AI Strategy & Business Case

AI Readiness: Technology, Data and Organization

Marin Niehues18 November 202410 min read

Woman drawing a system diagram

AI readiness describes whether a company's prerequisites in technology and data as well as in its organization are sufficient for an AI application to deliver reliable value in daily operations. The assessment becomes robust enough for budget and sequencing when it is made per application, because a customer service assistant has different prerequisites from a forecast in controlling. This article provides an overview and does not constitute legal advice.

How many companies are ready for AI?

Roughly one in eight companies in Germany with at least ten employees uses AI technologies. This is shown by the official survey on ICT usage in enterprises. Large companies with 250 or more employees use AI almost three times as often as the average.

The reasons given by the remaining companies are revealing. Most often they cite a lack of knowledge, followed at some distance by incompatibility with existing systems and difficulties with the availability or quality of data. The obstacles are thus spread across all three dimensions, and the largest is an organizational one. A readiness assessment therefore addresses all three dimensions and gives organization the same weight as technology and data.

How do you assess AI readiness?

You assess AI readiness by answering ten review questions from the three dimensions for each planned application and assigning each answer to one of four levels. The starting point is a view of the company as a whole: which systems and data sets exist, and who is responsible for them? For budget and sequencing, however, the assessment only becomes robust when it is applied to a specific application.

To do this, you choose an application, for example the pre-qualification of service requests, and check what is missing between prototype and regular operation. The answer differs for each application. The sum of the checks produces a sequence that concentrates investment where several applications share the same gap. The company's AI maturity then follows from these individual findings.

Technology: integration and operations

The technology dimension asks whether the application can be connected to the systems in which the work takes place. Three review questions are enough for an initial assessment. Are there interfaces to the required source systems that an application is permitted to query? Can the solution be embedded in the business units' tools, so that nobody has to switch programs to use it? Is operation organized, with monitoring and incident handling assigned to a named person or team?

Legacy IT landscapes usually stall at the first question, because data sits in systems that offer no modern interface. This can be solved and requires its own planning. It therefore belongs in the assessment before a business case is signed off.

Data: availability and ownership

The data dimension covers three review questions. Is the data the application needs available in a form it is permitted to access? Is the quality sufficient for the purpose, above all in completeness and timeliness? Is there a person responsible for maintaining and releasing each relevant data set?

The third question is the one most often underestimated. A data set without an owner deteriorates unnoticed, and after a few months the application delivers worse results than in the pilot. Where personal data is involved, a data protection review by specialists belongs in the plan before operation begins.

Organization: ownership and decision paths

The organizational dimension most often decides whether a working prototype becomes an operational solution. From a delivery perspective, four questions arise. Does the application have a business owner with budget and decision rights who is accountable for the benefit in their own area? Who decides on changes and shutdown during operation? Are the affected processes described in a way that allows the application to fit into them? Do employees have the AI literacy that the European AI Act requires of companies, and can they check and correct results?

If any of these answers is missing, the pilot remains a one-off. Where co-determination rights are affected, for example with applications that make employee performance or behavior visible, consultation with the works council belongs early in the timeline.

How can you tell whether your company is ready for AI?

Your company is ready for an AI application when all ten review questions in the following checklist are answered at level 2 or higher, meaning every gap has a plan with an owner and a deadline. The checklist brings together the questions from the three dimensions.

How a readiness assessment works

In the first step, the two to five most promising applications are described with the business unit: goal, workflow, expected benefit and owner. Interviews with IT and data owners, together with documentation on systems and data sets, complete the picture. The result is an assessment per application using the scheme above, from which the executive team reads which application starts first.

In the second step, the most critical gap of one application is tested with a small prototype in the real workflow. A prototype answers questions that no questionnaire can: how good is the data really? Does the business unit accept the result? In the third step, the transfer into operations follows, with clear responsibilities and training. The results produce the roadmap for implementation.

Niehues Consulting as a partner for the readiness assessment

Niehues Consulting assesses AI readiness per application and supports companies from assessment to operation. As a management consultant, Marin Niehues builds organizations that deliver and brings AI into operations. Projects range from VC-backed scale-ups to DAX 40 groups, in 20+ countries on six continents.

The approach follows the Niehues Delivery Model with its phases Understand, Build and Deploy. Three review questions matter most for readiness: who decides, who delivers, who uses? Does the prototype prove its benefit in the real workflow with real data? Who owns operations from day one?

The assessment belongs to the AI management service, which runs from business analysis through the business case to the selection and sequencing of applications. Prototypes follow in the same approach, as do operationalisation and deployment, so that assessment and delivery come from a single source. The entry point is a free 30-minute initial consultation, in which the first application and its most likely gap are discussed.

Frequently asked questions

What does AI readiness mean?

AI readiness is a company's capability to use a specific AI application in regular operation. The assessment covers technical integration, the data situation, and responsibilities and skills in the organization. The result differs for each application.

What is the difference between AI readiness and AI maturity?

AI readiness checks whether the prerequisites for a specific application are in place. AI maturity measures how reliably the company as a whole already brings applications into operation. Readiness is the question before the start, maturity the question after the first initiatives.

How long does an AI readiness assessment take?

For two to five applications the assessment takes four to six weeks, provided the interview partners from IT and the business unit are available. A prototype to test the most critical gap follows as a separate step.

Who should be involved in the assessment?

The executive team or divisional management appoints the sponsor. The business owner of the application, IT and the data owners take part. Data protection and the works council join as soon as personal data or co-determination rights are affected.

What data does a company need for AI?

A company needs the data that the specific workflow of the application produces or uses, in an accessible form and of sufficient quality. For a customer service assistant, that means requests, responses and knowledge articles, for example. A company-wide data program before the start is not required.

What IT infrastructure does AI require?

For the first application, AI requires interfaces to the source systems, governed access to a model and an operating model with monitoring. A shared platform pays off from the second or third application onwards, as soon as several applications use the same building blocks.

What AI literacy obligations apply under the AI Act?

The AI Act requires providers and deployers of AI systems to ensure a sufficient level of AI literacy among their staff. The yardstick is the technical knowledge of employees and the context in which the system is used. Role-based training with documented participation is the usual way to meet the obligation.

How much does an AI readiness assessment cost?

The cost of an AI readiness assessment is driven by the number of applications and interviews over four to six weeks. The market usually bills by consultant day. Most of the effort goes into the interviews with the business unit and IT.

Let’s talk about your initiative.

In 30 minutes we clarify where your organization stands and which step will pay off most now.

Request a consultation

30 minutes · free · no obligation