AI-Native Consulting: The Future of IT Consulting

The IT consulting industry is facing a structural shift. For decades, consulting firms have relied on the pyramid model: many juniors, few partners, scaling through headcount. Now, companies are emerging that build their business model from the ground up around artificial intelligence. Not as an add-on tool, but as an operating system. This changes not only the cost structure but the entire logic of the industry.

The Key Points at a Glance

• 📊 AI consulting will grow by 13.9 percent, the fastest in the German consulting market by 2025 (BDU 2025).
• 🔧 AI-augmented consultants deliver 40 percent higher quality at 25 percent faster processing (Harvard/BCG 2023).
• 💡 AI-native firms replace the classic pyramid with flat obelisk structures featuring three new roles (HBR 2025).
• ⚠️ Over 40 percent of all agentic AI projects will be abandoned by the end of 2027, according to Gartner.
• 🏢 The German consulting market will exceed the 50 billion euro mark for the first time in 2025 (BDU 2025).

The Consulting Market is Shifting

The Federal Association of German Management Consultants (BDU) reports a historic milestone for 2025: The German consulting market exceeds the 50 billion euro mark for the first time, reaching 51.8 billion euros. This represents a growth of 6.4 percent compared to the previous year. Within this market, one segment stands out: AI consulting is growing by 13.9 percent, more than twice as fast as the overall market.

However, these numbers only tell half the story. Because the growth is not evenly distributed. While large consulting firms are supplementing their existing processes with AI tools, companies are simultaneously emerging that work with AI as their core architecture from day one. The Harvard Business Review describes this difference as the shift from the pyramid to the obelisk.

Pyramid vs. Obelisk: Two Models Compared

In September 2025, authors David S. Duncan, Tyler Anderson, and Jeffrey Saviano analyzed in the Harvard Business Review how AI fundamentally changes the structure of consulting firms. Their key observation: The classic pyramid model with many junior consultants at the base and few partners at the top is under pressure because AI automates exactly the tasks that were previously handled by juniors.

The alternative they call the obelisk model. Instead of a broad base and narrow top, a slim, consistently high-quality structure is created. Fewer hierarchy levels, fewer heads, but each individual with significantly higher output. The consequence for personnel planning: An AI-native team with five seniors can handle projects for which classic firms would deploy ten to fifteen consultants.

Sources: BDU Facts & Figures 2025, Harvard/BCG field experiment 2023

According to the HBR authors, three new core roles emerge in the AI-native model. AI Facilitators operate the AI infrastructure and data pipelines. Engagement Architects define problem statements, interpret AI outputs, and translate them into actionable recommendations. Client Leaders maintain the trust-based relationship at the C-level. None of these roles can be filled by a beginner.

The model has consequences for training. Someone who starts today as a cloud consultant no longer begins as a research analyst who builds PowerPoint slides. Instead, the entry must start at a higher level: with the ability to critically evaluate AI-generated results, bring contextual knowledge, and independently manage customer relationships.

What the data show: 40 percent more quality

The most challenging study on this topic comes from a collaboration between Harvard Business School and Boston Consulting Group. In a randomized field experiment, hundreds of BCG consultants gained access to GPT-4. The results, published as a Working Paper by HBS in September 2023, are noteworthy.

AI-supported consultants delivered an average of 40 percent higher quality in their outputs. They completed 12.2 percent more tasks and were 25.1 percent faster. Most importantly: consultants who previously performed below average improved by 43 percent. The already strong consultants managed to add only 17 percent.

For the pyramid model, these numbers are a problem. If a senior consultant with AI support becomes almost as productive as a three-person team of juniors, the traditional scaling through headcount loses its legitimacy. The question is no longer whether AI will be used in consulting, but whether the existing business model can survive.

”

// Quote

Confuse knowledge workers with knowledge work. The knowledge work of today may likely be automated. But who said my goal in life was to sort my emails?

Cloud Consulting as a Test Field

The change is particularly visible in the Cloud consulting. Multi-cloud migrations, FinOps optimization and Kubernetes orchestration are areas where AI tools already deliver measurable time savings. Generate Terraform code, validate CloudFormation templates, detect anomalies in cloud spending data: These are repetitive tasks with clear rules that were classically handled by junior consultants.

