The Road To Artificial Intelligence And Digital Transformation

As I promised last week, this article is more of a "how to" than just the challenges in?Why Your Digital Transformation & AI Initiatives Are Going To Fail. This article will be high-level but will set the foundation for future articles on AI Initiatives and Digital Transformations.

Though I am linking the two,?Digital Transformation?and?Artificial Intelligence?initiatives are related but are different.?Both?are built on a foundation of good-quality data. Still, the roadmaps can differ depending on what your company is trying to achieve.

Definitions

Digital Transformation

Accenture?defines Digital Transformation as "the process by which companies embed technologies across their businesses to drive fundamental change." Where this is true, the transformation concerns?Culture,?People,?Process, and Data?TO embed the technology effectively.

Artificial Intelligence (AI)

In an article?How Businesses Are Using Artificial Intelligence in 2023, Forbes describes the business use of AI to, more or less, "improve and perfect their operations." I agree that this is how businesses are using it today. The advantages of having a 24/7 virtual assistant and analyst are great, but this is just the tip of the AI iceberg. McKinsey has a briefing on Artificial Intelligence (Real-world AI), which speaks to the potential benefits; however, to get the benefits of AI, your?Data?needs to be ready!

Data Foundation

As you can see, the common denominator is?Data, but for very different reasons.

Digital Transformations -?Good quality?Data?is vital since the primary value proposition is to eliminate the productivity loss inherent in people needing to complete, correct, and standardize what is being entered into ERP / Main Applications.

Artificial Intelligence?- AI isn't human intelligence. Humans are intuitive (or gut) thinkers, mainly using data to validate our hypothesis; AI is strictly data-driven. Insufficient or bad quality data means bad quality results. Good quality data is essential, especially as you look to Predict what will happen and Prescribe actions.

To succeed in either?Digital Transformation?or?Artificial Intelligence, you need to work on your data, which is what trips people up. Many Executives assume their data is good since they trust the people entering it, but?good?can be a misleading term since most?data?isn't bad because of maliciousness or carelessness but simply because no one has defined what?good?means.

Actions - How To

Though?Digital Transformations?and?Artificial Intelligence?are related, they both are extensive programs that tend to fail because companies need to see the value soon enough. Speaking to a CDAO recently, her company spent two years on their first Governance project (Customers) and still needed to get value from it. It is a rare company that can wait that long.

In organizations, I have either led teams for such transformations or have been involved in larger groups; success is related to knowing the roadmap but getting value by focusing on priority areas and solving them strategically.

High Level

However, you need training programs to help address skill gaps (such as?Data Literacy?and?Digital Tools). You need to have?Data Governance; you can build a strategic framework and start solving today's problems within that framework.

1. Ensure you have a foundation that supports your strategy.?Sometimes, you need to address some basic needs before you can start. I had one company running on an on-premise SQL database, aging storage, and using an older methodology that didn't allow business users to access data quickly. Since?Data Accessibility?is a fundamental need, it must be addressed first.
2. Make top-level decisions: For example, have an executive team and gain agreement at the top level on something simple, like "Process Owners are accountable for the Data"?(the person accountable for a business process is also responsible for the data). It all depends on the current company culture and organization, but this conversation can take a little bit of time.
3. Decide where to start: With that same team, pick a priority. My recommendation is to balance business needs with readiness (or simplicity). Being able to Forecast Sales for the next 20 years can bring incredible value. Still, it is not as doable as addressing a more straightforward but maybe burning need.
4. Focus and deliver on a small scope: Supply chain management may be a priority for your organization. Build your governance?just within that scope.?Keep it reasonable, focusing on the significant areas, and include all the stakeholders you can, but keep it manageable. Knowing you need?Data Quality, define what it means, check where you are today, and plan improvements.
5. Simplify: Do so if you are working on an?AI initiative?or can get the scope smaller. The idea is to build what you need strategically just for?this?problem. The smaller the scope, the easier it can be.
6. Rinse and repeat:?Using a similar process, you can grow as fast as the business can be active and consume. No matter how small you start, you will quickly get into more complexity (such as more Cross-Functional areas that cover Finance, Operations, Commercial ...., etc.), but start with a laser focus on the doable, smaller scope and grow as you mature.

Conclusion

Though these are extensive programs, an approach that "eats the elephant" in bites while providing reasonable value can be very successful if you do it strategically.

Be careful. You will be tempted to buy a short-cuts. Pointed Vendor Solutions can be valuable, but they rarely address how you do your business, and the limitations of the customizations needed often make it a losing proposition.

In next week's article, we will continue the theme of some?quick-win approaches for AI.