As AI accelerates, companies are faced with the challenge of developing an AI strategy that’s sustainable. It’s a crucial step in the AI adoption process. However, the idea of an AI strategy can be nebulous, and most companies don’t know where to start so they elicit outside help.
Consulting firms typically approach AI in one of three ways, each with its strengths and shortcomings. These three approaches are highlighted below along with our perspective on how to prioritize long-term capability over short-term flash.
This is the “starter kit” approach: organizations are provided a catalog of generic AI use cases and/or tools organized by function (think invoice processing in finance, resume screening in HR, predictive maintenance in operations, etc.). These decks are often well-packaged and tools seem easy to implement. The problem is, tools are not organization-specific and they don’t account for the quality of data at the company. It’s all quite generic. In the end, most organizations end up back at square one.
Pros:
Cons:
Here, consultants deploy a ready-to-use AI toolset—often a retrieval-augmented generation (RAG) engine or a proprietary platform—then look for problems it can solve. This can work for regulated industries (insurance companies or legal firms) where everyone has common models. But if a construction firm uses a pre-built tool for building a commercial high-rise, that tool won’t work for an offshore facility, for example. Another example is E&P land records. From the outside, it may seem like land records are pretty generic, but they’re not. A prebuilt tool that claims to analyze land records and guide the revenue cycle won’t work for every E&P company.
As a company with specific processes, there are unique aspects of the business that require more customization. Investing solely in a pre-built tool is risky.
Pros:
Cons:
The third and most sustainable approach starts with the business: identifying real, high-impact use cases grounded in organizational goals and available data. This involves a context-aware approach where tools are selected based on what the team can use and evolve on their own.
This approach involves stepping back, creating a blueprint before jumping in, and truly understanding the intricacies of the organization with the end goal in mind. With the right strategy, tools can then be built to fit the organization’s specific needs, or a hybrid approach might be necessary where there’s a mix of pre-built and custom tools.
Pros:
Cons:
Most companies today are seeing a mix of the first two approaches (playbooks followed by quick demos to generate momentum). But forward-thinking organizations are pushing for more. They want ownership, adaptability, and internal capability—aka the engineering mindset.
In the first approach, companies are limited to that generic binder of use cases. In the second, companies are directed to a specific type of solution or a specific use case. In the third, the sky’s the limit.
Companies that win will be the ones that treat AI like engineering, not magic.
To chat more about AI strategy, reach out to us at info@trenegy.com.
Trenegy is a non-traditional consulting firm based in Houston, Texas.