You can't automate your way out of a broken foundation

I've spent over 20 years in RevOps, mostly with West Coast software companies. I grew up figuring things out on spreadsheets, building models by hand, chasing numbers across tabs, and trying to make sense of data that never quite lined up the way it should. So when I say that go-to-market planning is getting harder, I mean it. I've lived it.

And here's the thing: AI was supposed to make this easier. In some ways, it has. But in other ways, it's created a whole new category of problems that we need to talk about honestly.

The world has changed fast. Maybe too fast.

Think about where we were with models like Claude just six or seven months ago compared to where we are now. The pace of change has been remarkable. Software companies have fundamentally shifted how they operate. The tools available to us have multiplied. And the pressure to move quickly, to use AI everywhere, to automate everything, has never been greater.

But speed without accuracy is dangerous. And right now, a lot of organizations are moving very fast in the wrong direction.

We're still carrying a lot of legacy software that's expensive to operate and requires whole teams just to keep it running. We've still got spreadsheets floating around everywhere, which means manual work, which means errors.

And ops teams are already stretched thin. Most of us don't have the bandwidth to properly vet every new tool or model that lands on our desks.

What AI actually helps with (and where to be careful)

Let me give credit where it's due. AI does some things genuinely well. It's much easier now to get visibility into your data. It automates the repetitive, time-consuming work that used to eat up hours of an ops professional's week. It speeds up analysis and helps with consistency. These are real benefits, and they matter.

But decisions? That's where I get cautious.

I've seen C-suite leaders plug AI models directly into their spreadsheets, feed in capacity numbers and targets, and then make decisions based on whatever comes out. And it's frightening.

We've had customers where multiple departments are constantly emailing IT asking why the numbers are wrong, because there's a whole layer of filters and logic that needs to be applied, and the AI is simply missing it. The model doesn't know what it doesn't know.

AI on bad data is dangerous. Full stop. If the underlying data isn't right, if the structure isn't sound, if the logic hasn't been properly applied, then the output is going to be wrong, and it's going to be confidently wrong. That's worse than being uncertain.

So before we talk about how AI fits into your GTM planning process, we need to talk about what a good planning process actually looks like.

The core elements of a proper planning process

When we think about building a solid go-to-market plan, there's a sequence that matters. You work through capacity first, then productivity, then efficiency, and only then do you get to targets and quotas. Most organizations do this backwards, and that's where the trouble starts.

Sales capacity

Sales capacity is the prediction of what your sales team should be able to achieve. The formula is straightforward: it's sales productivity multiplied by headcount. Weighted headcount by average production per seller.

That's it. That's how you calculate sales capacity.

When I ask audiences how many of them are calculating capacity this way, the hands that go up are always fewer than they should be. Most teams are using quota attainment as their productivity metric, and that's a real problem.

Quota attainment is unstable. It shifts based on how quotas are set in the first place, which means you're building your capacity model on a circular reference. You need a stable, independent metric, and that's average production.

To get to average production, you look at how much business has actually been closed in a specific segment or territory, divide that by the number of sellers who could have sold in that same period, and you've got your number. That's what you use to calculate capacity. That's the foundation everything else builds on.

Sales productivity

Once you know your average production, you can start benchmarking performance properly. You can identify where productivity is coming from and where it isn't. You can track the performance of the team against a metric that actually means something.