Your AI spend might be optimizing the wrong half of the funnel, and this issue gives you the benchmarks, search data, and ABM plays to fix it.
This week's edition covers:
- Kevin's Take: you bought AI to speed up work that was already fast, here's the spend you should be rethinking
- The Signal: Signal: fresh benchmarks on where CMO AI budgets are actually going, plus how B2B buyers really use AI search
- Tools & Tactics: a 10-week ABM playbook for cold markets, and where AI content generation is quietly heading next
- Quick Links: 53% of brands invisible in AI answers, 100% use AI but 13% trust it, voice vs text agents in B2B outreach
Kevin's Take
We bought AI to speed up work that was already fast
There's a piece in the Agile Brand Guide (link) that names something I've been thinking about for a while, which is that a lot of us went out and bought AI this year to speed up the parts of our funnel that were already running fine.
We automated the email send, we sped up subject line variants, we cut a few minutes off campaign setup, we generated more versions of the ad we were already running, and none of that touched the place where our deals were actually stuck.
When I look at where our own work with clients breaks down, it's almost never the campaign build. It's the stuff in between. It's the two weeks waiting on legal review, the sales team that never followed up on the MQL, the product marketing brief that sat in someone's inbox for nine days, the fact that nobody really knows what the ICP is anymore because it drifted three quarters ago and nobody updated the doc.
Those are the real bottlenecks and none of them get fixed by a tool that writes copy 40% faster.
We used to measure marketing productivity by throughput, how many campaigns we shipped, how many emails went out, and buying AI to speed up throughput made a kind of intuitive sense. But the constraint has moved. The constraint is decisions, alignment, and the handoffs between teams, and most of the AI we bought doesn't touch any of that.
One thing I've found useful is to actually map out where a lead or a project sits and waits before we go looking at any tool. Nine times out of ten the wait isn't in a step AI could speed up, it's in a step where a human owes another human a decision.
If we're renewing an AI martech subscription this quarter, the question this hard about is whether the thing we're paying for shortens a wait anyone was actually complaining about.
That's it for this week. Talk soon.
— Kevin Kerner, CEO, Mighty & True
The Signal
New survey benchmarks how CMOs are actually spending AI budget (1 min read)
Open Future Forum surveyed marketing leaders who control AI budgets to benchmark how CMOs and VPs of Marketing are actually deploying AI spend and leverage across their orgs. The report defines and segments respondents by decision-making authority over marketing AI investment.
Why it matters: This gives you a benchmark to see whether your AI budget and adoption is ahead of or behind your peers.
New data on how B2B buyers actually use AI search (1 min read)
Octane11 releases its second 'State of B2B AI Search' report with fresh data on how B2B buyers are using AI search tools to research and evaluate vendors.
Why it matters: This is original data you can use to justify shifting budget toward AEO/GEO tactics before your competitors catch on.
Tools & Tactics
An ABM case study for launching new products into cold markets (1 min read)
A first-person ABM case study details how itsme's CMO launched a new enterprise product (nextAuth) in an unfamiliar market with zero brand awareness, generating 40% account engagement, a 27% account-to-meeting rate, and a 17% account-to-SQL rate in 10 weeks. The piece breaks down the CMO's hands-on role and the sales-marketing dynamics that drove results.
Why it matters: If you're launching into a new market with no pipeline history, this gives you a real playbook and benchmarks for what a hands-on CMO-led ABM motion can produce in just 10 weeks.
Amazon's research shows how retrieval-augmented AI improves marketing content quality (1 min read)
Amazon researchers introduce MarketingFM, a retrieval-augmented LLM system that generates and evaluates customized marketing content at scale using multiple integrated data sources. The paper addresses limitations in existing LLM-based content generation by grounding outputs in retrieved brand and product data.
Why it matters: This shows you where AI content generation is heading: retrieval-grounded systems that can produce on-brand, customized copy at scale rather than generic LLM output.
Quick Links
News & Trends
- Most brands are invisible in AI answers, new data shows — Boring Marketing's analysis of thousands of AI platform checks finds 53% of brands never appear in AI answers, and 51.7% of citations link back to a brand's own pages
- Marketing uses AI everywhere but trusts it almost nowhere — Bessemer's Atlas report finds 100% of marketing teams use AI for content creation, but only 13% consider it core to their operations
- AI model memory predicts which brands buyers actually search for — geoSurge's study of nearly 4,000 AI model responses across 66 buyer questions found brands already 'remembered' by AI training data get searched for online roughly 3x more than unf (via TLDR)
- 13 weeks of data on AI voice vs text agents in B2B outreach — Docket.io ran a 13-week first-party test comparing AI voice agents to AI text agents in B2B outreach and shares proprietary performance data on which channel drives better results.
Work with us — Growth Consulting
Working through a strategic pivot? Mighty and True Growth Consulting partners with B2B tech CMOs on positioning, GTM rewires, and AI operating model work. Bring us the hard question.
Help us with a research report
We are putting together a research report on how mid-market tech companies are actually allocating budget in 2026. Not the aspirational version, the real one. If you are willing to share your numbers anonymously, reply and I will send the survey. You will get the full report before anyone else.