By Signe Enemark Kildesgaard and Cecilie Bonde Christiansen
December 16, 2025
Q4 is peak season for strategy development. Budget processes are running at full speed in most organizations, and strategy statements are used to frame our ambitions for the coming year — both within our own area and across the entire company.
Strategy statements are used as leverage to argue for the budget choices we make and the expectations we have regarding investments or cutbacks. What should we prioritize our resources toward? What results will our initiatives create? And how do we ensure that everyday work helps move the needle toward our big ambitions?
Most of us know the drill. You first have to learn it — figure out the strategic (read: political) way of wording things, secure buy-in from decision-makers, and create yet another PowerPoint to sell the message before everything ends up in Excel. But what does it actually take to develop and articulate strategy — and how can AI tools be used in the process?
We are often fortunate to be invited in as sparring partners during strategy development — and this autumn, the exercise looks a little different than usual. At first glance? Strategy documents have simply become better. The language flows more naturally, the wording (often in English) is more nuanced and sophisticated, and many now have a clearer structure and narrative flow that draws the reader in and helps them understand what the strategy is actually trying to achieve. It’s great to see leaders in Danish companies becoming better at articulating the strategic ambitions for their areas — and of course, we know they’ve had qualified help. There is no doubt that ChatGPT has saved them hours at the desk and done what it does best:
analyze text, identify patterns, predict appropriate responses one token at a time, and adjust structure and style based on your instructions. This means the text often feels human and meaningful — and we experience that strategy statements have improved.
So what’s the problem? You only really see it at “second glance” — when you scratch beneath the surface. And here three issues emerge:
- The beautifully worded strategies seem to have become longer. We simply write more, restate points in multiple ways, and add nuance to support understanding. But do these extra paragraphs actually help — or do they dilute our messages?
- The words we use in a strategy matter. If they are statistically generated by a language model, do they then lose their meaning — both semantically and in practice?
- And can leadership truly stand behind — and lead from — words, arguments, and messages that they themselves have only contributed to by prompting for them?
The golden cliché: Less is more
Most people have probably experienced how ChatGPT can make texts shorter and “sharper.” But new research also shows the opposite effect: the content we produce is getting longer — for example on social media like LinkedIn[1] and in recruitment, both in job postings and applications[2]. And it’s easy to see why it’s tempting to “just” add one more paragraph to elaborate and nuance your point when, with a simple prompt, you can get three versions of 20, 50, and 100 words. Then you’re sure you’ll appear both competent and sharp.
But classical behavioral and cognitive psychology tells us that it actually has the opposite effect. First of all, research shows that when we add more words to already clear points in order to elaborate or nuance them, we actually dilute those very points. It simply becomes harder for us to understand them[3]. And we also know that when a sender uses unnecessarily complex wording or phrasing, they are actually perceived as less intelligent — it simply looks like hot air[4].
It can be tempting to create a large, polished document that examines the issue from all angles and restates the messages several times. But in reality, what we need are short, concrete points so the messages stand strong and clear. And that places even greater demands on the words we choose.
The power of words
Words are absolutely central to strategy. Major strategies and transformations in global companies have been built around a single word or simple phrase as their essence. Think Different. Quality First. Good to Great.
And this actually creates major challenges when ChatGPT is your writing partner. It is neither as precise nor as consistent in its word choice (after all, it works probabilistically) as is required when we are defining strategic direction using only a few key statements.
Let’s take an example:
Amazon’s vision is:
“to be Earth’s most customer-centric company, where people can find and discover anything they might want to buy online.”
If you ask ChatGPT to “improve” the formulation (a poor prompt — but for the sake of example), the suggestions might look like this:
- “To be the world’s most customer-centric company by creating a place where everyone can discover and access whatever they need — anytime, anywhere.”
- “To set the global standard for customer-centered commerce by making every desired product easy to find, understand, and purchase online.”
- “To become Earth’s most customer-centric ecosystem, where people can freely discover, choose, and receive anything that makes their lives easier and better.”
