Are McKinsey Reports Actually Actionable? Where Theory Ends and Execution Has to Begin

One of the most-read article I have published here on Orr-Consulting.com is about the limitations of McKinsey's 7-S framework. That surprised me at first. It stopped surprising me once I looked at the search queries bringing people to it, because almost none of them are academic. They come from executives holding something impressive and trying to work out what to do with it on Monday.

One of those queries is why I am writing a follow-up. Someone found my site by typing this into Google: are McKinsey reports actionable or just high-level? It is an unusually good question, and a more honest one than most executives ask out loud after spending an afternoon with a beautifully produced piece of research. The 7-S piece answered a narrow version of it about one framework. This is the broader version, and it is the one people are actually searching for.

Here is the short answer.

McKinsey reports are usually actionable in the sense that they improve how you think about a problem. They are usually not actionable in the sense that they tell your company what to do next. A report can identify a trend, size a market, and benchmark an industry. It cannot determine what that finding means for your customers, your positioning, your economics, or your budget.

That distinction is the entire article.

I want to be careful here, because this is not an argument that McKinsey does poor work.

The firm is genuinely good at identifying macro shifts, synthesizing large datasets, benchmarking industries, and building frameworks that keep smart people from forgetting an important dimension of a problem. The published research is often excellent.

The failure happens in translation. An executive reads something important and true, feels the click of recognition, and walks out of the room believing the company now knows what to do on Monday. Those are two different states of knowledge, and the distance between them is where most of the actual work lives.

1. Are McKinsey reports actionable or just high level?

Think about how information becomes results:

DATA → INSIGHT → STRATEGY → DECISION → EXECUTION → MEASUREMENT

A strong industry report often does an excellent job of the first two steps, and occasionally gestures at the third. Everything after that belongs to your company. Nobody outside your business can cross that bridge for you, because the remaining steps require access to your customer data, your margins, your competitive position, your team, and your constraints.

Theory is not action. Research is not strategy. A framework is not strategy. A benchmark is not strategy. A trend report is not strategy. An eighty-page deck describing what high performers are doing is a description of other companies, and describing other companies is a different exercise from planning your own.

Consider a finding of the type that appears in nearly every major report right now: consumers increasingly discover brands through AI-assisted search.

That is true. It is also important. And no executive can implement it.

Now what?

To act on that sentence, a company has to answer a long chain of questions that the report does not touch. Are our buyers actually doing this, or is the trend concentrated in categories that are not ours? Which segments? For which purchase decisions, and at which stage of consideration? Which platforms are they using? What questions are they typing? When those questions get asked, which competitors surface, and why do they surface? Which sources are the models citing when they answer, and do we appear in any of them?

Does our positioning survive being summarized by a machine that has never seen our website? Where are the authority gaps? What content closes them? What third-party coverage do we need and how do we earn it? What technical changes matter and which are noise? What does this cost? What are we going to stop funding to pay for it? How will we know in ninety days whether any of it worked?

That chain is strategy. The report supplied the premise. Everything after the premise is company-specific analysis, and it is the part people skip.

I have watched this pattern repeat for years. A leadership team reads something credible, agrees that it matters, and then converts a macro trend directly into a tactic without any of the intervening work. The trend says AI is changing discovery, so the company schedules an AEO project. Maybe that is the right call. Nobody in the room has established it, because nobody looked at whether this company's buyers behave like the sample in the study.

2. What does a personalization finding actually require before it becomes a decision?

Take another common finding: personalization improves customer experience and commercial performance.

Also true. Also unimplementable as written.

Now what?

Personalized to whom? A brand with four distinct segments and a two-year repurchase cycle has a completely different problem from a subscription business with weekly touchpoints. Which behaviors are worth reacting to, and which are noise dressed up as signal? Which data do you actually have, and is it clean enough to trigger anything without embarrassing you? Which touchpoints matter, given that most personalization programs die in email while the website stays generic? What is the offer, since personalization without a reason to buy is just a name in a subject line? How deep does it go, because segment-level relevance and one-to-one dynamic content are separated by roughly an order of magnitude in cost? What technology does the depth you chose require, and does it integrate with what you already own? What incremental lift do you expect, and against what baseline? What privacy constraints apply in your category, which in healthcare is not a footnote? And what does implementation cost in money, in engineering time, and in the attention of a marketing team that already has a roadmap?

