DIGITAL TRANSFORMATION THROUGH COVID: NOTES FOR TECHNOLOGY LEADERS

Posted: 18 April, 2020

Link to source

I was originally going to follow up my previous post on transformation with more of a how-to article, but a virus pandemic threw me for a loop. This post is more of a shout out to fellow technology leaders and executives, echoing a few of the themes I’ve heard over the last month in terms of planning through this event. It would be fair to say that many of us have been brought into a lot of discussions around the business model of our organizations and what happens in the short term to the folks we manage, and we’re looking at decisions we didn’t plan for in business continuity plans. The other piece of context, which is no less important than business stabilization, is that we are all in this together – each of us, including our peers, our employees, our suppliers, our leadership, and our fellow Americans, has our own questions about safety for our families and returning to a “new normal.”


We’ll break up these notes into 4 sections: what is the fundamental challenge to transformation posed by the pandemic, what techniques do we have which can help our business adapt, how might we think through short-term decisions that have to be made, and how might we look to the future in decisions for the longer-term. I’m not going to specifically talk about the people management (or crisis management) side of this and have included a really good article from a law firm at the end of this post for your consideration.


COVID is changing customer behaviors


If we look at this as a transformation, changes in how we look at goods and services represent changes in the business model and in how much digital enablement is in the model. From a personal perspective, if I have a choice of a physician who offers tele-medicine check-ups and one who doesn’t, today I would probably choose the former based on trust, satisfaction, safety or other criteria.  That might not have been my choice a year ago. My sense of value has changed, which means that the physician’s practice needs to adapt. 


On a business model canvas, we would represent these as changes in certain parts of the model, from either one state to a new state, or as a large shift between two ends of a continuum. For example, the channelhow I get my service to market – may change from in-person to “over-the-air” or internet-delivered; a brick and mortar retail store may see a 60% change to online orders. My partners and suppliers may change – consultants may be asked to transfer knowledge to internal staff, or I may move internal logistics to contracted delivery for increased online orders. The value proposition or offer itself may change – my service which had been delivered in person for 95% of consumers is now delivered 85% online, but at a different price point or quality level[1]. Each of the business model changes reacts to consumer demand and also impacts top-line revenue. As a business, I want to track consumer demand and maintain a stable revenue through the pandemic. In order to accomplish these goals, I need to do two things: define and test the new business model coming out of the pandemic (maybe this is the “new normal”), and adapt the current business model to keep it relevant.


Techniques to adapt technology to new business models


One of the questions I’ve heard a few times is how should I select technologies to bridge this period of time between my current state and this ambiguous future state. Should I look at my current projects as an incremental path to the future, should I look at my emerging technology (innovation) portfolio as tools to help adapt, or do I have to look further? I won’t get into Darwinian metaphors, but both evolutionary and anti-fragile technology architectures have proposed this approach: organizations should select technologies supporting forward-looking business strategies based on their fitness and evolvability. In simpler terms, we’re not looking to the current portfolio of production systems; we’re looking at those emerging technologies and possibly outside that set even for flexible techniques and digital solutions. Though we may not know what the right set of enterprise systems will be in this future state, we will select complementary, low-cost technologies that let us experiment faster; this would be opposed to selecting best of breed types of systems for well-established requirements.


We already have customer journey mapping, business model canvas and service blueprinting to look at current business processes and understand where social distancing needs to be applied or identifying areas for self-service over in-person teller windows. Many of us have looked at these in the past as experimental ways to get closer to the customer experience, now we can look at them as a primary technique we use – and the good news is that they’re quick and lead to substantial dialog with the lines of business.

From the technology side,[2] we have green shoots as well – things we thought were the domain of prototypes, nails in search of hammers, maybe even technologies which surprised us by how well they worked in the move to working remotely (and might be ready for prime time):


  • Robotic process automation: used in HR scenarios internally and now maybe a way to fill out all those HIPAA forms before visiting the doctor’s office

  • Videoconferencing: useful internally to reduce travel, and now a way to communicate or provide service (training, tele-medicine, etc.)

