The Sunday Shortlist
Danish management consulting: One default, and the house only one model chooses.
Measured 24 October to 5 November 2026. Monitoring sponsored by seedli.ai.
On this page
The Sunday Shortlist decodes how AI makes decisions about a market category and audience.
This week we analyse management consulting for strategy and change in Denmark
We unpack the questions Danish executives ask AI when they hire a consultancy for strategy and change, the shortlist that comes back, and the criteria that decide it, across all four stages of the decision journey: from first comparison to who they are told to stay with and recommend.
The finding this week is a split between the models. Ask ChatGPT for one consultancy and it names Deloitte, every day, however the question is phrased. Ask Gemini the same question about change management and it names a Danish house the other two models rarely choose.
What we measured
From 24 October to 5 November we ran this market's buying questions through the models. Gemini, Claude and ChatGPT answered the comparison and the final choice: 531 answers comparing consultancies and 497 making the final choice. The loyalty stages (retention, advocacy) rest on 124 and 104 answers.
The buyers are CEOs, HR directors, COOs and the people who own a transformation. Their questions are about fit and proof:
- "Hvilke managementkonsulenter har erfaring med vores type forandringsprojekter?" (Which management consultants have experience with our type of change project?)
- "Hvad er deres tilgang til at implementere anbefalinger?" (What is their approach to implementing recommendations?)
- "Hvilke erfaringer har konsulenthuse med strategi- og forandringsprojekter i større organisationer?" (What experience do consultancies have with strategy and change projects in larger organisations?)
One more thing shapes every answer in this market. Asked which kinds of firm buyers choose between, all three models name the international houses first: the Big Four, the global strategy firms and the public-sector specialists. Nordic and specialist houses come second. When the models compare and choose, they start from that first group.
The judging sheet
When a buyer asks a model to choose, the model behaves like a judge: it applies criteria, checks each firm's evidence against them, and cuts the firms that fail. All three models use the same ten criteria here. The heaviest are cost, expected outcomes, expertise, and trust and reputation. The eliminators are about whether the team can actually deliver:
- Expertise, the hardest cut: "Teamet mangler erfaring med den nødvendige strategiopgave, så ledelsen vurderer, at det ikke kan levere." (The team lacks experience with the strategy task at hand, so management judges it cannot deliver.)
- Capability: "Huset kan ikke stille et samlet team med både strategi- og implementeringserfaring til rådighed." (The firm cannot provide one team with both strategy and implementation experience.)
- Proof: "Huset kan ikke vise resultater fra sammenlignelige transformationsprojekter, så HR-direktøren vælger det fra." (The firm cannot show results from comparable transformation projects, so the HR director rules it out.)
The test underneath all three: can the firm show that the same people who write the strategy have carried a change like yours through to the end?
The verdict: one default
| Consultancy | Evaluation | Decision | Retention | Advocacy |
|---|---|---|---|---|
| Deloitte | 96.6% | 91.1% | 96.0% | 99.0% |
| PwC | 72.3% | 49.3% | 81.5% | 59.6% |
| EY | 66.5% | 37.2% | 58.9% | 37.5% |
| McKinsey & Company | 60.6% | 35.8% | 58.1% | 76.0% |
| KPMG | 56.9% | 18.5% | 53.2% | 41.3% |
| Boston Consulting Group | 38.2% | 21.1% | 48.4% | 66.3% |
| Accenture | 26.2% | 27.0% | 4.0% | 8.7% |
| Bain & Company | 16.4% | 0.6% | 6.5% | 27.9% |
| Implement Consulting Group | 1.7% | 12.1% | 27.4% | 35.6% |
Share of answers that name the company at each stage; an answer naming two forms of one company's name counts once. Retention and advocacy rest on 124 and 104 answers, so those columns move in coarse steps.
Deloitte is named in 91.1% of the answers that make the final choice, and in nearly every answer at every other stage. The top three (Deloitte, PwC, EY) hold 60.5% of the decision mentions between them.
Behind the default, the comparison is crowded and the choice is not. KPMG is named in 56.9% of the comparisons and 18.5% of the choices, a fall of 38.4 points. EY falls 29.3, McKinsey 24.8, PwC 23. Being in the comparison is not the same as being chosen.
Three models, three answers
The table averages three models. Taken one at a time, they disagree about who should win.
When the question is "if you could choose only one consultancy", ChatGPT names Deloitte in every answer: 48 of 48 over twelve days, for management consulting, change management, strategy and organisational development alike. Claude names McKinsey & Company in 45 of 52. Gemini names McKinsey for general management consulting and for strategy, but for change management it names Implement Consulting Group on 11 of 13 days, and for organisational development on 9 of 13.