Answer those and you have a personalization strategy. Skip them and you have a vendor demo.

3. What does a marketing efficiency finding leave out?

A third example, and the one I get asked about most: high-performing organizations allocate resources differently and use data more effectively.

That statement is nearly a tautology, and it still gets circulated in board decks as though it were a directive.

Now what?

Allocation questions are answerable, but only with your own numbers. What does each channel actually cost you at the margin, which is a different question from what your dashboard reports as blended CAC? What is the contribution margin on the products those channels sell, because revenue-based ROAS targets have quietly bankrupted a lot of otherwise healthy brands? What is lifetime value by acquisition source, and how confident are you in that estimate given your repurchase data? Where are you saturated, meaning the next dollar into that channel returns meaningfully less than the last one? What share of your reported conversions would have happened anyway? Which channels are incremental and which are taking credit for demand created elsewhere? And what organizational constraints govern the answer, because a recommendation to shift half the budget into a channel nobody on the team knows how to run is a recommendation to fail slowly.

None of that is in the report. All of it is in your data, or it is obtainable.

4. How does a shift in consumer expectations become a marketing plan?

The fourth pattern is the broadest. A report identifies changing consumer expectations, usually around trust, transparency, speed, or value.

Now what?

The path from that sentence to a plan runs through customer research, then competitive analysis, then white-space analysis, then positioning, then messaging, then channel strategy, then creative, then testing, then measurement. Every one of those steps can invalidate the previous one. Customer interviews routinely reveal that the expectation described in the study is real but ranks fifth among the things your buyers care about. Competitive analysis routinely reveals that two competitors already own the obvious response, which makes the obvious response expensive and unwinnable.

The report told you the weather changed. It did not tell you whether to bring a coat, and it certainly did not tell you whether your competitors are already selling coats.

5. What does the report give you, and what do you still have to figure out?

Report-level insight	What your company still has to determine
AI is changing how customers discover brands	Whether your specific buyers use AI at the discovery or validation stage, which prompts surface your category, which competitors and citation sources appear, why they appear, which authority gaps you can realistically close, what you fund, and what you defund to pay for it
Personalization drives commercial value	Which segments justify the investment, which behaviors trigger which experiences, whether your data is clean enough to act on, what the offer is, what incremental lift you expect against a holdout, and what the technology and privacy constraints cost you
Marketing efficiency separates high performers	Channel-level marginal returns, contribution margin by product, LTV by acquisition source, where you are saturated, which conversions are incremental, and where the next hundred thousand dollars goes
Consumer expectations are changing	Which expectations your customers actually rank highly, whether competitors already own the credible response, where the white space is, and whether positioning, pricing, or product has to change rather than the campaign

The left column is where research ends. The right column is where a consultant, a CMO, or a very disciplined internal team has to start.

6. What has to happen after you read the report?

This is the part worth printing out.

Validate the finding against your own market

National and global averages describe a population. Your customers are a sample of that population selected by your pricing, your channels, your category, and your brand, which means they are frequently unrepresentative in ways that matter.

Start with what you already own. Internal customer data. Sales data by segment and cohort. CRM records and what the closed-lost reasons actually say. Site search queries, which are the cheapest and most ignored source of customer language in existence. Then go get what you do not own. Customer interviews, which remain the highest-return research activity available to a mid-market company. Conversations with prospects who chose someone else. Call recordings, if sales will let you near them. Reviews, including competitor reviews. Survey work where the question is genuinely quantitative and you have enough respondents to justify it.

I have had findings from major reports hold up perfectly against client data. I have also had them collapse entirely, which is equally useful information and considerably cheaper to discover before you fund a strategy around it.

Conduct social listening

Structured research tells you what people report doing. Social listening shows you what they say when nobody is asking.

Reddit is the most underused strategic asset in marketing right now, particularly in healthcare, telehealth, and any category where buyers are embarrassed to ask questions in public. TikTok and Instagram show you how the category is being explained to newcomers, which is usually not how you explain it. YouTube comments under competitor and review content are where objections live in raw form. Industry forums and private communities carry the B2B version of the same behavior. Reviews tell you what broke after purchase.

