The Setup: Bonterra makes software for nonprofits and charitable foundations. Its marketing team responds to reviews on software review sites like G2, TrustRadius, and Gartner, where business buyers rate the tools they use. Replying to reviews helps the team catch product feedback. But responding to each one was time-consuming, especially replying to hundreds of 5-star reviews. Each reply also had to follow Bonterra’s detailed brand guidelines.

The AI Solution: The team built a workflow with two AI agents in Jasper, an AI platform that marketing teams use to create on-brand content. Agent 1 sorts: It reads each review and labels it by sentiment, star rating, and type, such as praise, constructive criticism, or a 1-star complaint that needs a human response. Agent 2 replies based on the labels. It greets reviewers by name, adjusts tone based on the rating, points unhappy customers to helpful resources, and sends support issues to the right team. It follows Bonterra’s approved terms, such as using the term “solution” instead of “tool.”

The Results:

• A quarter’s worth of review replies now takes under 2 hours, down from 12+ hours, an 83% decrease

• Monthly batches of about 40 reviews receive replies in under 2 minutes

• Replies stay on-brand 

• The team spends the time saved on case studies and other marketing work

Why This Matters: The pattern goes beyond reviews: Sort first, then act. The same setup works for social comments, survey responses, and inbound emails. Any marketing task with high volume and a few predictable categories is a good fit for this kind of workflow.

Your Action: Recreate Bonterra’s workflow on a high-volume workstream of your own, such as social comments, survey responses, reviews, or inbound requests. Paste 50 into AI and ask: “Sort these into no more than five groups based on what each one needs. Name and describe the characteristics of each group and flag which ones need a human.” Then build one specialist agent per group, each with its own instructions, examples, and the brand rules that matter for that group. Run each group through its specialist agent. Then run all 50 through a single catch-all prompt and compare. If the specialists clearly win, split up the task. If the catch-all does fine, one good agent will do.

Anthropic signs its first news licensing deal, with the AP🔒 - Claude showed real-time AP election updates on midterm Election Night, attributed to the AP. Anthropic had avoided these kinds of deals until now, while rival OpenAI has signed more than a dozen.

OpenAI taps Magnite and PubMatic to make ChatGPT ads easier to buy at scale - Advertisers can set budgets, bid, and upload ad creative for ChatGPT through the SSPs. OpenAI continues to try to bring in ad demand from as many partners as possible.

OpenAI is making a major change to ChatGPT interface - OpenAI will combine ChatGPT’s Chat and Work tabs in a limited test. Everyday questions and longer tasks, like research and building documents, will happen in one place.

Product feeds are becoming the new ad creative in Google AI Mode - Google’s AI Mode groups products into shopping categories and appears to write the descriptions itself, using advertisers’ product data.

Chord buys its rival Zenlytic to build an agentic version of a CDP - Chord's AI agents help ecommerce brands organize and act on their customer data. It paid $5M for Zenlytic, which has an AI data analyst agent named Zoë that will largely replace Chord’s agent. Zenlytic has bigger enterprise clients such as Verizon.

Adobe report: AI traffic now generates 83% higher revenue per visit for U.S. retailers heading into the holiday shopping season - In September, shoppers who clicked through from AI tools converted 90% better than visitors from sources like email, paid search, and social.

The Trade Desk and Dstillery link AI agents to reshape audiences mid-campaign🔒 - The Trade Desk's conversational AI platform, Ask Koa, is connecting to Dstillery’s agentic platform, DS-1. Their agents can now work together to build custom audiences mid-campaign in minutes, down from about four days.

How a Democratic Congress could reshape AI ad innovation - After their midterm wins, Democrats are expected to bring more scrutiny to AI advertising, privacy, and data centers.

A brutal post-midterms sell-off for AI stocks sends Wall Street reeling, and the S&P 500 tumbles 4.4% - Stocks tied to AI data centers led the market’s worst day since April 2025, amid worries that Democratic wins could lead to pauses on new data center construction. The Nasdaq fell 5.5%.

Thanks for reading!