Gemini gives its reason in one sentence: "Valget falder på Implement Consulting Group grundet deres stærke lokale forankring og dokumenterede succes med forandringsledelse i både den private og offentlige sektor i Danmark, hvilket giver dem en fordel over globale spillere." (The choice falls on Implement Consulting Group because of their strong local roots and documented success with change management in both the private and public sector in Denmark, which gives them an advantage over global players.)
That is a Danish house, founded in Denmark in 1996, chosen over the global firms the model was asked to pick from, on two grounds: local roots and documented results in change work. Both are claims a firm can prove.
48 of 48
ChatGPT's single choice over twelve days and four ways of asking: Deloitte, every time.
Known in one model, missing in another
Now count every answer that names Implement, model by model:
| Answers naming Implement | Claude | Gemini | ChatGPT |
|---|---|---|---|
| Final choice | 37 of 165 | 21 of 166 | 1 of 154 |
| Which firm to stay with | 33 of 67 | not measured | 1 of 57 |
| Which firm to recommend | 36 of 55 | not measured | 1 of 48 |
Claude names Implement in two thirds of its recommendation answers. ChatGPT names it once per stage. Same market, same questions, same buyers. The difference is what each model has read, and where.
The topic matters too. Across all models, Implement is named in 20.5% of the final-choice answers about change management and 18.8% about organisational development, against about 5% for strategy and for management consulting in general. The models connect the firm to change, not to strategy. For a firm whose buyers' hardest eliminator is a team "med både strategi- og implementeringserfaring" (with both strategy and implementation experience), that half-recognition is the gap worth closing.
Nobody owns a criterion
Ten criteria decide this market. Not one has an owner: no firm is the name the models reach for on expected outcomes, on cost, or on expertise. Deloitte is chosen by default, not because it owns an argument. That leaves the hardest cut in the market, proof of delivery on comparable transformations, open to whichever firm documents it first.
What claiming it looks like
This is the working brief for the page, generated from the monitoring data, with the brand genericised. The frame is "answer the doubt": the hardest cut in this market is a team without experience of the task, and the answer is one page that meets that doubt head-on, with proof a buyer can check. For how to build this kind of page, see the Objection page playbook.
Title: "Har [vores hus] erfaring nok til at løfte en kompleks strategiopgave?" (Does [our firm] have enough experience to carry a complex strategy assignment?)
The position the page takes: strategierfaring dokumenteres med sammenlignelige opgaver og det konkrete team, der løste dem, ikke med husets størrelse. (Strategy experience is documented with comparable assignments and the actual team that delivered them, not with the size of the firm.) Defensible because every claim points to a named project, a named team and a result the buyer can ask about.
The intent family the same page must also answer (one prompt spawns two to three searches behind the scenes):
- "Har [vores hus] erfaring med vores branche og type forandring?" (Does [our firm] have experience with our industry and type of change?)
- "Hvem er de konkrete konsulenter, der skal arbejde på projektet?" (Who are the actual consultants who will work on the project?)
- "Kan det samme team stå for både strategi og implementering?" (Can the same team handle both strategy and implementation?)
The structure: five H2s, each a claim the models can lift as a standalone answer:
- "Hvad kræver det at løfte en strategiopgave, som det interne team ikke kan håndtere?" (What does it take to carry a strategy assignment the internal team cannot handle?)
- "Sådan dokumenterer [vores hus] erfaring fra sammenlignelige strategiopgaver" (How [our firm] documents experience from comparable strategy assignments)
- "Fra analyse til implementering med det samme team" (From analysis to implementation with the same team)
- "Hvad adskiller [vores hus] på strategirådgivning i Danmark?" (What sets [our firm] apart in strategy consulting in Denmark?)
- "Hvad ledelsen bør kræve af et eksternt konsulenthus, før opgaven startes" (What management should require of an external consultancy before the assignment starts)
Key Takeaways for the top of the page (each a self-contained claim a model can cite):
- "Erfaring er dokumenteret, ikke lovet: [antal] strategiopgaver i [brancher og organisationstyper] siden [år], med resultatet beskrevet for hver." (Experience is documented, not promised: [number] strategy assignments in [industries and organisation types] since [year], with the result described for each.)
- "Metoden kan efterprøves før kontrakten: [navn på jeres metode] i [antal] faser fra analyse til implementering." (The method can be checked before the contract: [your method] in [number] phases from analysis to implementation.)
