Aug 9 – Sep 8, 2026 · 6 items
The AI week, for a marketing manager
A real edition, written for: Marketing Manager
The one thing
The paid platforms you buy through are increasingly making the budget decisions, and their AI optimises for the seller — keep a human on the money.
Gartner: most ad spend will flow through AI-run buying platforms by 2028
Gartner expects most advertising money to move through self-serve platforms where AI shapes buying, costs and outcomes by 2028. Eric Schmitt, a Gartner analyst, warns marketers against handing those systems too much control. He argues the platforms' AI serves the seller, not the buyer. Recommendations often amount to advising advertisers to spend more. Schmitt says a human should stay in the loop on budget decisions. He notes the platforms still do not measure or work well across each other. His advice is to cut the number of variables and focus on the largest platforms. He also suggests bringing finance colleagues into the assessment. Independent measurement providers should check campaign results rather than platform reporting. Junior staff can operate the consoles, but experienced oversight remains necessary.
Why it matters for you
You buy paid across the same platforms this describes. Their optimisation advice defaults to "spend more", not to your margin.
- The warning is to keep a human on budget decisions, not on creative tweaks
- Cross-platform measurement is still weak, so platform-reported results flatter each channel
- Cutting the number of variables you test at once makes the AI's recommendations legible
Try this
List every automated bidding or budget rule running in your paid accounts, and who last approved it.
Paste this into your AI tool
I run paid campaigns for a retail brand across [name the platforms you use]. Here are the automated bidding and budget rules currently active: [paste or describe them]. For each one, tell me: what the platform optimises for, where that could diverge from my margin or return target, and what one check I should run monthly to catch it. Then list the questions I should be able to answer about each rule before I let it keep running.
OpenAI's GPT-6 Astra can operate a browser and spreadsheet directly
OpenAI released GPT-6 Astra, a frontier model built to operate computers directly. It works across browsers, spreadsheets, websites and desktop applications. It can fill forms, update CRM records, run web research and produce documents. OpenAI says this reduces the need for hand-built connectors to each business system. President Greg Brockman told a press briefing that the company is now in the AGI era. Rollout starts Thursday for enterprise customers in the Daybreak gated program. Paid ChatGPT tiers, the OpenAI API, AWS Bedrock and Microsoft Azure follow in coming days. Brockman argued buyers should compare price per completed task rather than per token. OpenAI omitted GDPval, its own benchmark for real-world occupational work. OpenAI also paused some frontier training for about two weeks after the Hugging Face incident and tightened infrastructure controls.
Why it matters for you
You already use ChatGPT for content. This version does multi-step work in the apps instead of just writing text.
- Enterprise access starts gated, so at your size expect an admin decision before it appears
- Campaign reporting pulls and competitor research are the shape of work it targets
- Paid ChatGPT tiers get it in coming days, per OpenAI
Try this
Pick one weekly campaign task you do by hand and write down its exact steps, ready to delegate.
Paste this into your AI tool
I am a marketing manager at a retail brand. Here is a task I repeat every week by hand: [describe the steps, e.g. pulling last week's email and paid numbers into a summary]. Rewrite it as a numbered instruction set precise enough that someone unfamiliar with my accounts could follow it exactly. Flag every step where a judgement call is needed and say what the decision rule should be.
Grok Bot: assistants that log into your tools and finish the task
x.ai launched Grok Bot, a beta service where AI assistants run on their own cloud computer, sign into everyday work tools and carry tasks through to completion, checking back only when approval is needed. People message a bot like a colleague, and can show it a task once so it repeats the steps later. It is available now to three paid subscription tiers on desktop and iOS, with enterprise access by waitlist.
Why it matters for you
Repetitive campaign admin is the target here. It learns a routine by watching you do it once.
- Enterprise access is waitlisted, so this is a note for your next tooling conversation, not this week
- It signs into apps as a user would, which is a permissions question for whoever owns your accounts
- xAI says heaviest use is outside engineering, which includes marketing ops
Try this
Ask whoever owns your martech logins how they would handle a bot with its own account credentials.
