August 21st-August 27th // Estimated Reading Time: 9 minutes
In This Edition!
Is Reddit losing its AI luster?
LLMs are becoming the new marketing “operating system”
Less Than 1 Month Left To Enroll For The AI Accelerator!
The AI Accelerator (sponsored by Blis) helps marketers level up on all things AI for marketing and advertising, like planning, creative, activation, and measurement, so they can make a meaningful impact on their business, fast.
We only have 40% capacity left! Enroll now using promo code EarlyBirdsLearnAI for 20% off!
Top Headlines🔥
What Reddit's ChatGPT citations collapse means for marketers
Source: Business Insider
August 26th, 2026
Summary: Reddit's ChatGPT citations dropped 86% in mid-August, according to new data from AI visibility firm Promptwatch; three other firms confirmed the drop. But researcher Suganthan Mohanadasan found ChatGPT is still pulling from Reddit just as often when researching answers. It's simply citing brand websites more often in the final response users see. OpenAI says it doesn't target a fixed visibility level for any site, and Reddit maintains it's still cited regularly.
Opinion: If your AI visibility strategy has been centered around Reddit content, rest assured, AI is still crawling Reddit and answering users accordingly. Citations going down means fewer users click out to Reddit and go deeper, which is not ideal, but if citations are just declining and / or shifting to brand websites, there isn’t much to be concerned about.
The bigger picture question is how marketers should be thinking about the constantly shifting nature of AI models, their sources, and consequently, AI visibility. Getting too anchored to any one AI visibility content strategy is risky. Using an AI visibility tool that not only gives you a good read on how your brand shows up in AI, but also is regularly monitoring how your brand showing up in the AI is changing, and why it’s changing, is critical.
Salesforce, Anthropic expand partnership amid ‘SaaSpocalypse’ concerns
Source: CNBC
August 26th, 2026
Summary: Salesforce and Anthropic unveiled "Claudeforce" on Wednesday, a Claude plug-in with 37 pre-built sales skills that lets people use Salesforce data directly inside Claude to do things like compose emails, update Salesforce records, run deal reviews, and more, without ever opening the Salesforce app. It's the first time Salesforce has ever put its "force" suffix on someone else's product and brand. The news came at the same time as a Q2 earnings beat that sent shares up 12% after-hours, even with the stock still down about 22% YTD due to AI disruption concerns.
Opinion: Giving "force" to Anthropic is quite the concession. It’s Marc Benioff admitting that people no longer want to work in the Salesforce UI (earth to Marc, no one ever wanted to work in the Salesforce UI anyways), they’d rather be in Claude, and Salesforce is now fighting to retain the backend. But what is the long-term value of that backend? Claude, and the other major AI chat platforms, continue to become an “operating system” of sorts for marketers.
New Products & Features 🚀
What It Does: Kokai is the name for The Trade Desk’s current buying interface. “Zuma” is the biggest update to Kokai since its rocky 2023 launch. This release focuses on UX improvements🔒, such as easier troubleshooting, bulk edits that forecast impact before going live, and lift studies that can be set up with one click and go live within 48 hours. Koa, the platform's AI assistant, also launched a closed beta with six specialized AI agents for tasks like campaign creation and audience building. The Trade Desk is opening those Koa agents to outside AI tools, including Claude.
Quick Take: The Trade Desk had already let external agents in🔒 some time ago (another example of the AI chat platforms becoming marketing “operating systems”). Zuma now gives it native AI capabilities. This is The Trade Desk playing catch-up in AI. Better late than never. It’s still early innings and no AI-disruptor has meaningfully moved the market yet. Now that The Trade Desk has “caught up”, can it make the next move to “get ahead”?
What It Does: Marketing operations startup Gradial can create entirely new, on-brand creative assets at scale by remixing a brand's existing approved assets instead of building from scratch. The AI pulls backgrounds, copy, logos, and product imagery from tools like Figma and SharePoint to produce roughly 1,500 asset variations per campaign, across channels, languages, and promo codes. Gradial charges based on marketing output, not token use.
Quick Take: Cheap AI generation has created a downstream burden on brand and review teams that have to clean up the output. Gradial's bet is that remixing only approved assets keeps output on-brand at scale, without the AI slop.
