Artificial Intelligence: Implementing a Successful Strategy in Your Company
[Note: This article was originally published in Portuguese on CI&T on December 12th, 2019.]
For years, Artificial Intelligence (AI) has been a hot topic in the market and has shown significant real growth in companies. To give you an idea, a study from the MIT Sloan Management Review indicated that 58% of organizations expected AI to bring significant changes to their business models by 2023. A Forbes article from 2019 also noted that 73% of U.S. executives aimed to significantly increase investment in the technology. In a similar vein, the "Trends to Transform Your Company in 2020" survey conducted by CI&T in December revealed that 49% of leadership in large Brazilian companies claimed AI was indispensable for business growth.
Despite these impressive numbers, we observe that a large portion of these companies has not made progress beyond experimental uses of the technology. This is because, when thinking about Artificial Intelligence, there is often a vision associated with something spectacular, almost a realization of science fiction predictions. If it's not about creating something grandiose, like an intelligent robot, it seems to lose interest and purpose.
This romanticized idea ends up distancing the possibility of a more pragmatic use, one that truly brings value to companies and their customers. The main—and most immediate—value at this stage of technology maturity in the market comes from its ability to facilitate customer decision-making. In other words, it's about reading the tastes, desires, and needs of customers with increasing accuracy to enable increasingly personalized offers.
With so much information and so many possible options, the problem to solve is how to reduce this volume, how to reduce noise, and deliver relevance. Consumers no longer have the time — or patience — to waste. Someone searching for sneakers, for example, doesn't want to browse through 30 pages of product listings unrelated to their taste. They prefer entering an e-commerce platform that offers a page with options aligned to their preferences.
In this context, the differentiator is the quality of the filter and the ability to unravel patterns and predict desires. And, here, AI comes into play. Today, this type of technology use translates into better experiences, improved customer relationships, increased loyalty, and consequently, better results for the company. But, as mentioned before, this use doesn't cause much excitement, and companies still struggle to understand Artificial Intelligence as just another tool among many, a facilitator to generate impact, and not the impact itself.
It's time to reverse the logic.
I often compare this moment in technology to the early days of the internet, which, when it emerged, sparked fascination — and even some apocalyptic predictions. Today, the internet is a natural part of our daily lives, and we only notice its absence when it's gone. The same will happen with AI soon.
The focus needs to shift away from the technology itself and start thinking of it as an enhancer of the value offered to the customer. I say this because much of the discussion about Artificial Intelligence within companies still starts with the question: "What data do we have, and what technology do we need to work with them?" It's time to reverse this and start exploring your customers' needs. What are your customers' questions that, when answered, will generate the most value for them and the company? Only after having this clarity should we move on to finding the right data and applying AI. The starting point defines whether the strategy will yield successful results or not.
Let's think about an airline aiming to improve seat offerings. If the main customer questions are about the best time to buy tickets for a sports championship or another event, using AI with the company's internal data alone will not be sufficient to provide the right answer. At most you'd be able to tell the best day to buy seat 31B - and your customer is not interested on that. It will be necessary to gather external data about these events, which, combined with internal information about flights and personal preferences of customers, will provide the AI tool with the ammunition to leverage its full potential in delivering real value to the passenger.
This perspective will determine whether companies can move beyond the experimental stage with technology and actually use it in practice to generate real positive impacts for customers and move the business forward.
Guidelines for Implementing a Successful Strategy
To build an effective AI strategy that truly drives results for companies, some key points are fundamental:
Raise Awareness Among Your People:
The first step should be education. It is essential to familiarize everyone in the company with the possibilities of the technology. Business teams, leadership, product management professionals, and experience design teams need to understand the capabilities of Artificial Intelligence to identify and seize opportunities. On the other hand, development professionals need to go beyond technology and understand the real business problems.
Break Down Silos:
For knowledge to flow, there is a need to break down silos and unite technology and business teams so that, together, they can discuss the end-to-end journey. In addition to creating and maintaining focus and objectives alignment, this knowledge exchange, the multidisciplinarity of teams, helps accelerate operations, eliminating communication noise and unnecessary processes.
Value of the Correct Answer vs. Cost of the Wrong Answer:
Before adopting AI in the company's strategies, always consider both sides of the coin: if the company uses technology to get the right answer to a specific business question, will the result be greater than if conventional methods were used? In this equation, on one side, you must consider the costs of implementing the technology, and on the other, the learnings that can generate future gains.
Start Small:
To begin, identify a customer pain point that, if solved, can generate significant value. However, if the initiative goes wrong, it will not negatively impact their experience.
An interesting example is the application of AI tools in the customer journey of a clinical analysis laboratory. After identifying that a significant friction in the journey was the scheduling process, it was found that the problem was in registering the tests to be performed. Struggling to understand the doctor's handwriting, patients gave up online registration and turned to the call center.
To create a solution, it was noticed that most tests are correlated. For example, suppose someone is getting a bone density test; they may also need blood tests to measure calcium levels. The solution was to develop a recommendation system that, with the use of Artificial Intelligence, could predict possible tests linked to each other. Thus, the patient would only need to enter one of the tests from the prescription and identify the others among the options offered by the registration tool.
In the first month alone, the test group—receiving recommendations—completed 25% more registrations by themselves than those following the conventional online scheduling process. The result? With a simple use of technology, the customer saved time and had a better experience, and the laboratory gained agility, savings in call center structure and personnel, and, most importantly, user satisfaction.
Build Scalability:
After successfully applying AI in small initiatives, it's time to take a bit more risk and gain scale in other points of the journey and even in other business fronts that generate even more value.
Thinking about a more significant impact, there are uses related to the interpretation of the physical world, with image analysis. In industries, for example, the implementation of Artificial Intelligence in the automation of production processes brings significant results. If, in a production line, machines — or robots — have the ability to recognize and deal with new situations, they can make adjustments automatically without the need to be reprogrammed for each route change. Moreover, you can achieve much better results if the robots are learning from all the new information.
To move beyond the experimental stage and truly reap impactful results, now is the time to bring the consumer to the forefront of strategies and prepare your people to harness the potential of technology. Together, the collective intelligence of your company and AI have the great power to generate high value for your customer, your business, and captivate the market.