What you are looking for is language, frustration, alternatives, and drift. The exact words customers use, which are almost never your marketing words. What annoys them enough to complain about publicly. What substitutes they are comparing you against, which is frequently not the competitor set in your deck. And where attitudes are shifting before formal research catches up, because published research is a lagging indicator by construction. This is also where I do a large share of the work behind what I call the Lurker Economy, because a growing share of buying conversation now happens in channels your analytics will never see.

Analyze competitors properly

A logo slide is not competitive analysis.

Look at how competitors position themselves and whether that positioning has changed in the last year. Look at their messaging hierarchy and what they lead with. Look at product and service architecture, including what they bundle and what they hold back. Look at pricing and pricing structure, which are separate decisions. Read their reviews, especially the three-star ones. Look at what they rank for organically and what they surface for in AI answers. Look at what they are paying for, how long the ads have been running, and what the creative is arguing. Look at their content and where it gets cited. Look at partnerships and distribution. Read the complaints. Catalog their claims and, more importantly, the proof they offer for those claims, because unproven claims are the softest place to attack a competitor.

Find the white space

This is the step that separates strategy from summary.

Given what customers want and what competitors already credibly own, where can we win?

A market trend cannot answer that question, because a trend is available to everyone in the category simultaneously. If a report tells you and your four largest competitors the same thing on the same day, following it is a defensive move at best.

White space requires four conditions to hold at once. A real customer need exists. Competitors are weak there, or absent, or making claims they cannot substantiate. Your company has a credible right to make the claim, which is a constraint people love to ignore. And the opportunity is large enough and profitable enough to be worth the redirect. Three out of four is a common and expensive failure.

Determine whether positioning has to change

When a market genuinely shifts, the correct response is often not a campaign.

The company may need to change what it stands for, what it promises, which benefits it leads with, what evidence supports those claims, which segments it prioritizes, which products it pushes forward, how it prices, or where it sells. Running new creative against outdated positioning produces exactly what you would expect, which is better-looking versions of the same result.

Translate strategy into channel decisions

Now get specific, because this is where strategy becomes budget.

What changes in Google Ads, in keyword strategy, in match types, in what you are willing to pay for a non-brand click? What changes in SEO, and separately in AEO and GEO, which are related but not identical problems? What changes in paid social, in organic social, in email flows and segmentation, in PR targets, in influencer and partnership selection, in the website itself, in sales enablement material, in CRM nurture logic?

And the question almost nobody asks: what does not change? Strategy is a set of choices about where to concentrate, which means most of it is a decision to leave things alone. A plan that changes everything is not a plan, it is a reorganization, and it usually destroys the baseline you needed in order to measure anything.

Put actual dollars behind the decisions

A strategy that does not change resource allocation is commentary.

The most clarifying question I ask clients is this one: if I hand you the next hundred thousand dollars of marketing budget, where does it go and why? The answer surfaces everything. Whether the team understands marginal return or only average return.

Whether anyone knows where saturation begins. Whether CAC is being measured against contribution margin or against revenue. Whether LTV is a real estimate or a spreadsheet artifact. Whether the growth priority is new customer acquisition, retention, expansion, or category creation, since those four demand very different allocations. And whether anyone has considered opportunity cost, which is the discipline of naming what you are giving up.

Measure incrementality

Tracking KPIs is not measurement. Plenty of KPIs improve while the business does not.

What is appropriate depends on scale. Smaller advertisers should be running structured experiments and holdouts, which are cheap, unglamorous, and more informative than any dashboard. Geo testing works well for anyone with enough regional volume to split.

Attribution has a place, provided everyone understands that it describes correlation with paths, not causation.

For larger advertisers with several years of spend history across multiple channels, marketing mix modeling answers questions a benchmark report structurally cannot. What did each investment actually contribute to this business? How quickly are returns diminishing in each channel? What happens to total revenue if we move budget from one place to another? Where does the model say we are overspending relative to marginal return?

Not every company needs MMM. Below a certain spend level and data history it produces confident-looking output with intervals wide enough to drive a truck through, and I will tell a client that directly rather than sell a model they cannot support. Incrementality testing is often the better starting point, and it is available to almost everyone.