What It Does: Influencer, a creator marketing agency, and Profound, an AI visibility platform, have partnered to help brands shape how they're recommended by AI tools like ChatGPT by leveraging creator content. Profound identifies how LLMs are surfacing brands and their competitors. Influencer then builds creator campaigns around those findings to help those brands optimize their AI visibility.
Quick Take: This is the second creator-focused AEO (answer engine optimization) offering🔒 we’ve heard about this year, after a similar product from Later was launched in May. It's worth watching whether AI models are able to, or even care to, filter out thinly veiled paid content from their crawlers the way people do.
What It Does: X Ads MCP connects LLM agents like Claude and ChatGPT to 23 X Ads tools covering campaign creation, analytics, and management. Advertisers can build campaigns with plain-language prompts inside the LLM they already use rather than X's native dashboard.
Quick Take: X joins platforms like Meta, Pinterest, TikTok, and Snapchat in launching MCP integrations over the past year. We’ll say it one more time: AI chat platforms are becoming the marketing “operating system”!
AI Use Case of The Week💡
Inside Groupon's AI Max bet on hyper-local search
The Setup: Groupon was under financial pressure, spread too thin across local service vouchers, travel, and physical goods. Groupon narrowed its focus to local services and had to solve a specific problem: Local search queries are too varied and unpredictable for any manual keyword list to cover, so high-intent searches were falling through the cracks. For example, a consumer search for a specific stylist in a specific city would default to a generic "hair salon" match and send the user to an irrelevant national deals page.
The AI Solution: Groupon became an early beta tester of Google's AI Max for Search to gauge intent behind queries instead of matching them to a fixed keyword list. Using AI Max, Groupon could parse hyper-local searches down to the service, provider name, neighborhood, and state, then serve a relevant local ad. Groupon tested it first in contained verticals like Health, Beauty & Wellness, and Things To Do before rolling it out further.
The Results:
• Groupon’s strategic shift to local and AI helped it grow billings and revenue for the first time in a decade
• Groupon has automated millions of local search queries using AI, eliminating manual keyword-building work
• An AI agent caught a data attribution error missed by Groupon’s standard reporting, flagging a revenue-losing issue early
Why This Matters: AI amplified a focused strategy, but it wouldn't have saved an unfocused one.
Your Action: Where does your keyword strategy fail? If you use Google Ads, export your Search Terms report for one vertical over the past 60-90 days. Feed the converting queries to an AI tool like ChatGPT or Claude, asking it to cluster them by what the user actually wanted rather than which keyword was matched. Compare that grouping to your current keyword list. A cluster with real conversion volume but no dedicated keyword is an opportunity for testing an intent-matching tool like AI Max.
Other Notable Headlines📌
OpenAI tests exclusion targeting for ChatGPT ads🔒 - OpenAI is letting a small group of advertisers exclude contexts they don't want their ads to appear next to. Studies found significant ad mismatches on the platform, including one that found only 38% of ads were directly relevant to a ChatGPT conversation.
source: SE Ranking
USA Today reformats content to attract AI licensing deals🔒 - The publisher is restructuring its content into AI-readable formats to improve discoverability in AI and strengthen its content licensing pitch.
Not all human-in-the-loop steps are equal, says Scott Brinker - His core argument: The highest-leverage use of human judgment is often shaping a system's rules upfront, not reviewing every output it produces.
Nobody wants Anthropic's best AI model anymore now that there are way cheaper alternatives - Spending data from Ramp shows companies reserving Fable 5, Anthropic's priciest model, for complex tasks while routing everyday work to cheaper models.
That’s It For This Week 👋
The U of Digital AI Edge Newsletter is intended for AI Edge Pro subscribers, but occasional forwarding is okay!
To subscribe visit Uof.Digital/Newsletters/ai-edge or AI Edge Pro, or contact us directly for group subscriptions.
And remember, U of Digital helps teams drive better outcomes through structured education on critical topics in marketing / advertising like AI, mar tech, programmatic, CTV, commerce media, and more. Interested in learning more about how we can supercharge your team? Let’s talk!
Thanks for reading!
