CDP Institute

There’s probably no limit to Amazon’s greed – because capitalism – but there are constraints as governments regulate BigTech more aggressively and, just maybe, the free market works its wonders. Amazon cited “seller sentiment related to the fee” in its statement pulling back from its planned 2% fee on sellers who don’t use its Prime fulfillment services.  But looming antitrust cases likely had something to do with it.

Oracle held its CloudWorld conference last week.  There weren’t any CDP-specific announcements but they did announce their next-generation Fusion Data Intelligence Platform, which includes comprehensive data models, prescriptive AI models, interactive analytics, and applications leveraging all the above.  They also added guided campaigns for marketing and sales and pretrained large language models for marketing, sales, and service applications.

Fear of losing your job to AI is never far from the surface.  This study from job site Indeed puts marketing in the top 20% (9th of 47) of jobs that generative AI can most easily perform. It may seem odd that jobs like driving, heathcare, and manufacturing are deemed least at risk, but remember this is specifically about generative AI, not all kinds of automation.

Salesforce is buying Airkit.ai, which builds low-code, AI-powered e-commerce customer service agents.  You might have thought Salesforce could already do that, but presumably Airkit does it better.  It doesn’t hurt that the company founders sold a previous business to Salesforce, the company was partly funded by Salesforce Ventures, and the product is already listed on AppExchange.  Price was not announced.

Microsoft Advertising has expanded its search ad buying platform to include video and connected TV campaigns.  The integration will let advertisers take advantage of search data to target the video and CTV audiences.  They also announced that Microsoft Monetize, the ad sales platform that was formerly AT&T Xandr, will make the ID5 universal ID available to its publishing clients, expanding the addressability of their cookieless inventory.

Forget right message to the right person at the right time.  The real Holy Grail of marketing is tracking behavior after those messages are delivered.  Just-launched Channel99 uses a combination of website tags and ad pixels to measure exactly that, at least for B2B accounts.  Woot.

We’re not quite ready to party like it’s 2019 but there was a touch of déjà vu yesterday when Klaviyo’s share price soared 23% after its initial public offering, which had already been repriced above its target.  The offering gave Klaviyo a valuation of $9.2 billion or roughly 20x its 2022 revenue.

Marigold is a mash-up of martech stalwarts including Campaign Monitor, Cheetah Digital, Emma, Liveclicker, Sailthru, Selligent and Vuture.  They just launched a Relationship Marketing Solution that combines the components into an endless buffet of hyper-personalized marketing goodness.  If you’re wondering how closely the products are integrated, well, that’s a very good question.

Google announced general availability of its automatically created assets for search ads, which creates tailored headlines and descriptions based on the ad context.  They’ve started a beta test of a conversational interface that uses AI to build search campaigns based on little more than the advertiser’s website URL.  Users are still allowed to modify and approve the AI-generated “suggestions”.

Personalization engine Movable Ink has dipped its toe into CDP waters with Universal Data Activation, an option that assembles data from all sources and makes it available for personalization.  No mention of identity matching or sharing the profiles with other systems, so it’s probably not a proper CDP.  But they are scratching the same itch.

The United Kingdom’s Competition and Marketing Authority (CMA) has been pondering generative AI since back in May.  They just released their findings in the form of a 129 page report that covers the topic from a definition of foundational models through how they work through their impact on consumers.  They finish with seven proposed principles to guide model development and a promise for further discussions.

Most surveys about generative AI report that managers are enamored of the concept and some describe how it’s being used.  I rarely share them because you already know the answers.  But this Harvard Business Review paper actually measured how much gen AI improved the output of professional knowledge workers.  Quite a bit, it turns out.