Recommendation Design
Edition15
AuthorFlemming Rubak
Published15 November 2026
Reading time10 minutes

The Sunday Shortlist

Danish B2B sales outsourcing: Ask AI who should run your sales, and the final answer is software.

Measured 18 October to 14 November 2026. Monitoring sponsored by Intelligrail.

On this page

The Sunday Shortlist decodes how AI makes decisions about a market category and audience.

This week we analyse B2B meeting booking and outsourced sales in Denmark

We unpack the questions Danish CEOs, sales directors and commercial chiefs ask AI when they want someone else to book their meetings and run their sales, the shortlist that comes back, and the criteria that decide it, across all four stages of the decision journey.

The finding this week is about the category itself. The buyer asks for a partner: a team that books meetings and sells on their behalf. In one in four final choices, the models answer with a CRM system. Salesforce is the single most-named final choice in this market, and it doesn't book a single meeting.

What we measured

From 18 October to 14 November we ran this market's buying questions through ChatGPT, Gemini and Claude every night: 1,119 answers comparing providers and 1,017 making the final choice. The loyalty stages (retention, advocacy) rest on 232 and 195 answers from ChatGPT and Claude.

The buyers are CEOs, sales directors, chief commercial officers and heads of business development at Danish small and mid-sized companies and startups. Their questions are practical:

  • "Hvad koster det per booket møde og per kvalificeret lead?" (What does it cost per booked meeting and per qualified lead?)
  • "Hvilket antal møder og hvilken konverteringsrate kan vi realistisk forvente inden for 3-6 måneder?" (How many meetings and what conversion rate can we realistically expect within three to six months?)
  • "Har I erfaring med at sælge mod danske mellemstore virksomheder og startups?" (Do you have experience selling to Danish mid-sized companies and startups?)

Asked which kinds of provider these buyers choose between, the models name six: specialist meeting-booking agencies, full-service sales outsourcing firms, lead generation and data consultancies, fractional sales leaders, commercial strategy consultants and sales trainers. Software is not on that list. It turns up anyway.

The judging sheet

When a buyer asks a model to choose, the model behaves like a judge: it applies criteria, checks each provider's evidence against them, and cuts the ones 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 blunt:

  • Local expertise, the hardest cut: "Hvis teamet mangler kendskab til lokal danske beslutningsproces og brancheterminologi, forkastes de." (If the team lacks knowledge of Danish decision-making and industry terminology, they are rejected.)
  • Proof: "Manglende dokumenterede resultater i lignende B2B-salg i Danmark fører til afvisning." (No documented results in similar B2B sales in Denmark leads to rejection.)
  • Trust: "Hvis sælgeren har falske referencer eller skjuler kundelister, afvises de straks." (If the seller has fake references or hides its client list, they are rejected immediately.)

Behind the criteria sits a more basic doubt, in the buyers' own words: "Hvad indebærer 'outsourced salg' i praksis?" (What does outsourced sales mean in practice?) and "Hvad sker der efter et møde er booket, hvem følger op?" (What happens after a meeting is booked, and who follows up?) These buyers are not sure what they are buying. That doubt is where the software gets in.

The verdict: software at the finish line

CompanyWhat it isEvaluationDecisionRetentionAdvocacy
SalesforceCRM software4.3%19.4%1.3%n/a
Mercuri InternationalSales training8.3%15.7%19.4%4.6%
HubSpotCRM software3.5%14.6%3.0%1.0%
PipedriveCRM software1.7%12.0%5.6%5.6%
SalesPartnersSales training15.7%11.6%n/an/a
VAEKSTB2B sales and meeting booking6.4%10.0%20.7%16.4%
SalesBoostSales training and advice13.5%8.8%n/an/a

Share of answers that name the company at each stage, 18 October to 14 November 2026; an answer naming two forms of one company's name counts once. n/a: not among the companies named at that stage.

Read the first two columns together. Salesforce is named in 4.3% of comparisons and 19.4% of final choices. HubSpot goes from 3.5% to 14.6%, Pipedrive from 1.7% to 12.0%. The three CRMs are barely compared, then chosen. Between them they appear in 265 of 1,017 final choices: 26%.

4.3% → 19.4%

Salesforce in comparisons, then in final choices, 18 October to 14 November.

