Hard truths from building sales engines, where AI belongs, and where it's going. Opinionated, and updated as I learn.
Be clear who you're for, and selling gets easy.
The right customer, at the right time, beats ten times the activity. I disagree with "always be closing": a few of the right customers bring more revenue, sooner and for longer, than lots of small ones. Tell the rest, "we're probably not right for you."
Every stage should give the buyer something: the first call, the demo, the proposal and the negotiation. When someone pushes on price, push back with value, not a discount. That takes confidence: you set the price for a reason, so hold it. Treat the product as valuable and buyers respect it more.
Each layer caps the one below. No process sells to the wrong customer, and AI on a broken process only scales the break.
SDRs book what AEs reject when nobody has defined a good deal. Write one definition, and use it everywhere.
It shows what a company really values, whatever the strategy deck says. People do what they're paid to do.
Return on effort. Busy teams push uphill on deals that won't close, and on admin no buyer ever sees.
Do they know what's expected, can they do it, do they want to, does the system let them? Only one of those is about will.
For a growing company, most revenue comes after the signature: first impact, renewal, expansion and referral.
Build the way you sell into the CRM and the playbook, so the whole team and its agents can follow it, not just your best rep.
Two more I live by: trust is reliability, not niceness, and keep asking why until the buyer says, "that's exactly it."
The moments that matter are where a buyer decides whether to trust you. The work between them is work an agent can do.
Someone has to put their name to it, and the buyer has to feel it. People buy from people they like. AI can do a great deal, but on its own it feels cold.
I don't scale an AI workflow until it sends what we would send, at least 95% of the time. One bad email to your best prospect can cost you the deal.
Take your last 100 leads, and the replies your best people actually sent.
Let the agent research, score and draft a reply for every one.
Your best seller marks each draft: would I send this? If not, why not?
Turn the feedback into rules, examples and categories. Run the next 100.
Go live with a person approving every message, and track how often they change it.
When the reviewer is just pressing send, send, send, automate it, and move to spot checks.
What I'm seeing from inside sales teams right now. I'll update it as things change.
Teams replace too much of the human experience, scale before the message is right, and flood buyers with slop. Bad in, bad out.
Find where you lose customers and time, then automate the repetitive work around it. Pour AI into a leaky bucket and you just lose more, faster.
A dozen point tools make a dozen silos. Build one system you control, with one source of truth, and grow it over time.
Voice agents will ask the qualifying questions at a volume no team can match. A lead SDR runs them and joins the calls that go well.
A strategic head of sales, a few skilled SDRs and AEs each running many agents, and RevOps keeping the system sharp.
The right customer, the right message at the right time, and a buyer who feels understood. AI makes each step faster. It skips none of them.
The system behind these ideas. Five levers, seven layers and 28 focus areas, scored on evidence. That's the Commercial Engine.
See the EngineIf you understand people, show them the value and say it in the right way, everyone wins: the customer, the company and the team. When something is genuinely good, it deserves to be understood. That's the job.
One lever or one AI workflow per issue, what I'd change this month, and what happened when real companies did it. Short, practical, no hype.
Want to dig into any of this? I talk about these ideas on podcasts, panels and at portfolio workshops, and I'm always up for a good argument about them.
Invite me to talkApply in one minute. If it's a fit, you book a call straight away and leave with at least one thing to fix this month, whether we work together or not.