Three AI budgets, no shared reckoning

Advertorial · in collaboration with the Evernine Group

Almost every company can quantify its AI expenditure down to the last euro. Few inside the organisation can demonstrate what the money actually delivers in return. According to McKinsey, more than 80 per cent of companies have yet to see any tangible impact from their AI use on business results. A reality check therefore belongs on the table before the next investment.

Related:The hyperscalers’ billion-dollar bet and your cloud bill  /  Turnaround lessons from thyssenkrupp to Continental

Uncontrolled growth is systemic

The CRM was introduced years ago and has since been maintained like a filing cabinet. The AI licences were added in 2025 simply because they became available. Marketing works with a stack of tools that do not recognise each other. Each of these line items has its own contact person, its own budget and its own success story. No one can add them up. Technology accounts for the smaller part of success in AI transformations, the rest is organisation. That is exactly where it gets stuck.

The consequence is visible in every budget round: management is expected to approve further AI expenditure without anyone being able to demonstrate what the previous investments delivered. This is not a side issue. According to a Gartner forecast, at least one in three generative AI projects is discontinued after the proof of concept, usually because of unclear value contribution or escalating costs. Blind flight then leads to stagnation or the next million spent on speculation. Both are expensive.

Reality check with a traffic-light system

What is an AI Check? An AI Check is a compact, numbers-based analysis of a clearly defined area such as sales, market visibility or marketing processes. It takes one to three weeks, assesses the status using a traffic-light system and concludes with a prioritised action list that management can understand without further translation. The decision remains in-house; the foundation for it becomes robust.

The Evernine Group has tailored this format to three areas. The AI Sales Check breaks down CRM, data quality and sales processes and concludes with three to five prioritised use cases along with the figures behind them. The AI Visibility Check measures whether a company appears in the answers of AI search systems and where competitors are already being cited. The AI Efficiency Check examines the marketing stack for silos, duplicate tools and reporting gaps and presents a 90-day roadmap.

Private Equity does not ask for the vision. Private Equity asks what the measure delivered.
– Alexander Roth, CEO Evernine Group (2026)

Visibility now arises in the AI answer

The most underestimated of the three areas is visibility. B2B purchasing decisions increasingly begin with a question to ChatGPT or Perplexity. If a provider does not appear in these answers, part of the market simply never finds it. This can be measured precisely: what questions the target audience asks, whom the systems cite, which gaps can be closed at what cost. A diffuse feeling becomes a prioritised list.

The trend is measurable. Gartner expects classic search volume to decline by around 25 per cent as AI assistants increasingly answer questions directly. Gartner therefore advises companies to focus on content that demonstrates expertise and authority. It is precisely such sources from established specialist publications that AI systems draw into their answers. The AI Visibility Check shows whether and where a company is missing. Closing the gap is then the company’s own task. One of the tools for this is the GEO program from Evernine Media, which places specialist content in precisely these environments and makes its visibility in AI answers measurable.

Whoever measures leads the budget round

There is a second reason why the format works that is rarely discussed. As CIO, CMO or head of sales, whoever substantiates the need for action with a traffic-light analysis leads the internal discussion. The role shifts from defending one’s own budget to taking the initiative in the transformation process. This position can be leveraged in the investment committee, documented to shareholders and recorded in one’s own career progression. An 80-page expert report does not achieve this because no one reads it.

The hard stocktake at the start

It can be argued that analyses rarely solve the problem. The objection applies to the long formats. A reality check is different: it creates the decision-making basis on which implementation can be prioritised at all. Whoever transforms in stages needs precisely this small, hard stocktake at the beginning of each stage. The alternative is implementation by instruction, the results of which no one can then measure.

The first step before the next budget round

The sequence is unspectacular: choose the area with the greatest pain, commission the appropriate check, take the traffic-light rating into the next management meeting. Anyone who knows the growing AI line items in their own cloud bill understands that the question of return will not get smaller. It will only become more expensive the later it is asked.

Frequently asked questions

More from the MBF Media network

Image source: AI-generated (July 2026)