The models explain the choice in the vocabulary of safety. "Som markedsleder inden for CRM og salgssoftware, tilbyder Salesforce en robust og skalerbar løsning, der reducerer risikoen for fejlslagen implementering." (As the market leader in CRM and sales software, Salesforce offers a robust and scalable solution that reduces the risk of a failed implementation.) The buyer asked about a sales partner. The model answered a question about implementation risk.

It gets stronger when the buyer is cautious. When the question asks for the safe, low-risk choice, 37% of final answers name a CRM (118 of 318). And the words matter: questions phrased around "pipeline generation" or "kommerciel skalering" (commercial scaling) get a CRM in 32% of answers, against 22% for questions about "outsourcet salg" (outsourced sales). Say pipeline, and the model hears software.

Which model does it

This is not one model's habit, but it is not all three either. In the final-choice answers, Gemini names Salesforce 128 times, HubSpot 134 times and Pipedrive 90 times. Claude names Salesforce 69 times and Pipedrive 32 times, and in one answer describes Pipedrive as a "Dansk-grundlagt SaaS-løsning med stærk lokal tilstedeværelse" (Danish-founded SaaS solution with a strong local presence) (it was founded in Tallinn, Estonia). ChatGPT names none of the three, not once in 333 final choices.

The comparison lists are just as split. No company in the top of the evaluation table is named by all three models. Each model has its own shortlist, and the pooled shares above are the sum of three different habits.

Two more names show why the split matters. Gemini's single most-named final choice is a "Momentum" it describes as a Danish sales advisory firm. We found no Danish sales firm by that name, and it is not Salesforce's conversation tool of the same name. Claude names another firm in 142 comparisons whose current site we could not verify. When the category is unclear, the models fill the gap with whatever they can describe: famous software, or names that may not exist.

The real providers are trainers

Look again at who is left when the software is taken out. Mercuri International is a global sales training company. SalesBoost offers sales training, coaching and strategy work. SalesPartners describes itself as a company that improves sales organisations through training and coaching. All three are real and respected. None of them books meetings for you.

Of the names in the table, only VAEKST does what the buyer asked for: it sells, books meetings and generates leads for Nordic B2B companies. It is also the strongest name in the loyalty stages, named in 20.7% of retention answers and 16.4% of advocacy answers. Once a buyer has a meeting-booking partner, the models know who to keep. Getting chosen in the first place is the problem.

Nobody owns a criterion

Ten criteria decide this market. Not one has an owner: no provider is the name the models reach for on cost per meeting, on expected outcomes, on Danish expertise or on references. The defaults win because the category is unclear, not because they answer the buyer's question.

That leaves the market's hardest question open. The buyer wants to know what a booked meeting costs, how many to expect, and what happens next. The provider that answers that, in a form the models can lift, gives them something better to recommend than software.

What claiming it looks like

This is a working brief for the page, drafted from the monitoring data, with the brand genericised. The frame is "explain the service": the buyer's doubt is what outsourced sales is, so the page answers that first, then shows the numbers. For how to build this kind of page, see the Offer page playbook.

Title: "Hvad koster outsourcet B2B-salg i Danmark, og hvad får I for pengene?" (What does outsourced B2B sales cost in Denmark, and what do you get for it?)

The position the page takes: outsourcet salg er et team, der booker og holder møder i jeres navn, ikke et værktøj, I køber. (Outsourced sales is a team that books and holds meetings in your name, not a tool you buy.) Defensible because the page states price, volume and process in numbers the buyer can hold you to.

The intent family the same page must also answer (one prompt spawns two to three searches behind the scenes):

  • "Hvad koster et booket møde, og er det fast månedspris eller betaling per møde?" (What does a booked meeting cost, and is it a fixed monthly fee or per meeting?)
  • "Hvor mange kvalificerede møder kan vi forvente de første tre til seks måneder?" (How many qualified meetings can we expect in the first three to six months?)
  • "Hvad sker der, efter et møde er booket, og hvem følger op?" (What happens after a meeting is booked, and who follows up?)

