The promise of digital transformation is to evolve into a more agile and enduring organization by taking advantage of advances in AI, cloud computing, data sharing, and increased connectivity. You might break into new markets, advance in your industry or design entirely new business models and services. But the statistics are sobering: Only 28% of major corporations are succeeding in their digital transformation efforts, according to Michael Gale in Forbes.
A while back an IT leader told me that the business was pushing back on a big data project. The project was failing to show value and failing to show ROI. In response, the team decided to attack the problem with machine learning.
Why not sprinkle some AI on top of that data lake?
Digital transformation projects do not fail because there is a shortage of AI. Business transformation is almost always a technology problem. But it is never just a technology problem. It is a technology, people, and process problem. To succeed with digital transformation, you need to evolve all three areas.
Evolving your technology
Digital transformation might have ushered in what we are calling the big data phenomenon, but it is only the continuation of age-old trends. Digital transformation is about the spread of information – and access to it. Today, we are replacing hardware with software and physical items with intelligence. Books, music, shopping, banking, education, insurance, and transportation have already transformed in front of our eyes – they transformed into bits and bytes. Into data. Based on data and technology, new businesses emerged that are crushing existing models: Transportation providers that own no vehicles, accommodation providers that own no properties, retailers without storefronts, and social platforms that write no content.
For established and incumbent organizations, the challenge is to reimagine technology, people, and processes in the new operating system created by digital transformation. As Kevin Kelly discusses in The Inevitable, this operating system has different priorities and urgencies:
Accessing resources is more important than owning or creating resources
Sharing content, resources and intellectual property is more effective than accumulating them
Decentralizing, distributing, and copying is more powerful than stockpiling originals
Connectivity and flow of data is the starting point for innovation and socializing.
Ownership and concentration are no longer the sources of competitive advantage, and applying these priorities shifts the focus:
– You shift from saving time and increasing efficiency for the organization to saving your customers’ and citizens’ time.
– You shift from a product-centric and organization-centric view to a customer-centric view.
– You shift from intuition-driven decisioning to data-driven decisioning.
Cloud and AI play an important part in these shifts, but it is not as simple as just moving data to the cloud and applying some AI to it. Digital transformation is not something you install, like a piece of technology. You need to evolve your people and processes too.
When it comes to getting started with digital transformation, I usually give standard advice, such as start small, pick a pilot project that can show value quickly and communicate the success of the project broadly. Lately, I have been modifying my advice to say: Start with engagement. This puts people at the center of your digital transformation project. Who do you want to engage: employees, partners or customers? Who are your customers, current and new? Do you know what they want?
If you put those you serve at the center of everything, everything will follow.
Once you understand who to serve and engage with digital transformation, your reason – the why – for transforming will be clearer. Consider these examples:
– A financial institution might pursue a more holistic view and understanding of customers using omnichannel customer engagement across product lines.
– A traditional retailer might replace catalogs and brick-and-mortar stores with a digital and virtual online shopping experience for customers.
– A manufacturing company might switch from scheduled maintenance to predictive maintenance.
– A university might move from classroom teaching to massively open online courses.
Another problem we see with digital transformation is a rush to try new technologies without considering the broader organizational strategy. One CEO I talked with recently put it this way: “We are experiencing a tools explosion, and individual departments are trying out the flavor of the day with no alignment to each other or the rest of the business.” It creates a mess that he now must untangle, with help from his IT team. It creates technical debt that he must pay down.
This might sound like a technology problem, but ultimately it is a process problem. If you have processes in place for requesting and supporting new technology, and a digital transformation strategy that is clearly communicated throughout the organization, you can solve this issue.
Likewise, you will find it important to put processes around data science and the use of machine learning models. As analytics adoption increases, data scientists churn out more models of greater complexity, using a variety of tools and programming languages. Getting these models into frontline operations should not require time-consuming and error-prone recoding steps to integrate the models into an operational system. At the same time, models that are built on softer features, such as customer behavior, can decay more quickly than those built on more stable features, such as customer age. And organizations increasingly view algorithms and machine learning models as an additional source of risk, for example, reputational risk from a potentially biased model.
The combined effects increase the urgency to understand, govern, and manage business logic that increasingly reveals itself through data processing. According to IDC, only 35% of organizations indicate that analytical models are fully deployed in production. Of those models that are deployed, 90% take three months to deploy – and 40% take more than seven months. This “last mile” of analytics is crucial to transforming big data into big value. Part of your digital transformation strategies should be putting a process in place for operationalizing models. Today, analytics is about more than just an algorithm. It is about integrating data-driven decisions. Who cares about the clever features of an algorithm if it never sees the light of day and never makes a difference for the organization?
Look for the opportunity
It can be hard to imagine the next idea that will disrupt an industry or transform a business. Most of us are stuck doing business the same ways we always have. Our own imagination is limited, and our creative energy is weighed down by keeping the lights on. But opportunities play out in every industry and field. How can you take advantage of those opportunities? I am asking you to think of digital transformation as a necessity to achieve competitive agility, to declutter and to simplify. I am asking you to think of digital transformation as an opportunity to make a better thing, rather than to do the same thing better.