Sources
Salesforce repackaged sales and service into three AI-inclusive editions
Salesforce replaced its sales and service packaging with three new editions. Core, Advanced and Max now apply to Agentforce Sales, Service and Industries. Each tier bundles AI agents, Slack, embedded analytics, data security and the Premier Success Plan. Core lists at $195 per user per month and Advanced at $395. Max stays at $550, the same price as the earlier Agentforce 1 Edition. Existing Agentforce 1 customers can move to Max at no extra cost. Each tier carries a set allowance of Flex Credits for running agents. Industry edition pricing takes effect later in the fall. Pricing on legacy editions stays unchanged for current customers. Salesforce says Max editions will later include Headless 360 capacity.
Why it matters for you
If your brand runs on Salesforce, campaign and service tooling now arrives bundled. That reshapes any renewal conversation.
- Each tier carries a fixed allowance of credits for running agents, so usage becomes a budget line
- Legacy edition pricing stays as-is for current customers, which removes the pressure to move fast
- At your size the useful move is a question for whoever owns the contract
Try this
Ask your CRM owner which edition your renewal maps to and what the agent credit allowance covers.
Gartner: only a fifth of large firms have scaled AI beyond one team
Gartner found that only 22% of large organisations have scaled AI across several business units. It surveyed more than 1,300 leaders at firms with over $50 million in revenue, between January and April. Spending plans are undented, with 85% of technology leaders raising AI budgets next year. About 11% of respondents could not say what they spent on AI in 2025. Gartner's Tina Nunno warned that weak measurement tied to business outcomes wastes resources. Firms that track returns continuously and shut down weak projects reported gains on 81% of initiatives. Popular uses such as cybersecurity, threat detection and IT service desk automation often return less. The best returns came from IT asset and cost optimisation, synthetic data generation, and automated code generation. Separate reports from Infosys and Deloitte found similar gaps in measurement and readiness.
Why it matters for you
You are asked to produce content faster with AI. This gives you the language to show it worked.
- Firms that tracked returns and killed weak projects reported gains on most initiatives
- A notable share of respondents could not say what they spent on AI last year
- Automated code generation and cost optimisation returned more than the fashionable use cases
Try this
Define one measurable baseline for your content output before you scale any AI use further.
Paste this into your AI tool
I am a marketing manager at a retail brand using AI to speed up content production for email, social and paid. Help me build a simple before-and-after measurement for it. Ask nothing; instead propose three metrics I could plausibly already track [I will tell you which I have], say exactly how to capture a baseline for each in one week, and name the one metric that would most convince a sceptical finance colleague. Keep it to things measurable without new tooling.
Meta cancelled its second AI-driven layoff round after agents underperformed
Meta cancelled the second wave of layoffs planned under an internal reorganisation that would have shifted much of the daily work of thousands of employees onto AI systems overseen by small expert teams. The company still cut about a tenth of its staff in May, but internal measures showed autonomous AI agents were not delivering the expected productivity, while technical and security incidents rose. Zuckerberg told staff in July that the pace of AI agent progress had been misjudged.
Why it matters for you
You are being pushed to do more with AI. This is evidence for pacing that, not resisting it.
- Output volume rose sharply while incidents and fix times rose faster
- Staff sentiment fell alongside the change, which is a management signal not a tooling one
- Useful counterweight when someone proposes replacing a process rather than assisting it
Try this
Name the one campaign step you would not automate yet, and write down why in two lines.
Sources
Build this
Every week, one small thing to build with AI in something you actually care about. No work in it. Five minutes to set up, and worth keeping if it earns a second run.
A weekend plan for somewhere you actually want to go, stress-tested until only the parts that survive remain.
Five minutes to set up
I have two days in [place] in [rough month]. Plan the two days hour by hour, with specific named places and travel between them. Be concrete, not general.
- Reply: "Now attack this plan. List everything that will realistically go wrong — timings, closures, weather, queues."
- Then say: "Rewrite the plan keeping only what survived your own attack. Cut anything you could not defend."
- Finally: "You cut one thing I might actually want. Argue the strongest case for putting it back."
- Keep the three follow-up lines somewhere reusable — the sequence works on any trip, not just this one.
Yours arrives Thursday.
This one was written for a marketing manager. Tell us what you do and the next one is written for you — same news, your job, once a week.