd.in/dfwSmg5R Edit: There was a typo in one of the reported experiments' results. The difference is 11.18% vs 7.19% (= 55.5% difference) and not 11.8% vs 7.19% (= 64.1% difference) as initially reported. The text above has been corrected (but I'm unable to change the image of this post, the 64% should be 55%). Thank you for your understanding.

🎓 New scientific research: Be careful, referrals can backfire 👇 🔬 Scientists ran 7 experiments. When people recommended a product to a friend, but the friend did not buy it, they were: - 15.7 times more likely to change their initial choice of a pen - 4.3 times more likely to change their initial choice of a virtual roller-coaster ride - 25.5% less likely to say they would buy a travel mug again 🧠 Why? - We think our opinions are sound - So we give recommendations - But if they are rejected, we question our knowledge and lose confidence - So we change our original choice 👉 So be careful when encouraging recommendations Ask customers to recommend you only to friends who would likely choose you (e.g. “Tell a friend who also likes [what you offer]”) 📉 Otherwise, you risk losing your customer 🎓 Want more science-based practical insights? - Follow me - Hit the bell icon 🔔 - Subscribe to Ariyh, my newsletter

🎓 Should you use AI in your products? Here’s when science says you should (or not) 👇 🔬 Scientists discovered that it depends on what your product does: 👉 Does it perform several, complex tasks? (e.g. track your heart rate, your running, etc.) 📈 Then AI-powered adaptive algorithms will make it MORE attractive 👉 Does it perform a straightforward task? (tracking the time, making coffee) 📉 Then AI will backfire and make it LESS attractive 🧠 Why? - Adaptable algorithms seem more creative - In smart complex products, creativity is desirable - However, creativity can be unpredictable - So for simple task products (e.g. locking and unlocking a door) we prefer simple settings 👉 So don’t blindly follow the hype of ‘AI makes it better’. Sometimes it does, but often it doesn’t. 🎓 Want more science-based practical insights? - Follow me - Hit the bell icon 🔔 - Subscribe to Ariyh, my newsletter

🎓 Should influencers sound super excited? It depends how big their following is 👇🏼 🔬 Scientists analyzed 20,000+ Instagram and TikTok sponsored posts and found that a 10% increase in the excitement of the language led to: -> 5.4% higher engagement for micro-influencers -> 8.4% lower engagement for macro-influencers 🧠 Why? - Macro-influencers have a large following - So their sponsored content feels more like an ad - Exaggerated emotions feel like an attempt to sell - But we consider micro-influencers to be average people - Their high-energy content seems genuine - So we trust them more, and like their excitement 👉 So if you following ir ‘normal’ (<100k), feel free to sound SUPER-excited!! ✋ But if you’re working with a macro-influencer, encourage them to ‘tone it down’ (and focus on providing information, rather than hard selling). 📈 Your posts will be more effective. 🎓 Want more science-based practical insights? - Follow me - Hit the bell icon 🔔 - Subscribe to Ariyh, my newsletter

🎓 What works better: Customer reviews, or experts’ recommendations? 📈 Here’s when you should show one or the other, so that you can get more customers. 🔬 Scientists ran 3 experiments and found that people will prefer your product when: - It’s easy to judge its quality and you show customer reviews (e.g. movie) - It’s harder to judge its quality and you show reviews from experts (e.g. doctor) 🧠 Why? - We relate to people like ourselves - So for services that are easy to judge, we prefer experience of our peers - But when a service is hard to judge, we don’t trust reviews from fellow customers - So we rely on expert opinions instead 👉 Try to understand how easy or difficult your product is and provide the correct social proof 📈 People will be more likely to buy. 🎓 Want more science-based practical insights? - Follow me - Hit the bell icon 🔔 - Subscribe to Ariyh, my newsletter

🎓 Should your price be easy to divide? Yes, when you’re selling multipacks 👇 🔬 Scientists ran 15 experiments and found that people were: - 45% more likely to buy 6-pack tissues for $18 (vs $17) - 27% more likely to buy an 11-bag cashew multipack priced at $11(vs $10) - 20% more likely to buy a 3-pack car freshener at $9.00 (vs $8.98) 🧠 Why? - When it is easy, we calculate the price per unit - We then consider how long the individual unit, not the entire pack will last us - So we think it will run out quickly - And we justify buying the multipack 👉 When selling bundles of identical products (e.g. dish soap) price the multipacks with an easily divisible price. 📈 People will buy them, even for a higher price. 🎓 Want more science-based practical insights? - Follow me - Hit the bell icon 🔔 - Subscribe to Ariyh, my newsletter

🎓 Add a Q&A section to your product page. It can boost your ratings by +0.5 stars ⭐️ 🔬 Scientists analyzed 5,077 products. When a Q&A section was added to a product page: - Products got higher ratings - Negative reviews were less likely to focus on customer expectations not being met The effect was the strongest for products with lower average ratings. 🧠 Why? - Q&A section allows to ask questions - The extra info tells us if the product suits us - We are less likely to be disappointed 👉 So include a section where people can ask questions about the product before they buy. 📈 Customers will be happier with their purchase. 🎓 Want more science-based practical insights? - Follow me - Hit the bell icon 🔔 - Subscribe to Ariyh, my newsletter