  • Bots and lightweight artificial intelligence: in healthcare this is serving as a quicker way to provide interactive diagnostic guidance where we only had static web pages before

  • Digital twin and simulation technology: for both understanding complex problems like contact tracing and also to enable robots to enter hazardous environments in lieu of people

  • Social and shared data “exchanges:” an unstructured view of the end-users multiple interactions across organizations (possibly with machine learning or stream analytics)

  • Mobile applications: user interaction for order-taking, now combined with sensors in the environment like location and proximity to others and real-time alerts

  • Public cloud: resilient web sites quickly (especially with automated deployment, Jira feedback collectors, IFTTT and dashboard integrations)


Each reduces direct human contact, replaces paper-based processes, and improves trust in the digital customer experience over the previous version.


Finally, think about evolutionary or adaptive[3] architecture. As we pair up new views on the business with possible technologies, we want to construct systems that at least are robust when faced with similar recurring changes in the environment – e.g., Netflix viewership has been up 35% in two weeks, Zoom up 15M users in the same time period, or even the usage of GrubHub and DoorDash for takeout orders replacing in-restaurant orders. The technology we use should be able to survive radical changes in demand (a known stressor), but can we also use design techniques so that the system has resilience to stressors we cannot accurately plan for?[4] In the longer term we may be able to use this technique to design new systems.


Thinking through short term projects: focus on stabilization


The priorities for the short term are clear – stabilize the cash flow and revenue so that the we can get some time to do longer term planning. Who will tell us what’s going on with revenue?  Customers.  As leaders, not just in technology, but in finance, HR and the other folks around the planning table, we’re trying to gather data as close to the source as possible to build a plan and make decisions. This may mean listening to line-of-business leaders (those folks who own a single service or group of products) as a proxy or talk directly with customers in envisioning labs. Ask, what do you as a customer need from our organization over the next 6 months?


From a project and product perspective, it’s no longer an IT-only plan, with continuing budgets and diverse value. The project list will be a cross-organization plan, and many asks will be made for budget to support whole-organization priorities[5]. What this means for our current list is that we’ll go through it with a fine-toothed comb (some might be more ruthless), to find only the small (not resource intensive and short-duration) projects which are absolutely critical to keeping customers. Some leaders have told me that the whole project list is being revisited, some have adopted freezes on all major projects. There certainly will be some things like critical technology fixes, security controls and even capacity growth that may be unavoidable. Do not, however, push ahead on previously prioritized projects without revalidating with other business leaders that these are still worth doing and have quantitative value in the short term. Upgrading to version 7 of that database or consolidating web platforms are costs you could better incur elsewhere.


In product or major systems (ERP or core business) terms, consider going to a skeleton crew on the current program and getting the team’s head around a design cycle for v.Next. For short-term decisions realize that v.Next may be quite different and there’s a lot we do not know about it yet – the business model has changed. We have our saying “never let a good crisis go to waste” which is sort of a way of saying that when conditions change, we need to go back to re-validate some of the rules we’ve put in place during the calm period.

Some organizations are being more aggressive in the short term by converting open positions (hiring onto the current team structure) into training for the long term to position teams better. Larger cuts in short term operational spend on technology may be deeper as part of an overall organizational strategy. 


Hire for the long term and plan v.Next transformation


Most of this article is based on observations of how we make decisions as technology and business leaders. If we anticipate major changes in the model and our assumptions, we want to look broadly at alternatives and ideas that may come from unlikely sources. One of the reasons for this is that we are creatures of our prior experiences (see Kahneman’s book below) – we are subject matter experts on what came before and it is easier cognitively to go with a path we’ve been down before, assuming past success is valuable in this event. To counter this natural bias, we need to address the unknown unknowns of this specific challenge with a broader pool of advice: getting inputs from our customers in real time, our peer leaders, and quite likely from outside our organization. 


