Can AI do strategy? What StrategyFrame's Strategic Intelligence report means for M&A and corporate finance teams

AI can do everything, but can it also do strategy? StrategyFrame's Strategic Intelligence report answers this question and explores what it means for M&A and corporate finance teams.

StrategyFrame is one of the partners we work with most closely at StrategyBridgeAI. Their team relies on our platform to build the analyses behind their client engagements, and the relationship runs both ways: our CEO Deniz Schütz and data scientist Geisi Shima contributed a guest article on AI in company valuation to StrategyFrame's latest publication.

That publication, "Strategic Intelligence," is part two of StrategyFrame's four-part series "Hoffnung ist keine Strategie" (Hope is not a strategy). It looks at how AI is changing strategy work, from the first plan to daily execution, and it is directly relevant to anyone running deals, valuations, or transformation programmes for a living.

In this article you'll learn:

• Why StrategyFrame now treats strategy execution as an always-on operating system rather than a one-off project
• Which tasks specialised AI agents take on during market, competitor and financial analysis
• What an independent academic study found when it asked practitioners "can AI do strategy?"
• Why using the same AI tools as everyone else can erode, not create, competitive advantage
• How AI improves peer group selection and comparable company analysis in valuation work
• What real transformation projects at Fabromont and Caritas Krefeld & Meerbusch reveal about AI in practice

From a one-off project to an always-on operating system

StrategyFrame's core argument is that strategy execution used to be treated as a project with a beginning and an end: cascade the strategy, transform, experiment, adjust, done. Their view now is different. Execution is an operating system that keeps running in the background, and the question has shifted from "are we implementing the right things?" to "how does AI change our work, step by step?"

That shift only became possible after a lot of unglamorous groundwork. Before any AI was introduced, StrategyFrame broke strategic work down into roughly 18,500 discrete tasks through its StrategyFrame® methodology. Structure came first, then the system. It is a useful reminder for any organisation evaluating AI tools: the technology amplifies whatever process discipline already exists. Where strategy work is well organised, AI makes it faster and better. Where it is not, the gaps just become visible faster.

What AI actually does in strategy development and execution

The report splits strategy work into two phases. Development runs through four steps: plan, analyse, focus, adapt. Execution runs through another four: launch, implement, run, adjust. The first two are set up once, the last two run continuously.

During analysis, the report describes a layered agent setup: a market agent that tracks structural shifts, a competitor agent that follows positioning and pricing signals, a macro agent that contextualises economic and geopolitical developments, and a dedicated financial agent that supports company valuation and helps interpret market reports, particularly around acquisitions and capital decisions.

During execution, an OKR assistant checks whether objectives at team, division and company level actually line up, and a project portfolio view flags where initiatives overlap, where gaps exist, and what happens if priorities shift before those conflicts turn into missed deadlines or budget overruns. A background monitoring layer scans market, competitor and regulatory data continuously and flags when underlying assumptions stop holding, turning course correction into an ongoing task rather than a quarterly event.

None of this replaces judgment. The report is explicit that AI sharpens what is already there. An unclear target picture gets analysed faster, not clearer. Conflicting priorities get named more precisely, without resolving themselves.

Can AI do strategy? What an independent study found

The centrepiece of the issue is an interview with Prof. Dr. Julia Hautz, professor of strategic management at the University of Innsbruck and one of the leading European voices on open strategy. Together with Dr. Thomas Ortner, she ran 15 interviews across seven organisations using StrategyFrame®AI, in a study still in progress.

The most common expectation, Hautz found, was a "magic button": upload documents, get a finished strategy. That expectation did not hold up. What did materialise was a genuine time saving on the grunt work of synthesising and documenting information, which in turn freed up time for the harder part: reflection and alignment between the people who provided the input.

Hautz also names what she calls the differentiation paradox. If every company runs the same analysis through the same AI tools, AI becomes table stakes rather than a source of advantage. In her interviews, this pushed the differentiator back toward people: proprietary data, organisational culture, and the ability to mobilise many employees rather than just speed up a single analyst. Her framing has since shifted from "can AI do strategy?" to "should AI do strategy with us?", a distinction one interview participant summed up bluntly: someone has to own the decision, and the AI will not feel the consequences of it.

AI in company valuation: fixing the peer group problem

Deniz Schütz and Geisi Shima's guest contribution tackles a problem every valuation professional recognises: the comparable company or transaction group you choose drives the result, and broad market medians or size-and-sector rules of thumb are a blunt instrument.

Using radar charts comparing two sample companies, the article shows how the picture changes depending on the benchmark. Against the total market, both companies look similarly average. Against their actual sector and closest peers, one clearly outperforms and the other clearly underperforms, which directly affects which multiples are defensible.

The article names three practical obstacles to building sector-specific peer groups: sourcing the underlying financial data, selecting peers when standard classifications like NACE or SIC miss relevant companies or misclassify them, and dealing with peers that simply do not publish financials for the years needed. It argues AI addresses all three: extracting data from unstructured sources at scale, ranking candidates by actual similarity in offering, customer base, region and sector rather than by classification code alone, and using machine learning to estimate missing or forward-looking financials so a valuation is not stuck relying on stale historicals.

Proof from the field: Fabromont and Caritas Krefeld & Meerbusch

Two case studies back up the argument with real transformations. Fabromont, the Swiss manufacturer behind the Kugelgarn® textile flooring category, used a StrategyFrame® situation analysis and discovered its market share was roughly half of what leadership had long assumed, prompting a shift from product-led to customer-segment-led strategy.

Caritas Krefeld & Meerbusch shows the same methodology applied outside the corporate world. The nonprofit adapted the classic BCG growth-share matrix to its own language, replacing "poor dogs" with "hamsters" and "question marks" with "penniless church mice" to describe socially essential but financially unviable services it commits to keeping anyway.

The bottom line

AI clearly changes how strategy work gets done: faster analysis, sharper peer benchmarking, earlier warning on scope and resource conflicts. It does not change who owns the decision or who is accountable for it. For M&A, valuation, audit and corporate finance teams, the practical takeaway is narrower and more useful than the hype suggests: use AI to fix the peer group problem, tighten portfolio tracking, and free up time for judgment calls, not to outsource the judgment itself.

FAQ

Can AI actually build a company's strategy on its own? No. StrategyFrame's own study found this expectation, sometimes called the "magic button" scenario, was widespread but unmet in practice. AI accelerates analysis, synthesis and documentation, but the decision and its consequences remain with the people accountable for the business.

Why does using the same AI tools as competitors reduce differentiation? If every company runs comparable analyses through comparable AI systems, the output converges. Prof. Julia Hautz's research describes this as a differentiation paradox: AI becomes a baseline requirement rather than a competitive edge, pushing real differentiation back toward proprietary data and organisational execution.

How does AI improve comparable company analysis in valuation? Instead of relying on total market medians or size-and-sector rules of thumb, AI can build peer groups based on actual similarity in business model, customer base, region and financial performance, correcting for the blind spots of standard industry classifications like NACE or SIC.

What does AI do about missing financial data in a valuation? Machine learning models can estimate missing or forward-looking financial metrics for a target company or its peers, rather than defaulting to stale historical data that may no longer reflect current performance.

What is project portfolio management's role in AI-supported strategy execution? It shifts the question from tracking individual projects to asking whether the organisation is doing the right things, in the right order, with the resources it actually has, and AI supports this by simulating the effect of reprioritisation before conflicts turn into missed deadlines or budget overruns.

See how StrategyBridgeAI supports company research, benchmarking and valuation work like this for teams in M&A, banking, PE and audit: book a demo at strategybridge.ai.