A letter from our CEO
I’m Aaron, CEO of Inflection. I’ve spent 20 years in B2B marketing — building the tools, and living with the ones that were built for me.
We’re watching the shift from the internet era to the AI era. Here’s how I see it playing out, and what it means for CMOs.
For thirty years the pendulum has swung from vendor control toward buyer control. In the last few years it went 10x with AI — and the tools we all still run were built for control.
01 The Control Shift
Control moved from the vendor to the buyer.
In the internet era we could genuinely orchestrate the journey — gate, score, nurture, educate. We owned the white papers. We owned most of what a buyer read before deciding.
Not any more. By the time someone fills out a form, they already know everything. They have an analyst in their pocket that beats Gartner, and it costs twenty dollars a month.
So we can’t control the buying process. We can influence it. What we can still control is the customer.
Nurturing didn’t die. It moved to the only part of the journey you still own.
You owned the white papers, the nurture, and the path to the decision.
By the time they're known, they already know everything. Everything before the decision can only be influenced — what you still control is the customer.
02 The Slop Flood
Anyone can generate infinite content, so everyone does.
AI took the cost of making content to zero, so everyone is pumping it out. LinkedIn shipped an “AI slop” tag you can put on posts.
Which puts us back to trust. Buyers listen to humans talking about your product — not to the content on your website.
Almost nothing gets through the glass any more. A peer, a customer, a community they trust — that’s it.
When anyone can generate infinite content, the scarce thing isn't content. It's a human willing to vouch for you.
03 Where Value Went
All the value went to two extremes. Everything in between hollowed out.
On one end: real humans talking about you — events, communities, customers telling your story. You can’t generate it, which is exactly why it’s worth something.
On the other: one-to-one messages powered by data only you have. Product usage, account history, the moment someone hits an aha.
In the middle: public data through the same models everyone else uses, producing the same message everyone else is sending.
We’re all using the same models. The only differentiation left is unique data that fuels unique context.
The middle is everything that could be bought, scraped or generated — which is now everything everyone else is sending too.
04 The Old Way
The 2006 playbook, encoded in software: gate, score, drip, hand off.
It only starts paying attention at the form fill — which is now the moment after the decision has been made somewhere it can’t see.
And it does almost nothing once someone becomes a customer — the one stretch of the journey you still control.
It’s not that the old stack is bad at its job. It’s that its job moved.
Everything it's good at happens after the decision has already been made somewhere it can't see.
05 The New Way
The funnel becomes a loop. Buyer experience and customer experience feed each other.
Attract, convert, adopt, advocate — running continuously on your own data, with no handoff and no dead end after the sale.
You create advocates through better onboarding and lifecycle, then activate them. Advocates feed the front of the journey — the part you can no longer control, but can still influence.
And you measure it full-funnel: pipeline and NRR, not MQLs.
This happens with or without you. Virtuous cycles turn vicious just as easily.
Advocates feed the front of the journey — the part you can no longer control, but can still influence.
06 How Work Gets Done
Everything so far was outside the four walls. Inside them, the work itself changed.
The platform got democratised — anyone in marketing can draft a campaign from a prompt, not just marketing ops.
Agents work like coworkers, not like software. You brief them once and they remember. The queue stops being one person.
A small team stops being a constraint and becomes the design — people plus agents, deliberately structured.
Before
One marketing ops queue
Now
The platform got democratised. Agents work like coworkers, not like software — so the bottleneck stops being a person.
07 Winners & Losers
Pun intended.
In 2026 buyers moved to AI, and the split has already started.
The winners: one-to-one communication only they can write, customers telling the story, small teams plus agents.
The losers won’t feel a moment of failure — just CAC creeping up and reply rates drifting down, quarter after quarter.
The buyer changed. The economics changed. Legacy tools haven’t caught up.
Same models, same public data, same playbook — the only thing left to compete on is what only you have.
Our website uses cookies. By continuing, we assume your permission to deploy cookies as detailed in our privacy policy.