Strategy Is Dead. Or Is It?

Why the businesses moving fastest on AI are often the ones skipping the thinking that decides whether it pays off.

Recently I met with a leadership team that explained that they had added three AI tools in a single quarter. When I asked what each one was supposed to change, the room fell silent. The silence that filled the room was palpable. It was more that the question had not come up, because the urgency had arrived first.

I have been in that quiet room before, on both sides of the table. Lately I hear versions of it in almost every conversation: a competitor announced something, the business or marketing professionals that they network with asked about their AI plan, and out of curiosity or panic, a tool was purchased. The tool is not the problem; the order is.

It is also why I keep hearing echoes that strategy is dying. Longtime marketing professionals tell me AI can handle the strategy and the insights now, and that nobody wants to pay for the slow part when they can move faster. For a while that will hold. Then the work gets into the weeds, and the wall shows up. The question is circulating publicly, too. A recent CMSWire piece opens with a marketer asking Reddit whether brand strategy is dead, and the author's answer was no, because AI can generate content but not trust. I agree with the author.

The question is circulating publicly, too. A recent CMSWire piece opens with a marketer asking Reddit whether brand strategy is dead, and the author's answer was no, because AI can generate content but not trust. I agree with the author.

Pressure doesn't kill the thinking, it postpones it

Some would argue that strategy has never been the glamorous part of business. Most owners will tell you it matters, and in the same conversation mention that nobody has time for it this quarter. AI didn't create that habit. Though it did put it under a spotlight, because now the cost of waiting feels like falling behind.

So again, the thinking gets dismissed, or more often delayed until after the rollout. I understand the instinct. When everyone around you appears to be moving quickly, standing still to ask questions feels like the riskier choice.

What I mean by the thinking is a few honest answers to important questions. What problem are we trying to solve? What would be different in the business if this worked? Who has to change what they do for that to happen? How will we know? Who needs to upskill? None of that is exotic, and none of it can be handed to a tool.

Trust is where the wall shows up first. In Supermetrics' 2026 marketing data report, only 18% of marketers reported high trust in AI, and 37% said leadership had not given them a clear vision for AI. A late 2024 NP Digital survey of 1,124 marketers found trust in AI was highest for analytics and lowest for strategic planning, and only 24% relied on it for campaign strategy. People are being asked to move fast with tools they do not fully trust, pointed at a direction nobody has set.

What skipping it looks like

The pattern is consistent enough that I can typically spot it in our initial conversation. A tool arrives before a defined problem. Nobody has written down what the work looks like today, so there is no baseline to compare against. Without a baseline there is no return to measure. The tool might be helping, and even if it is, nobody can speak to it.

There is still a quieter version of the same skip, and it shows up even when something works. My colleague Dawn Gagye recently shared with me that a business she works with received a lead through an AI search, and it was the most qualified lead they had encountered. It was also heading toward their fastest close ever. The owners were thrilled, and then came the harder question: what exactly did that person ask, and why did it lead to them? It was the unknown in the room. The win was real, but was it an accident? I ask because an accident cannot be repeated or scaled.

The specialists I trust on this are consistent about what holds up. Search specialist Jill Corbett with Mile 20 Marketing, who joined Dawn and me on our latest CMO Real Talk episode, put it this way: "The biggest thing out there right now on how you get found is to have that foundation, have it say what you actually want it to say, and then have that backed up in a bunch of other places." She was equally direct that the shortcuts circulating right now are short term, not long term. Her version of the foundation is plain: say clearly what the business is, answer the questions customers actually ask, and let reviews and testimonials back it up. The tools can only amplify what is already there.

I see this in my own client work too. The tools are often already built, and what is missing is the connection between them and how the business actually sells and serves. That connecting is the foundation, it takes time, and it is slower than buying or starting something new, which is a big part of why it gets skipped.

The pieces are usually all there: a CRM, a campaign, a website that was rebuilt last year. What takes time is connecting them to how a lead gets followed up on, and nobody sees that work until it is done. Before any connecting starts, a careful look often shows that a system is not doing what the team believes it is doing, so I audit first and connect second.

AI can't live in its own department

The most common version of this I see is an AI plan that sits off to one side. IT or an enthusiastic team member owns it. Marketing has its own separate set of experiments. The business plan, meanwhile, was written before any of it existed and hasn't been revisited. Three efforts, three sets of assumptions, and no one asking whether they point at the same goal.

A useful test is to take any AI effort and trace it upstream. Which business goal does it serve? Which part of the marketing work does it change? Who in sales or operations will feel the difference, and do they know it is coming? If the trail goes cold at any step, the effort is a silo, however impressive the demo.

The people who can answer those questions are usually already in the building. On the episode I said that no matter the size of the company, "the most untapped resources are the sales teams and the operations teams," because each holds a different view of the ideal customer. Jill agreed, and with her franchise clients that is where she starts, asking franchisees what they wanted to know and what made them sign. That knowledge is exactly what makes an AI effort aim at something. A tool pointed at a guess produces confident-sounding guesses faster.

This is also where I arrive at the unsettling idea that strategy is dead. Let me explain! What is dying is strategy that arrives as a deliverable, because AI can draft a plan in minutes. What is not dying is the judgment about whether that plan fits this business, this team, and this customer. The version that connects the business goal to the marketing work to the tools is not going anywhere. As I mentioned on the episode, it comes down to "business strategy aligned with marketing strategy in a coherent state," because without that nobody can tell whether the tools are earning their keep.

Where I'm least sure

I may be underestimating how much a business can learn by simply trying things. Jill's own advice to people new to this work was to experiment with the tools themselves and see what comes back, and I do believe that is right. A small, cheap test can teach more than a month of planning.

There is a harder version of this doubt, too. Some of the businesses moving fastest may never hit the wall. The tools keep improving, and a team with good instincts and a little luck can get further than a missing baseline would predict.

What I keep circling back to is the difference between experimenting and deciding. A test with a question attached, a time limit, and someone watching the result and who is thinking in motion. An AI tool purchase with none of those components is just perceived momentum. Admittedly, I am less certain where the line sits for a business with no spare time or budget to run careful tests at all.

Your turn

Think of the last AI tool your business adopted. What was it supposed to change, and what would you point to today as evidence that it did? If the honest answer is a pause, that is worth sitting with, and I would like to hear what you find.