Projecting ROI for Enterprise AI Integration Projects: The Tangibles and Intangibles
How many of you would like to start every day already on a roll?
How do you construct the ROI story for massive enterprise integration projects? Especially ones centered on emerging technologies in the realm of artificial intelligence? Board rooms and management consultants have been buzzing with the AI phenomenon for years, yet the executive teams are struggling to demonstrate ROI. The OEMs of specialized equipment understand that maximizing Industry 4.0 is a necessary next step in order to stay competitive, but the path to do so is not so clear.
One of the unavoidable (and most profitable!) elements of AI is its frequently intangible nature. Successful integrations begin by grappling with a tangible objective, and solving it. Where the serious return originates is from the occurrence that the output of a tangibly solved objective can frequently be reapplied to new and unforeseen value streams.
As explored by the Harvard Business Review, the data have always made it clear that “soft” intangible assets have provided undeniable competitive edge for companies — historically, and to date. In an AI-powered enterprise, these intangible assets become the secret sauce needed for successfully extracting true value from data.
Domain Expertise Drives Data Science – Not The Other Way Around
One of the common flaws is an approach that views AI as fundamentally a Big Data Science initiative. This approach disconnects the rest of the organization from stakeholdership, as it naturally tends to intimidate the true domain experts and business users. The technology becomes an IT project.
A good philosophy is to think of a true AI project being 50% domain. It has to be driven by domain objectives. Some tangible examples in the world of repair and maintenance would be reducing time to repair, improving call-back metrics, generating new sales revenue or improving product durability. These are all areas where AI can make a difference.
AI projects become exponentially more valuable when the business, product, and customer support teams contribute their specific mastery to the overall tapestry. The secret, here, though, is not that each project individually adds to the overall value. It’s that the value created by each business unit to solve their own problem can frequently be reapplied elsewhere in a transformative way. And when there’s a standardized framework for all of these initiatives to operate in, things really take off.
We’ll explain using an example from our experience with assisting enterprises with the repair and maintenance of complex equipment.
When projecting ROI for enterprise AI integrations, the tangible ROI comes from the specific objectives under the Design It, Sell It, Use It, & Fix It stages in the equipment experience. Intangible ROI comes from enabling those insights (the Repair Intelligence) to be fed back into the loop to create new business value.
Focus On “Democratizing” Intelligence - Re-Apply Value & Create a Common Framework
Five years ago, we set out to see whether we could build a system that helps guide technicians through complex repairs on unfamiliar equipment. We built Predii Repair Intelligence™, a platform with a foundational framework that transforms domain expertise and operational data into usable insights.
We processed hundreds of millions of service orders (Fix It), which gave us the historical trends on what fix was applied to which problem. We also took usage data to establish operating norms, in order to understand deviations (Use It). This intelligence was provided to technicians and it cut repair times by 50%, while increasing first-time-fix metrics.
Great, right? End of story?
What we realized was that the knowledge created by understanding historical service orders could be put to a lot more use than just guiding technicians through repairs. Knowing the historical trends for symptoms, failures, and resolutions is incredibly valuable to the quality engineers designing the next generation of equipment – they are deeply interested in understanding equipment performance at a granular component-level. (Design It)
Then we heard from a dispatcher, who mentioned that they would use this information to schedule their technicians with 100% easy repairs for their first appointments, so they could begin every day already on a roll.
The parts salesmen also chipped in, mentioning they could use repair data (in which there is a parts transaction) to assess the market of OEM and off-brand parts to more competitively position their products on the market. (Sell It)
Repair operations, design excellence, employee happiness and parts sales. The same knowledge, applied four times. This only happens when multiple departments are allowed stakeholdership — and, once they are, you never know what new application someone might think of. How many of you would like to start every day already on a roll?
These kinds of ancillary use cases rarely exist in the early project planning of massive AI integrations, but they are certainly there — if you let them. The key is understanding that the equipment experience involves designing, manufacturing, selling, using, breaking, and fixing. The knowledge at each step is incredibly valuable to each of its counterparts.
Select Special-purpose AI Partners Who Can Provide Use-case Scalability
These unexpected new applications are part of the intangible ROI benefits that come with creating a truly democratized AI ecosystem for the enterprise. When every level of stakeholder is involved, you never know what might happen.
Key to making this happen is ensuring that the AI initiative is with a dynamic partner, not one that is themselves so siloed that they struggle to adapt to new opportunities as they present themselves. That partner has to be learning from the data, continuously improving and adding new features, building new value right along with you.
At the same time, enormous cost savings occur when they have pre-built solutions ready for initial business problems you want to solve. The best ROI initiatives are not simply strategic partnerships, and not simply applications of off-the-shelf products: they are a little of both.
You begin with tangible objectives and benefits. “Can we guide a technician through a complex repair?” or “Can we assist a warranty officer with fraud detection?” You aim for a reduction in repair visits, a reduction in average cost per claim, increased Smart Products revenue.
But you have to also consider the more intangible aspects — for example, better prioritization of engineering efforts based on quality feedback, increased brand promise from both repair performance and equipment performance, the new and unpredictable products. And, of course, our favorite example: intelligence-as-a-service helping your employees be happier on the job.