Hi there,
We are in the middle of an AI gold rush. Marketers are deploying the most powerful technology in history, yet the majority are using it to ship the exact same generic campaigns, just faster, at higher volume, and at a lower cost.
That is not transformation. That is throughput.
Recently, Arielle from our team gave an incredible presentation at the AI for Marketers Masterclass, and she broke down a paradox that I haven't been able to shake:
While 87% of marketers now use generative AI in at least one recurring workflow, over 80% of organizations report no tangible EBIT impact from their investments [1] [2]. Adoption is nearly universal, but impact is nearly absent.
Why? Because AI is only as good as the data you feed it.
We all have access to the exact same models. The teams winning right now aren't the ones with the fanciest algorithms or the biggest budgets. They are the ones who fixed their data layer first.
Here is the uncomfortable truth about the three biggest AI marketing trends happening right now, and the one thing they all demand: you have to actually see your customer. Everywhere they buy you.
Effective personalization is impossible without quality first-party data. Over the last few years, privacy regulations like GDPR, CCPA, iOS changes, and cookie deprecation have wiped out 30% to 40% of previously trackable conversions. If you felt your attribution get shakier, that is why. [1]
The brands winning are the ones who responded by building robust first-party data strategies, recovering up to 75% of that lost signal. They didn't do this with a clever new pixel; they did it by giving customers a real reason to identify themselves.
Now, layer this on top of our omnichannel reality. For a healthy omnichannel brand today, maybe 20% to 30% of revenue comes through DTC. The other 70% to 80% is happening at Target, Amazon, Whole Foods, Sephora, or TikTok Shop. For that 70%, brands were already flying blind before the privacy cliff. [1]
Marketers who have actually unified their data are 42% more likely to respond to customers in real time, and 60% more likely to scale successfully with AI agents. [1] That is the compound difference between brands with a data foundation and those without it.
Personalized content at scale used to be a pipe dream. A creative team of three could make maybe 30 variants of a campaign. Now, a single prompt produces 300. The bottleneck has moved. It is no longer about whether we can make the creative; the bottleneck is context.
You cannot give a customer a personalized recommendation or reply if your AI doesn't actually know who they are.
Right now, 84% of marketers using AI admit they are still running generic campaigns.[1] They point generative AI at a blank canvas and get back content that is technically fine, technically on-brand, and completely interchangeable with their competitor's content. Generic AI content is just spam with better grammar.
However, AI content drafting delivers a 3.2x ROI when it is fed three specific things: brand voice, campaign context, and customer data.[3] What did this person buy? Where? When? AI content that personalizes to a real customer on a real journey outperforms generic AI content by an order of magnitude.
Predictive AI's promise is real, but it demands a closed loop. Companies that successfully deploy predictive analytics see three times higher campaign performance.[4] But that only happens when your offline purchase data gets back to the model.
If you are running Meta ads and Meta can only see your DTC purchasers, Meta's model is optimizing against 25% of your reality. It is going to keep spending against lookalikes of your DTC shoppers and completely miss the audience that is actually buying your product at Target or on TikTok Shop.
When you close the loop between retail purchase data and ad platforms, the results speak for themselves. We have seen brands achieve a 36% uplift in tracked conversions relative to ecommerce data alone, with the exact same ad spend and the exact same creative. The only thing that changed was that the platform could finally see the full picture.
Three trends. One requirement.
Every single one of these trends demands that you see your customer wherever they buy. Personalization demands it. Predictive modeling demands it. Even generative content demands it.
If you are an omnichannel brand, that context isn't sitting in your website analytics. It is hidden in the 70% of your revenue that is invisible today. The brands that win the next era of retail won't be the ones with the best attribution models. They will be the ones that build direct, owned relationships with their customers, no matter where the transaction happens. Period.
Talk soon,
Kait
A few extras:
Some of my favorite brands are hiring…
L'Oréal - Manager, Amazon Retail Media
Timberland - Director, Retail Marketing
References:
[1] Salesforce. State of Marketing, Tenth Edition. 2026.
[2] McKinsey & Company. The State of AI. 2025.
[3] Boston Consulting Group. The Future of Marketing with GenAI. June 2025.
[4] Forrester. Predictions 2026: B2C Marketing, CX, & Digital. 2026.
Kait’s free Sunday newsie on what’s next in performance marketing, AI, and omnichannel growth.
