Looking back at 2020, most forecasts were simply discarded. The epidemic and government responses ripped through most market sectors. Before the epidemic smashed into the world economy, there were rosy forecasts about the coming digital transformation. The details of what this meant varied by market, industry, and nation, but the sentiment was strong: get on board or be left behind. 

Three Kinds of Digital Transformation

How has that urgent advice played out? It varies. In this series, we’ll examine three kinds of digital transformation that created “two cities,” with some organizations living in digitally-driven success and others living in towns where digital technology has failed or even been abandoned.  

Therefore, the three digital divides are digital enterprise, cloud migration, and big data adoption. 

Digital Enterprise in Organizations

The first of these, a truly digital enterprise, has seen stark differences. Successful organizations have some common attributes. However, unsuccessful groups tend to have the opposite attributes. 

The most important characteristic seems to be how digital an organization was before the epidemic and recession. Those who accelerated their digital-driven results were corporations and government agencies who were either “born digital” or well along in enterprise-wide transformation. 

Enterprise-level transformation requires data use horizontally.

The root of this seems to be the vertical stovepipes within large businesses and agencies. Mature organizations have a bewildering collection of databases, spreadsheets, and data aggregation methods. 

Each box on the org chart owns a data set and jealously guards them. In some cases, the motivations are political, but in most cases, they simply want to protect their ability to do their job and deliver what they perceive to be their mission. 

Enterprise-level transformation of the enterprise requires data use horizontally, across stove-piped verticals. That takes painful, time-consuming, expensive work. Data sets are not collected in synchronized methods; they lack consistent definitions and often lack attribution or provenance.

Even within a stovepipe, data might be moved from one application to another by manual methods (spreadsheets, sneaker-net), so no “data set” actually exists in that unit. 

Should data be shared across all stovepipes in an enterprise?

The more extensive and more geographically diverse an organization is, the more difficult it is to attack these issues. There is no data lake. There are data ponds, data puddles, data swamps, data canteens.

Not all the data in an enterprise needs to be shared across stovepipes. In retail supply chains, item-level RFID is a good example. It is one narrow kind of data. It provides the means to track each item from the source to the point of sale. Most RFID benefits were targeted in retail outlets. But some integration was needed across purchasing, suppliers, shipping, distribution, marketing, and retail locations.

Firms that had achieved critical mass with this kind of integration could pivot their business models, understand their frozen supply chains, and generally outperform laggards. 

Some markets, like air travel and energy production, have been hit hard by low demand. It’s easy to argue they “need” digital transformation to survive, but it’s impossible to think about digital transformation while large swaths of the company are being laid off. 

The same patterns emerge in government agencies.

Among government agencies, the same pattern emerges. We see some agencies who had made steady progress before the epidemic hit, and they have been able to accelerate their visibility and control despite challenges in their human institutions. But those who didn’t have a foundation to build on have made little progress, and some have regressed. There is a risk from post-election changes in leadership in the U.S. federal government, which often stalls change efforts for a year or more.

Sorting out difficult organizational issues across the organization is challenging work. Consequently, doing this work via web meetings in a giant organization is simply impossible. That’s why the digital elite moves farther ahead of laggards and why advice varies depending on which city you live in.

Digital Enterprise in 2021

Entering 2021, if your organization is among the digital elite, how can you extend and increase your lead?  Here are two suggestions for consideration:

  1. Move beyond visualization and business intelligence. Those are a given for digital enterprises. Future leadership will come from the ability to make accurate predictions and prescriptions.  Soon everyone will be doing these basic digital functions. They won’t keep you in a leadership position. Of course, this is what Lone Star does, and we’d love to help.
  2. Use digital leadership to help your customers who can benefit from your progress. A great example of this is ABB’s Wellhead Manager for oil production platforms. This digital offering leverages existing data streams in oil production and requires zero new computing infrastructure. Most firms can improve their cash flow from oil and gas wells before sending the first payment to ABB.
What to do for your organization

If you fear your enterprise is lagging as you enter 2021, what should you do? Here are two suggestions, which should apply to most organizations, whether lagging or not:

  1. Be honest about your size. Small and medium enterprises can connect more types of data and get stakeholders aligned in ways big outfits can’t hope. The scope of your ambitions should take this into account. The larger the firm, the narrower your focus can be, at first.
  2. Start with the data you have under control. If possible, don’t try to bridge incomplete data sets or uncooperative units.   This achievable span varies with the organization and with digital maturity. But it is best to avoid an effort that starts with a new data collection or infrastructure.

Finally, there is one suggestion, which applies to everyone considering their next step in digital transformation: 

Focus on what kind of results you hope to achieve.

Are you hoping to drive your top line, gaining more customers and revenue? Do you want to drive the bottom line, improving profits, productivity, and output? These usually involve different parts of the enterprise, different workstreams, and different kinds of digital transformation. They also involve different levels of acceptable error. Being wrong 95% of the time in digital ad targeting is world-class performance. Even more so, few firms ever achieve a 5% click rate. Of course, being wrong 95% is catastrophic in most other areas. The methods we use for ad targeting, or product recommendation are probably unsuitable for anything with safety considerations.