CDP Institute

Marketers may be using generative AI primarily to generate content, but its potential to speed data integration is at least as important.  Case in point: Informatica has just launched CLAIRE GPT, which lets users build data management processes with a natural language interface.  They’ve also added closer integration with Microsoft Azure and an expanded program for developers to integrate with Informatica tools.

It may or may not be “helpful and natural” to place advertising within a chat flow, as Microsoft promises, but it’s surely inevitable.  Citing experience with the chat-enabled Bing search engine, they’ve released a new API that lets web publishers, app developers, and online services deliver ads within a chat platform from Microsoft or someone else. If details were provided, I missed them.

Whether “Amazon Anywhere” sounds like your sweetest dream or worst nightmare, it’s now a (virtual) reality.  The new offering will let people buy physical goods – you know, t-shirts and stuff – within video games, virtual worlds, and mobile apps.  Once you select something, the experience will match the familiar Amazon store with product information, pricing, availability, delivery dates and the rest.

Apart from some pesky AI gurus who think it might destroy humanity, pretty much everyone else is excited to see what generative AI tools like ChatGPT do next.  This Botco survey finds that 73% of marketers are already using it to create some type of content, and 66% of users see a positive return on investment.  Training, cost and privacy concerns are the biggest roadblocks.

AI should eventually help IT teams manage data more smoothly, but right now they are one pack of unhappy campers: as this report from StreamSets puts it, “69% are frustrated non-data experts think you can click a button and data magically appears” and “59% say line of business owners are ‘blissfully unaware’ of how hard it is for IT to deliver data”.  In a companion report, 95% say line of business teams creating their own datasets creates problems including inaccurate data, non-compliant use, data breaches, and new data silos.

The data managers interviewed for Monte Carlo weren’t quite so self-pitying, but they did see lots of problems: downtime lost to data problems is up 86% year over year, time to resolve data incidents rose 66%, and – insult to injury – 74% said most data problems are identified first by business stakeholders, up from 47% last year.  The size of the task is daunting: 27% of companies manage more than 100 tables in their data warehouse, lake, or lakehouse.