With the pandemic, organizations had to undergo 2 years transformation in 2 months. There are multiple components to a successful digital transformation. Jace An, in his book ‘’77 building blocks to digital transformation’’, highlights all of them.
Artificial Intelligence is one of these components that make a big difference and that will drive a lot of the future.
How can an organization be successful at AI?
Have a data-driven mindset: it is about decision-making, measuring success for technology, and for business. Data should be drivers. Digital gives the means to have visibility on everything through data. Not leveraging them, is a huge mistake. However, in order to use the data correctly, everyone and everything needs to be driven by them.
Look holistically at the organization: do not operate in silos. All departments should come together, and you should build bridges between them. It is no longer viable to have Technology work independently from Business or Marketing work independently from Product for example. Success is only possible with a holistic approach.
Have a trusted advisor: advisors are here to provide an outside perspective and be able to call out the misfunctions that we tend to not see any longer when involved deep into the business. An advisor is also someone who has the expertise and who has done it over and over so they can prevent you from a lot of mistakes.
Have a governance: define rules, processes, communication channels. Smoothen and fasten the process by engaging everyone. Having a governance also means fostering a culture of testing, questioning, measuring, agreeing, and moving forward.
At #MicrosoftAISummit, Wesley Rhodes (VP R&D and Technology Transformation at The Kroger Inc.) says that nowadays ‘’you have to be a tech company or you will not survive…AI is the way you have to go’’.
How to select the right AI solutions?
What are the goals? For AI the goals should be to automate or augment the way we work. Doing it now rather than later could help the company be seen as a thought-leader in the market and it can just be for brand positioning, marketing, and competitive advantage. It could be because it is improving the customer experience, keeping a market share, and be positioned along with competition or it can help cut costs and improve productivity. Once we understand the goals and why we are using new technologies, we will be able to make an evaluation aligned with the expectations. The business value being of a different type, the evaluation will also be different.
Who should be involved in the selection? It is always good to involve IT and business in all evaluations. As mentioned, we need to work holistically. From IT, we want the CIO/CTO but also some SME in the company that will be either developing or integrating the technology or who can advise on the challenges or things to be considered. From the business side, marketing, sales, and operations need to be involved as they need to be able to leverage the technology. A lot of technologies could have an impact on customer experience and need to be communicated clearly to customers. Product management must be part of the evaluation because they will write the requirements and be the ones defining how the new technology will mold the company's offering. Product Management will also supervise the whole implementation and be the link between business and IT.
What success criteria should be used to support a final decision? We need to go back to the goal. The metrics will wary depending on the goal:
Brand: will be the NPS score, number of publications and posts
Customer experience: number of engagements, conversion, users, average order value, time spent on the new application.
Improve productivity or cut costs: costs saving, speed of a process (a process might be faster going from 15 min to 5 minutes for example or it needed 5 people and now it needs 3 which then can be translated into dollars)
Is it useful to seek external support for additional insight? It very much depends on the savviness of the internal team. External support is a must-have if the team is not familiar at all with the technology. However, having an understanding from an outsider always gives a different perspective and can help in decision making. It helps make sure that all elements got considered or evaluated. The external support has also certainly done it before so they can advise on the pitfalls and prevent some errors, so the implementation is smoother and more efficient.
When should a technology implementation be reconsidered or postponed?
If the benefits are not clearly visible. Even if it is done for marketing purposes, for example, but that customers or the market are absolutely not reactive and have no interest, it is better to consider it for later. Doing a small survey with users is usually a good way to understand also the market interest.
If the current infrastructure cannot support it. Old systems, data challenges, the organization is already going through changes so cannot take on more, no internal people to be able to support it and take over after. If internally the organization is not ready to support new technologies, it is better to postpone it.
If the new technology costs a lot of money. It is better to assess if it is worth the investment. In the end, embracing a new technology is a test. If the test costs a lot without possible long-term or short-term benefits, we can wonder if the investment is necessary. The company needs to be able to sustain this investment and it should not prevent the company from making another investment that would be more profitable for now.
What's the first step to take when deciding to move forward with a new technology? It is necessary to put together a plan highlighting what do we want to achieve, what is the business case, what are the needs, how we are going to do it, what is the budget, what is the scope, what is the project plan, what is the roadmap. It is necessary to start small and put together multiple phases to the project so that each step teaches something and justifies the need for the next step.
It is important to evaluate the introduction of a new technology at the right time. Too often, organizations introduce technologies too late and then they have to play catch up, they lose market share and do not leverage the benefits for their customers. Organizations should always watch the market and have a budget for innovation. Nobody wants to lose the train. Some organizations are afraid, want to see how it will behave, wait, and then wait too long but will have to be reactive at some point because it will quickly start hurting their business. Introducing new technologies will have, in the short term, an impact on the bottom line and might cost 1 cent or 2 to the EPS. Managing shareholders and expectations is never easy but sometimes it is better to take a small hit now rather than a bigger later because no matter what, it will have to happen.