Big data, with its complimentary AI & ML engines attached, relays unthinkable opportunities for company performance over the next decade. And we so easily envision this future bounded inside current organizational constructs, like inside the existing organization chart. Then some external force – like the COVID pandemic – comes around and flips our understanding upside down and we realize that the organization structure and work routines are not what they are because of some fundamental truth, but simply the result of repeated beliefs that evolved over the years. The frightening part is that they evolved to function perfectly well in an analog world, and we are now in a digital world.
The VCI FORUM involves 31 members, each being an executive directly accountable for digital transformation of their companies. The companies range from global organizations employing hundreds of thousands of people to more national companies with a few thousand people. The Forum meets virtually every 2 weeks and its purpose is to collaborate and learn – We believe 31 brains focusing on solutions beats the lone wolf every time.
Insights from the data architecture and management discussion included a contextual view that we’re probably headed to an ecosystem of communities of knowledge, and a lot more freelance and gig type of work. So, we are not thinking of big data and AI as a construct of the company but as a construct that bounds and benefits the commercial ecosystem. This insight spawns two areas to explore:
- How should one transact with this model in mind?
- Where are the boundaries of the requisite Culture of Innovation?
What makes this stream of consciousness so compelling is that the economics of the situation carries the argument. This ecosystem design approach beats every other model when it comes to growth and cost efficiencies.
Thinking differently about big data also includes displacing the chunky big corporate planning approaches with a much more agile and iterative process. The merit of plowing ahead and just doing things with the quest to reframe how businesses will operate in the future will probably be a wiser strategy than attempting to define everything up-front before anyone gets a budget or a go-ahead. See, there are always good arguments against moving from your current position – the incumbents of the current have the facts, and the visionaries that see tomorrow have only dreams, ideas and concepts. So, if you really want to stop or slow down transformation progress, it’s not that hard.
A good example is how companies do data cleansing and restructuring. Old school thinking would conclude that the company has to first go through a mega project of both cost and time, to define its data and then clean and properly structure it all. However, maybe the smart way is to get going with the basic data concepts and start doing things and learn as we go. If what we learned is that data is not good enough to solve that problem, we fix it and move on. This gets us to that entrepreneurial space where getting stuff done is the beacon, not complying to some grand project schedule. We need to plow ahead and realize that the data will never be perfect. So just get on with it. Obviously, we must be careful that we don’t draw the wrong conclusion from some bad data. But that’s why we have smart people working on things as well. In any case, bad data isn’t bad per se, it’s just anomalies and outliers. And once we’ve identified that, they’re not. If they’re not applicable according to operational requirements, then they become discarded data. We now view data as a new and innovative platform that is changing the way we do things.
Instead of having a large data cleaning effort, it may be more valuable to spend time on making everyone in the company a data champion. Get people to understand what they need to do in terms of the data to produce the results that are expected. A focus on the business question will tell you which data you need to spend your time on and if you intend to transform the company. Let this notion become engrained, like the culture where you do things because you believe in the end goal. Again, we touch on transparency and just having that real time data will also drive accountability.
People traditionally think of data cleansing as a precursor to forecasting because everybody wants to run their business looking forward through the windshield versus in the rear-view mirror. But there is a sort of independent value, which is shining a bright light on what is happening to create accountability. It’s not just looking forward as you drive, but also having accountability in the present.
Another challenge with data is an inability by the folks involved to translate between operational engineering standards that aren’t always explicitly written down and transactional data that is standardized. Operations data flows in their own kind of normative culture, and that’s not always explicit. How this translates into the business rules that govern data systems via ERP, data stores, etc., that operate all separately, is extremely complex. Designing the connections between them is awfully challenging and there’s a role for all involved to help with the translation between these two.
Given all the complexities, there is a prevailing sense in not boiling the ocean. Document the working minimum and plow through whilst keeping your eyes wide open. Look for those quick wins wherever you can. Show the value of cultural change and show the value of some sort of data analytics. Equally so, we should not try to do digital transformation without or around the IT division. We must have those hard conversations to transform IT along with the rest of the organization, rather than trying to tiptoe around them and hope that they don’t notice.
We should appreciate that the IT group has for years fulfilled a defensive role of compliance and providing pragmatic solutions with seriously constrained budgets. To let them loose on an agile unbounded digital initiative is to ask for a quick asphyxiation of the project. Big data projects require innovation and agility and most of all, a mobilization of all functions to want to partake and share the spoils. Part of sustaining momentum, if you don’t have command control role power, is to establish a continuous stream of quick wins based on simple business cases. So, the three horizons planning model comes into play. We have the grand vision and strategy, but we are equally comfortable to add the enablers and the short-term plays into the data architecting narrative. On top of that, we need to remain fluid and probably not spend too much time down the multiple rabbit holes. So, there definitely needs to be agility.
Digital transformation, with data at its heart, is the challenge of the day simply because we are transforming the method of management simultaneously with transforming the business itself. We are, metaphorically speaking, rebuilding the car we are busy driving into a self-driving car while the other team is transforming it into a helicopter. We are digitizing the company whilst at the same time rearchitecting its business model. For that, high frequency collaboration and an agile culture is essential.
(VCI FORUM, Notes and Insights)