Particularly in the third suggestion — which ChatGPT describes as “ambitious and forward-looking” — we’ve drifted quite far from the original.
Because it does matter whether we call ourselves a company or an ecosystem. And whether people should simply be able to buy anything they want online — or whether it should also make their lives easier and better.
In ChatGPT’s world, the wording remains faithful to the original intent and strongly appeals to the reader — but it does not necessarily reflect the strategic choices and trade-offs we want and need to make as a company. And that, ultimately, is the purpose of strategy.
Strategy formation is a critical leadership process
Finally, we should remember that there is a reason it’s called a strategy process.
AI tools like ChatGPT have made it easy to create a strategy document with fluent language, solid structure, and a compelling narrative. And many may recognize that the strategy process can become somewhat rushed when — for good or less-good reasons — it falls in a period when companies are already extremely busy ahead of Christmas and year-end. And AI is, of course, a convenient way to speed things up.
But perhaps we need to remember — and even protect — the balance between process and product. Because one thing is that the strategy document as a “product” becomes the starting point for the organization when budgets are prepared, tactical plans are developed, communication happens, and execution begins.
But the strategy formation itself is a critical leadership process — where, through dialogue, we set direction, coordinate and prioritize our efforts, and commit — together. The process is therefore absolutely essential to the leadership we must collectively produce across the organization.
That is also why the strategy document becomes evidence of the strategic discussions, the choices and trade-offs made by leadership. Your final strategy document is just as much a confirmation or acknowledgment that leadership has created a shared direction and committed to it — forming the foundation for your tactical execution in the coming year.
Start with people, end with (a little) AI
The challenge with big visions and strategy has always been that they risk becoming either too flat — and therefore not truly visionary, ambitious, or strategic — or too lofty and abstract, and thus so detached from everyday life that they cannot be translated into action. And that is exactly what strategy must do: provide a long-term plan for our work that guides our priorities and activities.
But ChatGPT does not provide this simply by rewriting strategy statements.
It still requires you as a leader to understand your market context, know your customers’ needs now and in the future, understand what it takes for your employees to thrive and develop the capabilities the future demands. And to articulate this in messages you can personally stand behind — and communicate to your organization.
So how should we work with strategy development in the age of AI?
Here are 5 recommendations:
- Select — possibly with ChatGPT as a sparring partner — a number of key questions you and your colleagues can use to clarify your strategic direction. Remember: the better your prompt and the more information you share, the more qualified the output — within the policies of your organization.
- Spend three hours or more offline with trusted colleagues, sharing thoughts and perspectives based on those questions. Together, you sketch the contours of a strategic direction.
- Sit down and write. Start with a blank page and formulate your conclusions in accessible language. What is your overall strategic ambition? What are your strategic goals — and what will you do to achieve them?
- Ask ChatGPT to act as a critical voice on your first draft — get it to ask questions or provide feedback rather than taking over the writing task.
- Finally — feel free to use ChatGPT to refine language and structure, but stay sharp and ask it to show you the specific changes. Think of it as an editor — not a co-author.
This article was written without the help of ChatGPT. We know it’s too long — but we’re glad you made it to the end.
[1] https://originality.ai/blog/ai-content-published-linkedin?utm_source=chatgpt.com
[2] Silbert, Jesse (2024/2025). Job Market Paper. Title: Generative AI and Labor Market Signaling.️ https://jesse-silbert.github.io/website/silbert_jmp.pdf
[3] Dilution effect: Nisbett, R. E., Zukier, H., & Lemley, R. (1981). The dilution effect: Nondiagnostic information weakens diagnostic judgments. https://deepblue.lib.umich.edu/handle/2027.42/24406
[4] Oppenheimer, D. M. (2006). Consequences of erudite vernacular utilized irrespective of necessity: Problems with using long words needlessly. ️ https://onlinelibrary.wiley.com/doi/abs/10.1002/acp.1178