From No. 5 to No. 1, ahead of Designit, Accenture Song, McKinsey and Frog

When Manyone began using 3RD, the company ranked fifth in strategic design. Four months later, Manyone was at the top and had achieved the greatest progress among all 37 organisations tracked by the platform in the category.

Key figures

Baseline: 24 April to 24 May 2026. Latest measurement: 18 July to 17 August 2026.

What was at stake

Manyone is an international strategy and design company with offices in seven countries. The company's clients are often leaders and transformation executives in large organisations, and a place on the right shortlist can have significant commercial value.

Those shortlists are increasingly influenced by AI.

A potential client might ask: “Who should we consider for design-led transformation in the Nordics?” The response typically condenses the market into a short list of companies. The companies that are mentioned move on to the buyer's research. The others risk never being considered.

The development can be seen clearly in the B2B software market. In a March 2026 survey of 1,076 B2B software buyers, 51% said they were more likely to begin their research in an AI chat than in Google. 71% used AI chatbots at some point in their research, while 69% had chosen a different supplier from the one initially planned based on guidance from an AI chat. One in three had bought from a supplier they did not know beforehand. The survey concerns software purchases, but it illustrates how much influence AI has already gained over B2B research and shortlists. G2, The Answer Economy, 2026

For Manyone, the problem was that this part of the customer journey could not be seen in the existing analytics tools. There was no consolidated report showing how often the company was mentioned, which competitors were recommended instead, or how its position varied between categories, markets and languages.

At the same time, Manyone competes with companies such as McKinsey, Accenture and Deloitte. These are brands with an enormous amount of historical coverage, content and external authority for language models to draw on.

Without data, it was natural to assume that the largest companies would also automatically dominate AI-generated responses.

How Manyone used 3RD

Manyone and 3RD began by configuring a market that reflected the company's actual business and the way its clients ask questions.

The setup included:

• 1,470 prompts phrased in the way buyers ask their questions

• Five languages: Danish, Swedish, Norwegian, German and English

• 12 markets across seven countries

• Nine topics matching Manyone's business areas

• 37 competitors and comparable organisations

• Daily measurements over more than 100 days

The prompts were divided into three layers: must-win, core and foundation.

The must-win layer contains questions with clear purchase intent, where a user compares suppliers or puts together a shortlist. The core layer covers Manyone's central business areas, while the foundation layer tracks broader research around the market.

This division made it possible to distinguish between general visibility and visibility at the points when a buyer is approaching a decision.

Manyone's own team then continued working in 3RD.

The platform collected the AI-generated responses, measured development against the 37 organisations and showed where Manyone gained or lost visibility. Data and tasks were brought together and prioritised in the platform so the team could choose what to work on first.

Manyone interpreted the data itself, selected the tasks and carried out the work. The work included structured organisational data, a Wikidata entity and an FAQ structure, which made central parts of Manyone's expertise easier for the models to find and understand.

This required neither a new internal department nor a lengthy external agency engagement. Manyone used 3RD as a tool and completed the tasks with the team and expertise the company already had.

During the same period, Manyone moved from No. 5 to No. 1 in strategic design.

The result

From No. 5 to No. 1

At the first measurement, Manyone was at the back of the leading group. Four months later, visibility had increased from 18.0% to 26.9%, corresponding to growth of 49%.

Manyone had thereby overtaken Designit, Accenture Song, McKinsey and Frog.

The increase of 8.9 percentage points was also greater than that of any other of the 37 organisations tracked in the category.

The greatest growth came from purchase prompts

Manyone's visibility grew across all three prompt layers. The largest percentage increase came in the must-win layer, where buyers are actively researching and comparing potential suppliers.

Here, visibility increased by 45%, from 3.8% to 5.5%.

This is important because general coverage and visibility close to a purchase decision do not have the same value. The must-win prompts show whether Manyone appears in the responses where a potential client is deciding whom to investigate more closely.

Progress in several categories

The clearest position is still in strategic design, but the development is not limited to the main category.

In industrial design, visibility almost doubled, and Manyone reached approximately the same level as McKinsey. In brand design, the company moved ahead of Frog, McKinsey and AKQA during the measured period.

