Kevin's Take
Stop Using AI to Hire Fewer People
Everyone's talking about AI like it's a better spreadsheet. Faster emails. Quicker reports. Content at the speed of light. I call this Efficiency AI, and yes, it matters. But if that's your whole strategy, you're thinking too small.
The real game is what I'm calling Opportunity AI: using these tools to do things that weren't possible before. Not "write my blog post faster" but "now we can personalize our entire content library for 47 different buyer scenarios." Not "answer support tickets with fewer people" but "proactively solve problems customers didn't even know they had yet."
I'm seeing two very different conversations in boardrooms right now. The first group sees AI agents and immediately starts running headcount math. How many SDRs can we replace? How many writers can we cut? Every agent is a potential pink slip. They're optimizing for Q2 expense reduction. The data backs this up. Forrester Research found that 55% of companies that cut headcount citing AI as the driver now regret it. Quality degraded, institutional knowledge walked out the door, and some, like Klarna, ended up quietly rehiring the people they cut.
The second group is asking completely different questions. How do we use AI to let our best people work on higher-value problems? What can we do now that was literally impossible six months ago? What new revenue streams open up when content production costs approach zero?
Here's the Framework
Efficiency AI asks: How do we do the same thing cheaper?
Opportunity AI asks: What can we do now that we couldn't do before?
Both matter. But only one compounds. Companies optimizing purely for efficiency are playing defense in a market that rewards offense. You might save 200K on headcount this year. Great. Meanwhile, your competitor just figured out how to serve 10 new micro-segments profitably because their AI isn't replacing people; it's multiplying what those people can create.
The Short-Term Trap
Yes, AI will create efficiency in the short term. Your content team will move faster. Your operations will get cleaner. But the companies that win are the ones using that efficiency as fuel, not as a finish line.
When your writers suddenly have 10 extra hours a week because AI handles the first draft, you have a choice. Fire two writers, or redeploy that time to create entirely new programs you've been saying "we don't have bandwidth for" since 2022.
When your marketing ops person can suddenly spin up complex workflows in hours instead of weeks, you can cut the role to part-time, or you can finally build that account-level intent scoring system that's been on the roadmap forever.
The Opportunity Mindset
The Signal
LinkedIn Is the Only B2B Ad Platform Delivering Positive ROI — Dreamdata
Dreamdata's LinkedIn Benchmarks Report 2026 found that LinkedIn is the only major advertising platform delivering positive return on ad spend for B2B marketers at 121%, while Google and Meta both show negative ROAS. The research surveyed B2B companies and compared actual revenue attribution across platforms, not just engagement metrics.
Why it matters:If you're splitting paid budget equally across LinkedIn, Google, and Meta "to test everything," you're lighting money on fire on two of those channels. This is the kind of data you bring to Q2 planning when finance asks why LinkedIn gets 60% of paid budget. The report essentially gives you permission to stop pretending Meta works for enterprise B2B.
Virtual Event Platforms Aren't About Events Anymore — Forrester
Forrester's 2026 Virtual Event Management Platforms Landscape reveals that live event delivery is now table stakes, and vendor differentiation has shifted entirely to post-event engagement and content repurposing capabilities. The firms that used to compete on "better webinar experience" are now competing on "what happens in the 90 days after the webinar ends."
Why it matters: If you're still evaluating event platforms based on registration flows and chat features, you're optimizing for the wrong thing. The real ROI is in turning one keynote into 40 pieces of content and keeping attendees engaged for months. This should change both your vendor selection criteria and how you staff event programs—you need content operations people, not just event coordinators.
ICONIQ GTM Report Shows Deals Closing Faster With Shorter Contracts — SaaStr
ICONIQ surveyed 150+ GTM executives at B2B and AI software companies and found that deal velocity is up significantly in 2026, but buyers are demanding shorter initial contract terms. Companies are closing faster but getting less revenue locked in upfront. The report also shows teams are getting leaner while quota attainment is holding steady or improving.
On Our Radar
HubSpot's 2026 research shows that 58% of marketers report visitors from AI search tools convert at higher rates than traditional organic traffic. If you're still treating AEO as experimental, you're leaving conversions on the table.
Marketers are quietly reallocating SEO budgets away from traditional search optimization toward GEO, optimizing for visibility in generative AI outputs instead of clicks. The shift is happening faster than most CMOs realize.
FROM THE TRENCHES
The Site That Started Falling Apart on Page Two
The Problem:
A client came to us mid-project frustrated. They had been building their site with AI assistance for several weeks. Page one looked sharp. By page four, fonts had drifted, spacing was off, and two sections used components that didn't match anything from the original build. They assumed the tool was broken. It wasn't. Claude Code has no memory between sessions. Every build starts completely blank. Without a system to read from, it improvises -- and improvisation compounds fast across a multi-page site.
The Fix:
We built a seven-file structure that lives directly in the project repo. A CSS file where every color, spacing value, and border radius is defined once. A ten-rule constraint doc. An approved component library. A running log of mistakes already made so they don't get repeated. Claude Code reads all of it at the start of every session.
The Result:
The inconsistent pages rebuilt in one session. Two weeks later, the client needed five campaign landing page variants. They built in hours, not days, and every page matched.
We wrote up the full seven-file system here: Why AI-Built Websites Fall Apart After Page Two