AI-Assisted Best Next Action 260-percenter

"Sales organizations that provide sellers with AI-enabled next best actions are 2.6x more likely to achieve commercial growth"

The PR lead from a Vegas Gartner discussion earlier this year.

Wayhey, let's all go out and get AI'd up.

Well, not quite. For there's plenty of holes in such pronouncements. Which the promoter conveniently leave aside under clickbait amplifying their preferred buzz.

For starters, there's the stack of untested assumptions.

Take the classic; correlation ≠ causation. It's the big one. Where are proper controls for prior growth trajectory, company size, industry, sales maturity, or overall tech spend? Was this specific angle adopted and apparently fruitful because they were already higher-performing outfits? With more budget full stop, better processes, stronger talent, and more willingness to experiment, #failfast or iterate?

Next there's that "AI-enabled next best actions" is a fairly woolly and self-reported category. What does it even mean? A basic CRM recommendation engine? Or sophisticated agentic system? Maybe simply the salesperson’s own prompt engineering? Naturally with the seller alone deciding the 'magic' sparkled.

Then we've a litany from selection bias, of delegates predisposed to talk up their "commercial growth" (another famously fuzzy metric), who let Gartner write it all up, in a way that suits them.

Without doubt the best salesteams I've been involved with share incredible informal deal management for precisely this kind of 'next best action' formulation. When you couple this with a focus on true process - where a pattern of events will be known to practically guarantee you prevail - you're in a special place.

This type of figure - 📣 Two Point Six Times SalesPeople!! - is a marketing conceit yes, yet within the caveats is there some actual juice?

Other vested interests have pushed their marvels too. You can find reference to such as when 'AI recommendations are followed 50% higher win rates' accrue, and '89% of Sales leaders using 'agentic AI' reporting a positive impact on sales growth'.

Strangely in contrast to overall studies you might note though. Widely reported examples you may have heard include that '95% of generative AI pilots showed no measurable impact', with 'roughly twice the failure rate of comparable non-AI IT projects', and as of now '42% of companies abandoned most of their AI initiatives in 2025'.

Wouldn't you know it, Sales shines as a departmental outlier putting our laggard colleagues to shame, hey. If only...

For all my scepticism, I actually like the concept of 'next best action' bot.

I once developed a deck of 52 'playing cards' that when stuck on a bid, you could randomly cut to get an idea on what to do next.

It is by no means a huge leap to go from these kinds of prompts to having an AI back-end create it for you.

You'd gain consistency across the whole team, scale, capturing tribal knowledge that walks out the door. And force some kind of tracking too, as imagine you flag considered options and the one taken with its outcome.

And the crazy thing is, you likely have much of the data already at hand.

If your hand is on the lever in a sizeable concern, you can give this idea a go for sure. How you populate it, and then measure that needle moving, well, that's another matter entirely...