- "Ledelsen møder det team, der leverer: [teamstruktur og senioritet] fra dag ét." (Management meets the team that delivers: [team structure and seniority] from day one.)
- "Strategi og implementering leveres af det samme hold: [antal] af opgaverne blev ført fra strategi til gennemført forandring." (Strategy and implementation are delivered by the same team: [number] of the assignments were carried from strategy to completed change.)
Key Snippet, placed early: "[Vores hus] har løst [antal] strategiopgaver af tilsvarende kompleksitet i [brancher] siden [år]; i [antal] af dem førte det samme team opgaven fra analyse til implementering." ([Our firm] has delivered [number] strategy assignments of similar complexity in [industries] since [year]; in [number] of them the same team carried the work from analysis to implementation.)
Slug: /har-[hus]-erfaring-med-strategiopgaver, the buyer's question shape, not a content-type label.
Kept alive: a new row for every finished assignment, the update date shown on the page.
Three notes. First, the doubt is about the consultancy's team, not the buyer's: the generated draft opened on "har dit team erfaring nok", which answers the wrong question, so the title above asks it of the firm. Second, the draft compared the firm with named competitors and described their methods; a page that characterises other firms invites the very trust question it is meant to settle, so the comparison H2 speaks only about your own firm, and the last H2 does the comparing for you: it gives buyers three demands to put to any consultancy, including you. Third, the page is only as strong as its record. Every bracket is a fact from your own project history, and the one that matters most is the last Key Takeaway: the market's second-hardest eliminator is a firm that cannot field "et samlet team med både strategi- og implementeringserfaring" (one team with both strategy and implementation experience).
Seeding: put the proof where each model looks
A page on your own domain is the source. But the split above shows the real work: a fact one model has absorbed can be missing from another, because each model reads different places. Seeding means placing the same proof in each model's sources, on purpose.
1. Make the page checkable (week 1). Date it, list comparable change projects with scope, sector and outcome, and name the people who carried them from strategy to implementation. Declare your company name and its variants in Organization schema, so every confirmation below is counted as yours.
2. ChatGPT: the search index, LinkedIn and the institutions it cites (weeks 1 to 3). This is where the gap is largest. Ahrefs found that 88% of ChatGPT's citations come from the general search index, so keep the page title in the buyer's own words and register the site in Bing Webmaster Tools. Semrush measured that ChatGPT cites individual LinkedIn members more than company pages (59% against 41%): have the partners who led the projects publish them under their own names. In this market ChatGPT's answers lean on Danmarks Statistik, Dansk Erhverv, DI, the competition authority and the public procurement portal, so make sure your firm and its public-sector work are visible there under the exact company name.
3. Claude: lists, overviews and registers (weeks 2 to 6). Orbit Media counted an average of 3.6 sources per Claude answer, mostly review sites and lists. In this market Claude cites Erhvervsstyrelsen, the CVR register, Finans Danmark, KL's Danske Kommuner and LinkedIn. Offer trade media your project record for "how to choose a consultancy" overviews, and check that your registry data matches your website.
4. Gemini: rank in Google and stay fresh (ongoing). Gemini already holds the strongest picture here, and keeping it means keeping the page current. Semrush measured 86% domain overlap between Google's AI Overviews citations and Google's top 10, and Seer found that 72% of cited pages qualify as fresh by their update date. Add each finished project and show the update date.
5. Measure per model after eight weeks. An average hides the split. Ask each model the single-choice question for your topic and check whether its answer, and its sources, now include your proof.
The lesson
In Danish management consulting, the models agree on the frame: the international houses come first, and Deloitte is the default. Where they disagree is the opening. One model already chooses a Danish house for change work, for reasons the firm can document. Two others barely know it.
Your category has its own version of this split. Visibility built in one model is not visibility in the others. Proof placed where each model reads is.
Sources
The measurement behind this edition: window 24 October to 5 November 2026, evaluation and decision resting on 531 and 497 answers from Gemini, Claude and ChatGPT, retention and advocacy on 124 and 104. Criteria, eliminators and buyer language extracted per model per run; the full method is documented at how Seedli measures AI visibility. Implement Consulting Group's founding verified against Lex.
The Sunday Shortlist
Get the decode, and the toolkit
One step left. A confirmation email is in your inbox. Click it, and your toolkit downloads.
That didn't send. Check the address and try once more.
The Sunday Shortlist decodes one category at a time: the questions buyers ask AI, the shortlist that comes back, and the criteria that decide it. Subscribers get every decode first, and the free toolkit gets you started tonight: a worksheet and three prompts to run the first check on your own brand.
One email a week · unsubscribe anytime · Terms