Every sales feed this year sold you the same story. AI ate the SDR, the AE is next, hire a bot and go home.
The latest data is less convenient. The companies that gutted their teams are quietly rehiring. Half of all reps are missing quota. And the biggest names in software are spending billions to prove that AI with no human on top does not close anything.
Four pieces. Read them in order. There is a pattern.
Piece 01 · The great SDR downsizing, and the quiet rehire
Emergence Capital surveyed more than 560 B2B software companies. 36% cut SDR and BDR headcount in the past year, the steepest drop of any sales role. Only 19% added.
The headline misses the second half. Teams that swapped reps for pure-volume AI watched their domains get flagged and their reply rates fall through the floor. Inbox providers spent the year tightening the rules. Send a thousand generic emails a day and you do not book meetings, you burn the domain. The same filtering now runs on the phones: if your dials go quiet too, start with the voicemail flag.
Jeb Blount had this right in Fanatical Prospecting. Prospecting is a discipline, not a volume dial. The teams that won did not fire the floor. They put a trained operator on top of the AI, so the machine handles the grunt work and the human does the thinking.
So what: if you cut your team and the pipeline left with it, that was never an AI problem. It was a nobody-steering-the-AI problem.
Source: SaaStr on Emergence Capital's Beyond Benchmarks
Piece 02 · Half the field missed. The other half used AI.
The Bridge Group's State of Sales 2026, the tenth edition of the study, put annual quota attainment at 48%, down from 51% two years ago. More reps are sliding into the 0 to 30% danger zone than the study has ever recorded.
Now the split, from the same report. Teams with the highest AI engagement had 57% of reps hit quota. Teams with the lowest had 39%. Bridge Group calls it correlation, not proof. The two still move together for a reason.
Put those numbers side by side. This is not a down market. It is a divided one. The average rep is selling into tougher buying committees with the 2022 playbook. The AI-fluent rep qualifies faster and spends the day on the deals that actually move.
Keenan's Gap Selling names why that works. Buyers move when someone quantifies the gap between where they are and where they want to be. The AI-fluent rep gets to that conversation first.
So what: work harder stopped being the answer. The reps pulling ahead changed how they work, not how long.
Source: The Bridge Group, State of Sales 2026
Piece 03 · Salesforce just showed you where the stack is going
On June 1, Salesforce signed a definitive agreement to buy Contentful, reported at around a billion dollars, to bolt a content engine onto Agentforce. Three weeks later it shipped its biggest Agentforce Commerce release yet, with ChatGPT integration going live.
Read it plainly. The CRM is turning into an operating system for AI agents. Data, content and agents in one place, and every major platform is running the same direction.
So what: your stack is consolidating around agents whether you budgeted for it or not. The question is no longer whether to use AI. It is who on your team owns the agents and checks their work. Which is the last piece.
Source: Salesforce acquires Contentful
Source: Salesforce Agentforce Commerce release
Piece 04 · Everyone bought AI. Almost nobody is getting paid for it.
McKinsey's State of AI: 88% of organizations now run AI in at least one function, up from 78% a year ago, with marketing and sales among the most common uses. 62% are at least experimenting with agents.
Here is the catch. Only 39% report any profit impact at the enterprise level, and most of them put it under 5%. Buying the tool is not the same as banking the return.
The winners share a habit. McKinsey's high performers rebuilt their workflows and put human validation on the output, a person who checks the work before it ships. They are nearly three times more likely to have redesigned those workflows than everyone else.
So what: the teams making money from AI kept a skilled human in the loop. The teams that skipped that step bought a faster way to send worse emails.
Source: McKinsey, The State of AI
The bottom line
The thread through all four: 2026 did not reward the companies that replaced people with AI. It rewarded the ones that put trained people on top of it.
That is the whole game now. Not more reps. Not more tools. A skilled operator who knows how to drive the machine.
If your outbound went quiet this year, that is usually the missing piece. Subscribe below and reply to any issue with how your setup looks. I will tell you straight where the gap is.
Built to close,
Justin Power
Founder, 2CT.