Build an implementation roadmap

Everything above has to land on a calendar or it will not happen.

In the first thirty days you validate, instrument, and fix the measurement you will need later, because launching anything before the tracking is trustworthy guarantees an unresolvable argument in month four. In the next sixty days you build and launch the highest-confidence changes, usually positioning and messaging on the properties you control. By ninety days you have the first tests reading and the first reallocation decision in front of you. Across three to twelve months you tackle the structural work: authority building, third-party coverage, content that takes time to compound, product or pricing changes, and any measurement infrastructure worth standing up.

Every line needs an owner, an investment, an expected result, a KPI, and its dependencies.

Every roadmap also needs two things people leave out. What gets stopped, and what gets tested rather than assumed. That is the point where strategy finally turns into action.

7. What can a framework do, and what does it leave out?

A framework is a starting point. It earns its place by preventing intelligent people from forgetting a dimension of the problem under time pressure, which is a real and recurring failure.

I have written at length about where this breaks down in Limitations of McKinsey's 7-S Strategic Planning Framework, including why the big-firm model tends to misfire specifically for companies in the five to fifty million dollar range.

The short version applies to every framework ever drawn on a whiteboard. The framework does not contain your customers. It does not contain your competitors, your margins, your data, your capabilities, your positioning, your budget constraints, your organizational politics, or your market timing. Those variables determine the answer, and none of them are in the diagram.

A framework organizes the questions. The answers have to come from analysis of your specific business.

8. What did twenty years in academia teach me about this?

I spent roughly two decades teaching marketing before moving fully into CMO and consulting work. If anything, that made me more appreciative of good theory rather than less. Theory explains why things happen. It generates better hypotheses, and a better hypothesis saves you months of testing in the wrong direction. My doctorate is in marketing with concentrations in statistics and psychology, and I still read the literature.

The transition to practice made one distinction impossible to ignore. Being able to explain why something should work is a different competency from making it work inside a company.

In a paper, the sample is clean and the conditions are controlled. In practice there is a budget that is smaller than the plan requires. There is data with gaps in it and a tracking implementation nobody has audited in two years. There are entrenched competitors with more money. There is a sales team with its own view of the customer. There are customers who behave nothing like the research sample because they found you through a channel the study never considered. There is a technology stack that cannot do the thing the strategy requires. And there is an executive who needs an answer by Friday, not a literature review by the following quarter.

Good theory improves decisions. It does not make them. What produces results is theory combined with company-specific evidence, judgment about what to do when the evidence is incomplete, and the willingness to implement and measure.

The path from insight to action

Industry Research ↓ Company-Specific Evidence ↓ Customer and Competitive Analysis ↓ White-Space Positioning ↓ Strategic Choices ↓ Budget and Channel Allocation ↓ Execution ↓ Incrementality and Measurement ↓ Learn and Reallocate

The report is the beginning of the process, not the end.

9. When do you need more than a strategy report?

Some situations are genuinely solvable by reading good research and thinking clearly. Others are not, and the tell is usually that the information is already sitting in the building.

You are past the reading stage when leadership has absorbed the research and still cannot agree on what to prioritize. When three or four opportunities all look attractive and no one can rank them. When the market has moved and repositioning is on the table. When growth has flattened despite the same activity that used to produce it. When spend keeps rising and revenue does not rise proportionally. When two agencies are recommending contradictory things and both sound reasonable. When you have plenty of data and no decisions. When a board or a private equity sponsor wants a growth strategy that can survive contact with an operating calendar. And when you already have a strategy deck but nothing that tells the team what to do in the next thirty days.

That gap between analysis and action is most of what I do as a fractional CMO.

So, are McKinsey reports actionable?

They are valuable. They surface trends worth knowing about, challenge assumptions that deserve challenging, and give executives a sharper vocabulary for problems they were already circling. Reading them is time well spent.

Insight is not action. Theory is not execution. The value gets created when someone takes the external finding and works out what it means for these customers, this competitive position, these economics, and this quarter's investment decisions.

That work sits between the report and the result, and it does not happen by itself.

If your team has read the research and is now stuck on what to actually do about it, that translation is the work I do. You can book a marketing strategy call and we can talk through where you are.