The practice shows the effect already. The recruitment agency PageGroup reports that it has accelerated the creation of job advertisements by 75 percent with Azure OpenAI. Presentation content is curated 50 percent faster. Microsoft documents this figure in its Cloud Blog as one of over 1,000 customer cases. For cloud consultants, this pattern transfers directly: Writing offers, creating documentation, conducting code reviews. All of this can be done with AI support in a fraction of the previous time.

In the DACH region, the trend is particularly evident in specialized boutique consulting firms. Instead of building up 30 or 50 consultants, they work with teams of five to ten seniors who multiply their productivity through AI tools. The calculation behind this is simple: Fewer salaries, less overhead, but higher margin per project. At the same time, customers benefit from shorter project durations and direct access to experienced consultants rather than to newcomers who still need to be trained.

The advantage is particularly clear in AIOps projects. Here, AI analyzes monitoring data, detects anomalies, and suggests countermeasures. An experienced cloud architect only needs to classify and approve the recommendations. This significantly reduces the personnel effort for ongoing operations.

The Downside: Where AI Augmentation Fails

The Harvard/BCG study, however, also delivers an important warning. The researchers describe a so-called “Jagged Technological Frontier”: a jagged boundary beyond which AI not only does not help but harms. For certain complex strategic decisions, the AI-assisted advisors performed worse than their colleagues without AI access. The reason: They trusted the AI recommendations, although they should have doubted them.

Gartner reinforces this skepticism with concrete predictions. According to a press release from June 2025, more than 40 percent of all Agentic-AI projects will be abandoned by the end of 2027. The reasons: escalating costs, unclear business value, and inadequate risk controls. As early as 2024, Gartner predicted that 30 percent of all Generative-AI projects would be discontinued after the proof-of-concept phase.

For AI-native consulting firms, this means: technological advantage alone is not enough. The judgment of when to use AI and when not to will become the decisive competence. It is the difference between a consultant who masters AI as a tool and one who blindly trusts it.

What Customers Really Buy

From the customer’s perspective, something fundamental is also changing. Deloitte reports in its “State of AI in the Enterprise” from 2026 that 66 percent of the organizations surveyed have already seen productivity increases through enterprise AI. The study is based on a survey of 3,235 senior leaders between August and September 2025. At the same time, McKinsey’s State of AI Report 2025 shows that only 39 percent of organizations can demonstrate measurable EBIT impacts through AI.

This discrepancy is revealing. Companies feel that AI makes their teams more productive. But they often cannot prove it in hard business numbers. This is exactly where the opportunity lies for consultants who can not only implement AI but also make its value contribution measurable.

For cloud projects, this means specifically: An AI-native consultant who performs a FinOps optimization not only delivers the technical implementation. He can simultaneously calculate the ROI with AI-supported analysis, draw benchmarks against comparable companies, and generate a decision template for the CFO. What previously required three consultant roles is handled by an expert with the right tools.

Five Questions for IT Decision-Makers

Those who choose a cloud consultant today should not only ask about certifications. Five questions help with the assessment.

1. How does the consultant use AI in their own projects? Someone who only sells AI as a product to customers but does not use it internally has not understood the model. Ask about concrete internal use cases.

2. What does the team structure look like? Check whether you are paying for a project manager and four junior consultants or for three seniors who achieve the same performance with AI. The daily rates can be similar, but the quality of the results differs significantly.

3. Where does the consultant draw the line? A good AI-native consultant can explain which tasks they automate and which they deliberately handle manually. Knowing the “Jagged Frontier” is a quality feature.

4. How is the value contribution measured? Demand quantifiable KPIs. Not “we have rolled out Copilot,” but “cloud costs have decreased by 23 percent with a simultaneous 15 percent increase in deployments.”

5. How does the consultant handle sensitive data? AI-native consulting means that company data flows through AI models. Ask about data sovereignty, self-hosted models, and compliance concepts. This is especially important for NIS2-required companies and is not just a nice-to-have.

The DACH market is moving

McKinsey’s State of AI 2025, based on a survey of nearly 2,000 organizations from 105 countries, shows that overall AI adoption is rapidly increasing. 88 percent of the surveyed organizations now use AI, an increase of ten percentage points compared to the previous year. For Generative AI, usage has more than doubled from 33 percent to 72 percent. At the same time, only 7 percent have scaled AI across the entire enterprise.

