AI will make planning uncomfortable before it makes it better

The number one reason organizations fail to meet their overall business plans isn’t because the market is unpredictable or a certain department didn’t work hard enough. It’s because their planning is siloed.

Here’s a common example. Sales is aligned to revenue. SDRs are aligned to meetings booked. Marketing is aligned to generating MQLs. Stack that up and it looks like a good system. More MQLs should translate to more meetings, and more meetings should translate to more revenue.

But we all know how this story can play out. Marketing and SDRs can smash their numbers and Sales can still miss quota. Two teams can say they hit their metrics, great, but they didn’t move the needle on the one outcome that really matters. When functions don’t collaborate and share data, they can’t see how their efforts affect the rest of the funnel.

Applying AI to the planning process forces the organization to compare notes. It surfaces where assumptions don’t line up, or how one team’s KPIs aren’t actually driving business outcomes.

And it does so in a way that’s hard to ignore. That can be uncomfortable for functions trying to obfuscate their workarounds or bunk metrics (“These aren’t the droids you’re looking for”). But once you know where the weaknesses are, you can stop arguing about whose model is right and start to reconcile the KPIs, workarounds, and incentives protected by siloed systems. You can start doing what you should have from the beginning — integrate the plan.

How “local truths” form inside siloed systems

It helps to understand how planning misalignment happens. It usually starts innocently enough.

A function buys best-of-breed tools that work pretty well within their own world until, inevitably, the gaps appear. Then the team builds spreadsheets to track what the software can’t. They build decks to explain what the spreadsheet means. Pretty soon, they haven’t just created one shadow system. They’ve constructed a whole shadow city.

Each shadow system produces its own numbers. I call these “local truths.” In the glued-together decks and spreadsheets that serve as the function’s unofficial planning system, local truths rule the day.

When local truths collide in a company-wide plan, they don’t line up. Every department has gone off and created their own numbers without talking to each other. Their assumptions start to drift. And by the time teams have to present their output — say, MQLs generated — they’re so far out of alignment that everyone ends up arguing about the end results instead of fixing the inputs.

Silos survive, even when they hurt performance, because they protect territories

There’s a layer to this story people don’t talk about as much. Business silos aren’t always accidental. Some teams maintain them intentionally because they protect their turf.

If you own the model or the definition of a metric, you control the narrative. If the numbers live in your system, you decide what counts and what doesn’t. Siloed planning can help teams obfuscate inefficiency or risk, and give them negotiating leverage come budget season.

People get evaluated and rewarded inside their functions but punished for global problems they can’t control. So they build protection via a version of a metric that makes sense for them.

Everyone can hit their KPIs and still miss. AI shows you where

Because AI can connect planning across functions, it surfaces contradictory local truths quickly. And one of the biggest mismatches it exposes involves incentives.

The go-to-market example I shared earlier demonstrates exactly how local optimization produces global planning failures. SDRs and marketing were both hitting their KPIs, but that’s all they were doing. They weren’t motivated to reach for integrated outcomes like revenue because their bonus structures are tied to pipeline.

If you only incentivize a piece of the funnel, you’ll encourage behavior that maximizes that part, even if it creates dysfunction elsewhere. If you want real collaboration, you have to align incentives across the system.

AI makes this painfully clear because it doesn’t just look at one KPI. It connects marketing efforts to sales results. It spots patterns like high activity and low conversion. It can tell you the pipeline you’re generating isn’t qualified — without a quality measure on the backside, sales ends up with quantity that doesn’t convert.

These conversations aren’t comfortable. Siloed functions tend to avoid them because they’re forced to show their hands. But that discomfort is the cost of alignment. When teams see how individual metrics consistently fail to add up to shared outcomes, they have to come together and redefine what an MQL should look like.

When AI enters the planning process

When you apply AI across planning and start to see cracks in a narrative sales or marketing has maintained for years, people start to get defensive. AI has just cast a spotlight on your numbers, and boy, do they look off. Defensiveness is a natural reaction.

AI can actually help teams reframe and move through planning conflict in three specific ways.

First, AI can act as an independent arbitrator. It can help depersonalize the usual battle of ideas and egos. Instead of “my model versus your model,” you can say, “Here’s what the analysis shows. What do we think about this?”

Second, AI can stress-test the plan. It’s built to evaluate assumptions and flag anomalies at a scale that we miss as biased humans.

Finally, it speeds up the planning process when human alignment exists. It can seed a forecast, evaluate plan inputs, and help you identify where you’re on and off track.

Leaders have to be willing to get uncomfortable

As I write this, San Francisco has been locked in unusually chilly weather. Our heater just broke, because of course it did. We’d been running it nonstop for weeks — we put too much stress on the system.

That’s how planning works too. Companies only discover how fragile their planning models are when the business is strained. That’s why leaders need to use AI proactively before the cold snap hits.

AI doesn’t fix the truly hard part, the human part of this process. It just exposes the places where siloed planning can hide. Once you know where your local truths and ownership gaps are, it’s up to department leaders to actually integrate the plan.

Leaders must decide whether they align incentives to shared outcomes and reconcile metrics that don’t make sense across the full funnel — or defend them and stay in their comfort zones.