AI Success Starts Before the First Copilot Is Built - How technical implementation companies can win in the AI race
Why AI success requires strategy, governance, business value and advisory before technical implementation begins.
Table of Contents
Why the tech market has changed and what the implications are
Five years ago, organizations primarily contracted with technology partners to deliver technical implementations. Purchasing decisions focused on implementation expertise, delivery capability, references and scale. It was all about excellence, best practices, how certain technologies are implemented, how much experience a certain partner has and at which scale it can operate. As outlined in other articles the process is almost always similar: either an existing customer starts a new project with a tech partner that this particular organization already is working with, or there is a tender/RFP where many companies can hand in offer documents. Then after some back and forth a company is chosen, a project team is assembled and the projects start. Technology is implemented, potentially with some after sale, .. but that’s it. Nothing spectacular, nothing new, all routine for the last 20 years.
However, nowadays the world is changing. Fast. Where does AI fit into this process? This is an important question as modern AI, LLM based applications that is, is new for most organizations and technology implementation companies. AI initiatives are often structured around questions like: “Which use-case is really relevant and worth the investment?”, “How do I measure the ROI of the AI investment?”, “What are the risks of this technology for my bottom line?” .. Long story short: Traditional implementation-centric delivery models address only part of these questions. Organizations increasingly expect strategic guidance before implementation begins.
The AI buying journey
Consequently modern AI is, in most organizations, structured differently. An exemplary lifecycle is more structured like this:
- Business Problem (clearly defined)
- AI opportunity assessment
- Strategy & roadmap
- Governance & legal compliance (especially important because of AI Act and NIS2, GDPR)
- Business Cases (P&L justification towards the CFO office)
- Technology selection (Microsoft, Anthropic, Google etc.)
- Implementation
- Adoption by the organization
- Continuous improvement / further projects
We see that the point “implementation” comes way down the line in the process of most organizations. The impact of this market shift is larger than it may initially appear. To get to step #7 all the previous steps have to be taken first. And this can, for traditional IT companies, pose a problem. This is not related to size, tech stack focus but more to strategy of how to approach customers and create real value for them (measurable in currency).
Why implementation expertise alone is insufficient
AI is a huge market opportunity and a lot of companies who offer IT want to get a big chunk of the cake, plus numerous startups who come to the market. In the last 20 to 30 years traditional IT was all about implementation, which was fine until then. Technical excellence alone was often sufficient to build a successful implementation business.
However, nowadays the focus shifts, before anything is implemented.
Modern organizations demand experience and advice in business transformation, as AI will be very disruptive in many industries. Customers increasingly seek advice on how AI will reshape their industry and how their organization should respond. Those kinds of answers are typically given by strategy firms, that nowadays (unsurprisingly as their business model is also disrupted by AI) shift more and more towards implementation. Leading strategy consulting firms already make a significant amount of their revenue with AI projects. Their investments and hiring strategies indicate an increasing focus on combining AI strategy with implementation capabilities: They look for AI engineers & for strategy consulting. So they combine strategy and execution, or give advice with whom a customer should implement. This is an important insight when leading consulting firms expand their portfolio to implementation or at least leading the first stages of AI projects also on an operational basis.
For traditional system implementers/integrators this means that they might, to not be left behind, need to step up their game, expand their offerings, and offer more upfront advisory services themselves. Topics could range from organizational change because of AI, Security which becomes very important in today’s times, data readiness/governance and protection, regulatory compliance, executive alignment and concrete ROI justification & measurement of financial impact of AI (in concrete hard currency instead of just estimations and presumed value). Those are just the core topics and I could mention 20 more here, but I assume the direction is clear. Implementation remains essential, but it is increasingly no longer sufficient on its own.
How advisory creates influence and wins projects
In offering advisory services technical implementers not only get earlier to the table, but also increase their influence on decision making processes. When advisory services help shape corporate priorities, architecture, roadmaps, governance, vendor selection (so tech agnostic advisory) and related areas trust by customers will increase. In many cases, the advisory partner significantly influences later implementation decisions of who gets “invited” later to implement the AI in the whole corporation.
Winning strategy of modern AI projects
The firms most likely to succeed will combine strategic AI business consulting, AI strategy, governance, cybersecurity (very important as we see a huge data breach now each week), enterprise architecture & how to embed AI into the core enterprise architecture/design, technical implementation and organizational change management into an integrated offering. Core should be a strong business consulting branch that can combine business consulting, AI strategy, governance & legal compliance, enterprise architecture, AI experience (ideally from Microsoft :-) , implementation experience and adoption and change management. So…, people in those teams need a wide range of skills ranging from concrete compliance knowledge (so GDPR, AI Act, NIS2 etc.), to technical AI expertise in form of projects, experienced enterprise architects with international experience in multi country organizations are also needed next to risk- and change management experts.
Once the core consulting is done implementation will start. Without this approach it might be more luck than skill that determines if a technology implementer company grows in the era of AI or if it just loses market share/deals against competition. From an operational perspective such an approach also has the plus that the tech partner already knows the needs and goals of the customers in detail. If an RFP for a rough cost estimate comes in, the team should be able to deliver a rough estimate in 24h instead of 1 week/2 weeks/ 3 weeks … and then the competition is already implementing.
Technical implementers however have one core advantage over strategy firms like the ones mentioned above! It is their long history of technical projects. Experienced/Lead people in those firms have seen it all: Deployments crash on a Sunday morning 2am, data migrations to production gone horribly wrong despite being tested 5x beforehand in TEST, never ending requirements sessions and misalignment between departments of the customer that lead to contradicting requirements for the software, extreme time pressure to deliver a technical solution if something breaks in production and so on and so forth… All those “battle scars” deliver an incredible sense of reliability and trust in the statements made by a technical advisory/implementation firm. When an experienced Azure or Security architect makes a statement based on 15 years experience that a certain bullet point on a slide “will never work and break the architecture”, no strategy consultant will ever be able to challenge this. This kind of reputation and trust a strategy firm will never be able to get by any stretch of imagination. That is why combining all those strengths with the advisory part can be such a big leverage.
There is however a downside. Hiring! If a technical firm has operated for decades on technical implementation alone hiring respective experts could be a challenge. Not only because those experts often demand high salaries (try to find an AI governance expert who has global exposure, project experience in Fortune 100/500 and also has technical skills), but also demand environments to grow and meet like minded people. If this all needs to be built up that could be a challenge.
Microsoft Ecosystem perspective
As the author works in the Microsoft stack it’s clear that a part has to be said about Microsoft. Microsoft offers a wide range of very good AI tools from Azure Foundry, to M365 Agents and the more and more integrated MS Purview for governance. All those tools are super powerful if applied and used correctly.
However, the main focus should not be on the tooling - so on the “how to do things”, but rather on the “why to do it” or “should I do it”… so on the processes, business value, risks and how governance is organized around the tooling.
Professional observation, summary and wrap-up
I want to close the article with a couple of thought provoking questions for the readers. What do you think? Has the role of AI changed how consulting and technical implementations are done these days? Are technology partners ready for this substantial market shift? How can a traditional IT company shift its mindset from traditional implementation only towards a mixture of consulting and implementation?