For the German Mittelstand, this results in a paradoxical situation. The willingness to invest in AI is high. But the ability to successfully implement AI projects remains limited. According to Gartner, 63 percent of organizations lack suitable data management practices for AI projects. The BCG AI Radar 2026 predicts that companies want to double their AI spending: from around 0.8 to 1.7 percent of revenue.

It is precisely this gap that AI-native consulting firms fill. They bring not only the technical know-how but also the methodology to accompany AI projects from the concept phase to measurable value creation. The advantage over the large firms: faster decision-making processes, less overhead, and daily rates that are affordable for the Mittelstand. Instead of waiting six months for a study, an AI-native team delivers a functioning proof of concept in six weeks.

How this works in practice is shown by the Hamburg-based alfatier GmbH. Founder Kim Nis Neuhauss knows both sides: As CEO of Bright Skies, he built one of Germany’s strongest Microsoft-Azure partners, with 55 employees, around 500 customers, and 14 Microsoft-Gold competencies. Neuhauss himself was twice awarded Microsoft MVP for Azure. After the acquisition by Rackspace Technology in 2020, he started anew with alfatier, this time AI-native from the ground up.

The team consists exclusively of seniors and covers four pillars: Cloud Excellence, AI Transformation, Cyber Resilience, and Digital Sovereignty. The combination of Cloud and Security is particularly relevant: With the NIS2 registration requirement and the parallel DORA pressure on financial service providers, SMEs need consultants who can think about cloud architecture and compliance requirements at the same time. Those who offer both from a single source save their customer a second consulting firm.

Instead of a consultant pyramid, Neuhauss relies on a lean model that does not treat AI as an add-on, but as the operating system of the entire consulting service. alfatier achieved seven-figure sales within two years.

”

// Quote

The future does not belong to the best resume. It belongs to the best person.

This example illustrates a broader trend in the DACH region. Experienced cloud professionals, who know large organizations from the inside, found lean companies that tackle projects with AI support, for which classic firms would use significantly larger teams. For medium-sized companies that need enterprise quality but do not have enterprise budgets, this model is becoming increasingly attractive.

Conclusion: The future belongs to judgment

AI-native Consulting is not a trend that will disappear. The economic logic is too strong. If a team of five with AI support delivers the results for which 15 consultants were previously needed, this changes the entire competitive dynamics of the industry.

But the technology alone does not decide. The Harvard/BCG study shows clearly: AI makes good consultants better and bad consultants more dangerous. The decisive competence will not be to operate AI tools. Rather, it will be to know when to trust them and when not to. This distinguishes the AI-native consultant from a prompt operator.

For IT decision-makers, this means: Do not commission the largest consulting firm and also not the one with the most impressive AI demo. Rather, the one whose people bring the judgment to use AI where it creates value and to stop where it causes harm.

Frequently Asked Questions

What does AI-native Consulting mean?

AI-native Consulting describes consulting firms that have built their business model from the ground up around artificial intelligence. Unlike traditional consulting firms that use AI as an additional tool, AI-native firms use AI as the core of their approach. Typical are small senior teams without a classic junior pyramid.

How does the Obelisk model differ from the Pyramid model?

The Pyramid model scales through headcount: many junior consultants at the base, few partners at the top. The Obelisk model scales through technology: fewer, but more experienced consultants who multiply their productivity through AI tools. The Harvard Business Review describes the Obelisk as a lean, consistently high-quality structure with three new roles: AI Facilitators, Engagement Architects, and Client Leaders.

Is AI-augmented consulting always better than classic consulting?

No. The Harvard/BCG study shows that AI-supported consultants performed worse than their colleagues without AI on certain complex strategic decisions. The researchers describe a “Jagged Technological Frontier” with tasks that lie outside of AI’s competence. The ability to recognize this boundary is crucial.

How can IT decision-makers evaluate AI-native consultants?

Ask about internal AI use cases (does the consultant use AI themselves?), the team structure (seniors instead of a junior pyramid?), the data protection concept (self-hosted models?), and measurable KPIs (not just “we use Copilot”). A good AI-native consultant can clearly name where they use AI and where they consciously do not.

What does AI-native Consulting cost in comparison?

The daily rates of individual consultants can be comparable or even higher than at classic firms. The difference lies in the team size: Instead of four to six consultants for a project, an AI-native team often only needs two to three. In the end, boutique consulting firms report 30 to 40 percent lower total project costs for their customers.

Reading tips from the editorial team

More from the MBF Media Network

Image source: Pexels / cottonbro studio