🎓 When ‘AI-powered’ makes products better, or not

Up to 24% more people prefer a smart product (e.g. fitness tracker) that uses adaptive AI algorithms (vs. a product that doesn’t), but not when the product is simple (e.g. electric toothbrush).                                                                                                                                                                                                                                                                                                                                                                                                                 March 26, 2024 | [Read Online]( When ‘AI-powered’ makes products better, or not Up to 24% more people prefer a smart product (e.g. fitness tracker) that uses adaptive AI algorithms (vs. a product that doesn’t), but not when the product is simple (e.g. electric toothbrush). [Thomas McKinlay]( [fb]( [tw]( [in]( [email](mailto:?subject=Post%20from%20Ariyh&body=When%20%E2%80%98AI-powered%E2%80%99%20makes%20products%20better%2C%20or%20not%3A%20Up%20to%2024%25%20more%20people%20prefer%20a%20smart%20product%20%28e.g.%20fitness%20tracker%29%20that%20uses%20adaptive%20AI%20algorithms%20%28vs.%20a%20product%20that%20doesn%E2%80%99t%29%2C%20but%20not%20when%20the%20product%20is%20simple%20%28e.g.%20electric%20toothbrush%29.%0A%0Ahttps%3A%2F%2Ftips.ariyh.com%2Fp%2Fwhen-ai-powered-makes-products-better) New to [Ariyh](? This is a 3min practical summary of a scientific study 🎓 Join 26,757 marketers who use science, not flawed opinions 📈 [Subscribe here]( Planable is the content collaboration platform that makes marketing teamwork a breeze. Create, plan, review, and approve all your social media, blogs, newsletters, press releases, briefs (you name it!) — all in one place! Then, once you hit publish, simply [track your analytics]( in the same place. [Try Planable for free]( Want to sponsor Ariyh? [Here’s all you need to know](. 📝 Intro Your company is developing the new flagship model of your ultra-secure safe. Your brand has a decades-old reputation for selling ‘unbreakable’ safes. One of your suppliers is proposing you use their latest adaptive AI lock mechanism. The lock will make it faster for the owner to open and lock the safe. Your product manager agrees it could “take things to the next level”. You wonder what you should do. Would AI make your product more valuable, or not? Here’s what the latest scientific research has found. P.S.: When describing your products make sure to keep your descriptions simple. Too technical or difficult-to-understand [language can reduce sales by up to 16%](. Want to access hundreds more insights like these? [Explore Ariyh insights here](. People prefer adaptive AI algorithms in smart products that have many uses vs a simple use Topics: Product For: B2C. Can be tested for B2B Research date: November 2023 Universities: WU, University of Lucerne, University of St. Gallen, and Columbia University 📈 Recommendation When creating or marketing technical (e.g. Adobe Photoshop) and smart products (e.g. Alexa home assistant) include and highlight their AI-powered qualities to make them more desirable. Do not rely on high adaptivity and learning for products with simple functions (e.g. automatic light switch, coffee machine), it will backfire. People will be more likely to choose your product. 🎓 Findings - People are more likely to choose a product because it has an AI-backed adaptive algorithm when the product performs complex tasks (e.g. spatial sound system, smartwatch), but not when it performs simple tasks (e.g. alarm clock). - As part of 4 experiments, researchers found that: - 66.3% (vs. 42.6%) of people chose an adaptive (vs. non-adaptive) cooking recipe generator and considered it 47.8% more creative - People were 16.05% more likely to want to use a highly adaptive smart voice assistant (vs. non-adaptive) - The more complex the task was (eg. creating a recipe versus generating a random number), the more likely people were to prefer high adaptivity. - The effect reverses when products have a simple or straightforward function (e.g. bicycle lock, smoke detector). 