Today's news: AI products are struggling with retention

Per The AI Journal which featured a list of studies that to no surprise pointed to high adoption/conversion, yet lower impact and retention vs. non-AI apps.

The "Novelty Problem" of exciting new capabilities that ultimately diminish past the trial phase is well publicized already...

In the context of making the commitment to build your own AI and automation (internally, or with a partner like Capto 😉). It reinforced something I've been thinking about: this is one of the strongest arguments for building AI around your own workflows instead of hoping a generic product becomes part of them.

𝐴 𝑝𝑟𝑜𝑑𝑢𝑐𝑡 𝑡ℎ𝑎𝑡 𝑖𝑠 𝑢𝑠𝑒𝑓𝑢𝑙 𝑏𝑢𝑡 𝑛𝑜𝑡 𝑎 𝑝𝑎𝑟𝑡 𝑜𝑓 𝑦𝑜𝑢𝑟 𝑤𝑜𝑟𝑘𝑓𝑙𝑜𝑤? Highly unlikely when you're straight-up building something to fit your processes.

𝐶𝑜𝑛𝑠𝑖𝑠𝑡𝑒𝑛𝑐𝑦 𝑡ℎ𝑟𝑜𝑢𝑔ℎ 𝑐ℎ𝑎𝑛𝑔𝑒? The impact is less when you're the one in complete control.

Another factor the article didn't touch on is the constant re-evaluation of subscription software and the disruption/distraction this breeds. Every renewal becomes another opportunity to ask, "Are we still getting enough value from this?". Shoutout to my related post last week about Microsoft Copilot licensing: https://lnkd.in/ecAXuYgp

A well-built AI and automation continues creating value for as long as the underlying business process exists. Even if the ROI isn't quite what you hoped, you've still learned something about where to invest next rather than stuck overanalyzing the need to cancel another subscription.

TLDR - just another list of reasons why Capto is increasingly bullish on building AI that fits existing workflows rather than asking people to change how they work to fit another product.

Original article: https://lnkd.in/ehuXSQUp

—

Originally posted on LinkedIn:

https://www.linkedin.com/posts/petersteube_tir-today-i-read-about-how-ai-products-activity-7487696254567747584-eCQO