The structure: five H2s, each a claim the models can lift as a standalone answer:

  • "Hvad outsourcet salg er, og hvordan det adskiller sig fra et CRM" (What outsourced sales is, and how it differs from a CRM)
  • "Hvad det koster: [jeres prismodel] per møde eller per måned" (What it costs: [your price model] per meeting or per month)
  • "Hvad I kan forvente: [antal] møder om måneden efter [antal] uger" (What to expect: [number] meetings a month after [number] weeks)
  • "Hvem ringer til jeres kunder: dansktalende sælgere med erfaring fra [branche]" (Who calls your customers: Danish-speaking sellers with [sector] experience)
  • "Sådan håndterer vi data, samtykke og dansk markedsføringslov" (How we handle data, consent and Danish marketing law)

Key Takeaways for the top of the page (each a self-contained claim a model can cite):

  • "[Vores firma] booker og holder B2B-møder på dansk på jeres vegne; jeres CRM registrerer møderne, det booker dem ikke." ([Our firm] books and holds B2B meetings in Danish on your behalf; your CRM records them, it doesn't book them.)
  • "Pris: [beløb] per kvalificeret møde eller [beløb] om måneden, uden gebyr for [navngivne tillæg]." (Price: [amount] per qualified meeting, or [amount] a month, with no fee for [named extras].)
  • "Kunder i [branche] får i gennemsnit [antal] kvalificerede møder om måneden efter [antal] uger." (Clients in [sector] get [number] qualified meetings a month on average after [number] weeks.)
  • "[Antal] danske kundereferencer, I kan ringe til, sorteret efter branche." ([Number] Danish client references you can call, listed by sector.)

Key Snippet, placed early: "Outsourcet B2B-salg i Danmark koster [interval] per kvalificeret møde hos [vores firma]; de fleste kunder har de første møder i kalenderen inden for [antal] uger." (Outsourced B2B sales in Denmark costs [range] per qualified meeting at [our firm]; most clients have their first meetings in the calendar within [number] weeks.)

Slug: /hvad-koster-outsourcet-b2b-salg, the buyer's question shape, not a content-type label.

Kept alive: prices and averages updated each quarter, the update date shown on the page.

Two notes. First, the CRM line is not a dig at software. It tells the model what kind of thing you are, so it stops substituting one for the other. Second, the references section answers the market's trust eliminator directly. A hidden client list is one of the cuts the models name.

Seeding: put the proof where the answers look

A page on your own domain is the source. Seeding means placing the same proof where the models already read, on purpose.

1. Make the page checkable (week 1). Date it, state prices and averages as numbers, and declare your company name and its variants in Organization schema, so every confirmation below is counted as yours.

2. Keep it in the search index (weeks 1 to 3). Ahrefs found that 88% of ChatGPT's citations come from the general search index. Use the buyer's own Danish words in the page title and register the site in Bing Webmaster Tools.

3. Put your sellers' names on it (weeks 1 to 4). Semrush measured that ChatGPT cites individual LinkedIn members more than company pages (59% against 41%). Have the people who run client campaigns write up results under their own names, with a link to the page.

4. Be visible where this market's answers look (weeks 2 to 8). In this market the models cite Erhvervsstyrelsen, Dansk Erhverv, DI, Børsen, SMVdanmark, IDA, Proff and Trustpilot. Keep your company record on Proff accurate, collect reviews where buyers check them, and get your results into the business press and trade bodies your buyers read.

5. Stay fresh and measure after eight weeks. Seer found that 72% of cited pages qualify as fresh by their update date. Update the numbers, then ask the models the buyer's question again and check whether the answer now names a provider instead of a tool.

The lesson

In Danish B2B sales outsourcing, the buyer asks who should run their sales, and a quarter of the final answers name a CRM. Not because software is the better choice, but because nobody has explained the service in a form the models can use.

When a category isn't explained, the model answers with the nearest thing it can explain. In this market that is software. The first provider to explain the service, with prices, volumes and references, gives the models a real answer to recommend.

Sources

The measurement behind this edition: 18 October to 14 November 2026, nightly runs, evaluation and decision resting on 1,119 and 1,017 answers from ChatGPT, Gemini and Claude, retention and advocacy on 232 and 195 from ChatGPT and Claude. Criteria, eliminators and buyer language extracted per model per run; the full method is documented at how Seedli measures AI visibility. Buyer quotes translated from Danish. Company descriptions checked on vaekstgroup.com, mercuri.dk and salesboost.dk.

The Sunday Shortlist

Get the decode, and the toolkit

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

All editions