I do not mean to sound too much like a radical, but I am suggesting that we break the long-term planning effort into its own section simply because it deals with potentially a much different customer experience, different data we’ll base decisions on, and potentially represents a deeper change to the business model (even if we are just adding a new model alongside) to what we have today. Our thinking starts with what the concept of this new normal is and evolves backwards into technology programs of change. Makes that Windows 2016 upgrade decision look kinda puny compared to pivoting my whole hotel chain to certify employees are safe and rooms are cleaned in a new method and check in is all phone-based and room service ordering with a tablet in every room and I get my own Peloton in my room (you knew I’d throw that in here!) is the new quality of service. If you do not have an emerging technologies or prototyping or innovation group, start one. 


If you choose to break up the short term and long term, the long-term projects are all greenfield – they need new titles, new value definitions, new envisioning. The good news is that the new project list has a number of useful techniques we already know, and the better news is that these projects are also low cost experiments initially to define v.Next.  v.Next will be defined by a digital first approach – simply looking at current business processes or journeys and identifying how to remove the facilities-based actions and tasks where we want to apply social distancing now.[6] V.Next has its own customer experience, which is net new. This has to be compelling to the customer as a standalone service or offering; do not assume that since the customer bought v1, that they’ll move to v.Next (unless you make it highly incentivized to do so). The value proposition of v.Next is at least as good as the v1; that does not mean we have to do all the functions of v1, however. May Post-It notes may be gobbled up in this process – we are using the creative side of our brain, with a time deadline of a few months to emerge onto this new normal. We will also need this creativity to come from the business side of the house, so if you are hiring, consider bringing in an external viewpoint who represents the specific customers and industry to drive ideation. [7]

This effort will take all of your team’s focus and ingenuity so the reduction in short-term tasks should clear their plate of anything customer-misaligned. Do not take on legacy debt for v.Next – ripping and replacing an ERP mainframe system is going to be $M and take years to yield the same pre-pandemic experience: duct-tape, PowerShell (or a brokered/messaging architecture), and a new web application prototype may be $K.


I appreciate all the conversations I’ve had over the last few weeks here, so thank you for all of your inputs. Hopefully this approach of breaking the thinking down into a couple more manageable problems is helpful.


Since we are all in this together, if you or someone you know wants to brainstorm about some of these technology planning issues, www.princetondigitaladvisors.com is open to help with confidential advice and referrals. We will make it through this. Please stay safe and keep your family close.


For further reading:

[1] A good example here are some of the changes in higher education: most universities have moved quickly to online education, which helps in social distancing terms but not all universities are good at this new service offering. The New York Times reports here that quality varies widely, and students have protested that the online experience is not the same quality and should be offered at a lower cost. This shows that some of the changes we may be making in the short term may not lead to continued success coming out of the crisis. We know that customers will have choices and they will adapt their behavior to what they learn through these days ahead.

[2] Digital technologies help during this type of event more than other solutions, but we have to recognize that other changes in the business model may have other accelerators. Also note that there are a number of accelerators that help as well – 5G, microservices, etc. might be included as indirect technologies or ones which require work across organizations to implement

[3] My friend, Barry, describes this as anti-fragile in the references below

[4] Future article, we’re not going all Blade Runner yet.

[5] Business units may want to use some of the cash flow for discounts or incentives https://cars.usnews.com/cars-trucks/best-car-deals#bottomDesc or cleaning services for dorms https://www.nytimes.com/2020/04/15/us/coronavirus-colleges-universities-admissions.html or other things not in the annual spend plan

[6] Though this sounds like a traditional “requirement,” it is certainly architecturally significant and more similar to a “stressor” in antif-ragile terms than a functional one.

[7] Each organization approaches hiring differently; some see an event like this as an opportunity to bring in talent that they would not have been able to reach with the right role previously, and others may look at training up their own internal talent first. My recommendation – which I use in my own organization – is that a new model may mean I have to look for that quality of out of the box thinking and for internal candidates for this new role, look at how big the gap is to bring in new experiences.