The digital innovation process has made organizations more agile, but the shift could present its own set of IT challenges for modern companies. Genpact’s Gianni Giacomelli explains why.
Gianni Giacomelli: What we see people most benefiting from right now is significant cost reduction in a number of business processes. Banks, for example, have been able to optimize a lot of their IT processes through automation.
There are companies that are embracing the business model transformation power of digital. They have used digital to create completely different business models. At that point it’s not just about cost anymore, it’s about giving customers access to things that were not possible before. Tokio Marine, for example, has a new business model that allows you to buy insurance by the hour.
How has your innovation strategy changed over the years?
Giacomelli: The way we do innovation now is different than what we did 15 years ago. We had a much heavier strategic planning exercise with a lot of marketing research, and that would take months. Today, the process is much more agile. The innovation process is also a lot more human-centered, a lot more iterative and the technical realization of it is generally a lot easier. We try to test fast, test often and fail quickly.
When it comes to innovation, you obviously need to have an understanding of the problem that you are trying to solve and the outcome that you are trying to achieve. You have to look at it through the eyes of your intended beneficiaries, the people, processes, business leaders and customers, which will help catalyze the process of innovation.
What are the biggest roadblocks to innovation?
Giacomelli: First, the process for innovation is often hindered by the inability to take risks. We have seen that humans change less fast compared to the technologies. Very often the problem is a cultural resistance to not just change — it is normal that people fear change — but also the resistance to take risks or big bets. The middle and lower management very often fear taking any sort of risks. That’s a big hurdle.
Secondly, companies have invested in enterprise technology for many years. Much of that investment is now legacy and rigid. At the same time, business processes have gotten more complex, and while business process reengineering and lean management have become more prevalent, most processes in most companies are suboptimal in today’s reality.
Third, any intervention, be it on technology or processes, should not be seen in isolation. Trying to optimize only one step of an enterprise process typically leads to over-engineered and possibly underperforming solutions. Instead, it is important to look at a process end to end. For example, a customer’s experience when making an order, of which an individual step, say billing, is a part. The billing then can be optimized by looking at its relation with the steps before (for instance, the choices in the web cart) and after (e.g. the ability to cross sell, or the retrieval of bills online).
How important is design thinking in the digital innovation process?
Giacomelli: Design thinking is incredibly important because it allows you to do prototypes — what we called low-fidelity prototypes — that you can put in front of the people to elicit their response. There has been empirical evidence that companies that have really managed to use digital well excel with both change management and human-centered design.
The design thinking process is important when companies are in the process of implementing changes to the way they work to ensure that those changes feel positive and aspirational, as opposed to feeling someone is forcing new things. You use design thinking because you need to have design-centered choices.
The design thinking process is also pretty good at getting people from cross functional groups like the business, sales, marketing and IT to work together.
What are some of the biggest innovation trends that we will be seeing the rest of this year?
Giacomelli: This is going to be the year of artificial intelligence. There are parts of AI that are completely ready for prime time. Computational linguistics, which is basically natural language processing, is a part of AI that has gone from being 70% accurate to 90% accurate in 24 months or so.
We have invested significantly in this space and we think NLP and computational linguistics have the ability to automate and process semi-structured information, which is completely revolutionary. My bets are that in December we will have a bunch of jaw dropping implementations of this.