🧠Why it works - We consider adaptable algorithms to be more creative. - In certain products, a higher [level of creativity]( is a desirable characteristic (e.g. we expect voice assistants to [provide creative, new answers to a repeated request](). - So we prefer smart products with smart, adaptable features. - However, smart algorithms and AI can generate unpredictable results. - So for products that are meant to do a simple task (e.g. locking and unlocking a door), we [trust]( and prefer simple, pre-programmed settings. ⚖️ Get expert feedback to support your marketing decisions - [try it for $0]( Our brains aren’t well tuned for judging our own work. So how can you be sure you are making the best marketing decisions? Simple 👉[Â]([SenseCheck it](: get pragmatic, confidential feedback on any work eg websites, ads, designs etc from 5 of our huge panel of marketing experts. Back your judgment and confirm effectiveness with an informed second opinion from outside your bubble. Curious?[Get your first feedback report for FREE](. Or [join the panel](. [Get your first feedback for free]( This announcement was sponsored. Want your brand here? [Click here](. ✋ Limitations - The research did not consider other technical factors of the AI algorithm (e.g. how well it actually worked, what data was used to train them, the privacy conditions that come with the product), but only whether a product claimed to use AI or not. - The study compared low versus high adaptivity algorithm products to each other, but not to a human (e.g. an AI shopping assistant vs a real person assistant in-store). It is unclear how these would compare. 🏢 Companies using this - Over 50% of Fortune 500 companies that sell tech products refer to algorithms or AI in their online presence. - 200 people rated their preferences of high adaptivity AI in 55 products and found that items: - Ideal for AI include voice assistants, smart home systems, and financial advisor apps - Not ideal for AI include bicycle locks, coffee machines, and air purifiers - Many tech companies correctly include adaptive algorithms and AI in their products, for example: - Oportun app uses AI to analyze spending and automatically transfer small amounts to the user’s savings - Citizen’s smart-watch uses AI to learn about the wearer's habits and increase personalization - But some have likely overengineered adaptive algorithms into simple products, for example: - Oral-B’s AI-powered toothbrush sends signals to an app when applying too much pressure or not brushing long enough in a certain spot - Samsung’s smart fridges that adapt to daily routines and allow users to check the weather and stream videos Adobe PhotoShop, a complex software, correctly highlights its AI-powered generative fill and expand functions. ⚡ Steps to implement - If you are building smart or technical products (e.g. complex software) consider powering them with AI-driven algorithms that are capable of learning. - When marketing such products highlight their creativity and adaptability (e.g. “the daily playlist software will select songs according to your agenda for that day”) - If the product was developed in partnership with a university, highlight that - people [will be willing to pay more for it](. - However, if you are creating or selling simple, binary function products (e.g. on/off electric toothbrush), refrain from using AI. - Focus on explaining how these products will perform a task predictably and reliably (e.g. “turn on with a single push of a button”). 🔍 Study type Lab and online experiments 📖 Research [Unveiling the Mind of the Machine](. Journal of Consumer Research (November 2023) 🏫 Researchers - [Melanie Clegg](. Vienna University of Economics and Business - [Reto Hofstetter](. Institute of Marketing and Analytics, University of Lucerne - [Emanuel de Bellis](. Institute of Behavioral Science and Technology, University of St. Gallen - [Bernd H. Schmitt.]( Columbia Business School, Columbia University Remember: This is a new scientific discovery. In the future it will probably be better understood and could even be proven wrong (that’s [how science works](). It may also not be generalizable to your situation. If it’s a risky change, always test it on a small scale before rolling it out widely. Rate today’s insight to help me make Ariyh's next insights 🎓 even more useful 📈 How was today’s insight? [Loved it]( | [Great]( | [Good]( | [Meh]( | [Bad]( Here’s how else Ariyh can help you: - 📈 Access 100s of insights and learn how other evidence-based marketers apply them, with [Ariyh Pro]( - 📘 Supercharge your business with Ariyh’s [Playbook of Pricing & Promotions]( or [Playbook of Ecommerce]( - 🎓 New to Ariyh? If this was forwarded to you can subscribe below for $0 [Subscribe here]( [fb]( [tw]( [ig]( [in]( Update your email preferences or unsubscribe [here]( © 2024 Ariyh Calle Bailen, 11 Barcelona, Barcelona 08010, Spain [[beehiiv logo]Powered by beehiiv](