The figures are still lower than in strategic design, but the movement shows that Manyone's position can be extended to more of the areas in which the company works.

More specific buyer questions led to Manyone

In addition, Manyone was now mentioned for 17 questions where the company did not appear at all during the baseline period.

Sentiment around Manyone was also consistently positive. The company had a sentiment score of 86.9, and only 19 out of nearly 3,000 registered mentions were negative.

The value for Manyone

Manyone has gained a documented position in strategic design that can be tracked from day to day and compared directly with named organisations.

The company now knows which questions it is recommended for, where its position is strongest and which areas have the greatest potential for further growth.

With 3RD, Manyone can track the development itself, see where the potential is greatest and have the tasks that can move its position generated and prioritised. With agentic fulfilment, the team can also execute tasks directly in the platform.

This gives Manyone a consolidated workflow from measurement and prioritisation to execution and follow-up.

Strategic design can be supported as a documented position of strength in sales and marketing. In areas such as business transformation and AI enablement, Manyone can use data to investigate why other companies are selected more often and which signals can be strengthened.

At the same time, the development can be tracked across markets and languages. This allows the team to work with the areas where the potential and commercial relevance are greatest.

3RD made the market and the development measurable. Manyone itself used the insights and tasks from the platform to continue working on its position.

The case therefore shows that an internal team can create measurable progress with 3RD in its toolkit, without building a separate GEO function or outsourcing all the work to an agency.

The sources revealed a brand risk

3RD also made it possible to examine the sources used by the AI models in their responses.

85% of the registered citations came from external sources. This is positive in principle because independent coverage can strengthen the company's credibility.

However, the source analysis showed that the most-cited source about Manyone was Glassdoor. The other prominent results included a job site for “Many One L.L.C.”, which belongs to a different US company.

The models' foundation therefore consisted both of relevant external coverage and sources that provided only a limited or inaccurate picture of Manyone's expertise.

This is a type of brand risk that rarely appears in traditional marketing and web analytics. Manyone could now see it and prioritise the work on the sources the models actually used.

The analysis identified, among other things, four industry lists that together were cited more than 110 times per month for relevant buyer questions. Between one and five of Manyone's competitors appeared on each list, while Manyone was not yet represented.

This gave the team four specific opportunities to pursue. Each placement was selected based on documented relevance in AI-generated responses, rather than a broad assumption that more content or more coverage would automatically create an effect.

What the Manyone case shows

Positioning is also shaped in AI-generated responses

Companies still determine their strategy, narrative and desired market position. But a growing share of buyers' first impressions is created by the way AI models describe and recommend them.

There can be a gap between the two pictures.

That gap is difficult to detect without continuous measurement because the conversation takes place outside the company's own channels.

An internal team can move the position

Manyone went from No. 5 to No. 1 in four months, while the internal team itself selected and completed the tasks based on 3RD.

The work was based on existing knowledge and expertise. The platform showed where the information lacked structure, where the models had an unclear picture and which tasks were most relevant to address first.

This makes the GEO work more operational. The team can interpret the market, choose its efforts and track development in the same tool.

Size does not necessarily determine the category

Manyone achieved greater progress than all other measured organisations in strategic design and moved ahead of companies with significantly greater global reach.

The result shows that competition in AI-generated responses can be category-specific. A specialised company can build a stronger position in a particular area than a competitor that is larger overall.

This creates a genuine opportunity for companies with clear expertise and documentation, even if their overall marketing presence is smaller.

Individual AI-generated responses can be misleading

Manyone's total daily visibility fluctuated between 1.0% and 8.5% during the period.

A measurement on 14 August would have shown 1.4%. A measurement on 31 July would have shown 8.5%. Both measurements were correct on the day, but neither alone provided a representative picture of the development.

Stable insight emerges through repeated measurements across prompts, models, markets and languages.

Do you know how your company stands in AI-generated responses?

3RD shows how often your brand is mentioned and recommended, which competitors take its place and which tasks can move its position.

The platform enables your own team to measure the market, prioritise the work and track the effect over time.