Big Digital Transformations Come In Small Packages

Part 1 of our digital transformation series: keeping it simple is the best way to begin

What’s the best way to begin any digital transformation journey? Find a project that is achievable within one month, with the resources you have today, and that is of immediate value to the business.

That’s it. That’s your Plan A. Compare that with most digital transformation projects, which can take a year or more, require significant new headcount and resources, and whose value to the organization is an unproven projection. In our experience, keeping it simple and starting small is the best way to begin any digital transformation journey, not least because several teams can start small in parallel.

Now, we do acknowledge that this approach may not be viable for every organization—or may not be ambitious enough to meet your objectives. By culture or by policy, most enterprises require some level of up-front planning and commitment before allocating resources, which this approach goes against. However, for the vast majority of organizations, there’s more to be gained from starting small than you might think. Let’s take a look at why that is.

Challenge #1: Prove your ability to execute

By ability to execute, we mean that every person involved in a project needs to have the necessary know-how to make meaningful contributions, and the team needs to be able to collectively make informed decisions in a timely manner. And chances are, both of these attributes are in short supply. But despair not: You increase your ability to execute simply by doing the actual work—there’s no way to whiteboard your way around it. Better to do what you can today and take the risk of having to solve the same problem twice, than to overthink it once and do nothing in the interim. Worst case scenario: you’ll learn a lot along the way. Always. Be. Iterating.

For example, if your role affords you some budgetary discretion, you could grab your company credit card and set up a cloud instance, within minutes while you wait for your finance department to sort out centralized invoicing. You can always attach your project to another billing account later. Or this: while Central IT considers how best to synchronize your company’s central user directory to allow for seamless single sign-on, you could manually create a handful of user accounts for your team.

Instead of asking for permission, look for precedent in other parts of your organization. Your Marketing and Sales departments are a good place to start, because they have had to quickly embrace cloud-based services to maintain their competitive advantage. Which is to say: think of your first project like your department would think of setting up an email newsletter or an appointment reservation service or a shared documents folder on the Web, rather than as competing with your data center.

Challenge #2: Mind the bystanders and the skeptics.

It’s not that they lack imagination or ambition: it’s the fact that they are right to assume your efforts will fail. 

Until you produce tangible results that the business can value, looking from the outside in, your ambitions are a distraction at best, or worse, are undermining other departments’ efforts to drive their own digital strategies. Every week that goes by in which you don’t deliver the outcome that you promised, the skeptics are proved right and you are shown to be wrong.

Rather than make a lot of noise and concern yourself disproportionately with how to share your progress with various stakeholder groups, it’s easier to fly under the proverbial radar and focus on delivering results quickly. It’s much easier to ask for forgiveness than for permission when you can point to a successful outcome—especially if its scope was small enough to not call into question anyone’s role at the company.

Be prepared to have your solution’s greatest weakness compared with the incumbent solution’s greatest strength. Even after you have successfully landed your solution and can confidently point to its benefits, it likely won’t be superior to the old solution in every way. If you are truly starting small and gradually building your competencies, then v1.0 will need to be followed by v1.1 and so on. “But it doesn’t do X!” is an objection you will hear often. The comparison might seem myopic or unfair (don’t forget: the big picture is only obvious to you), but the good news is that it’s easy to anticipate and to mentally prepare for. You might even want to preempt this challenge to your solution by being upfront about it: “Here’s this new solution’s greatest weakness. And here are all the benefits that outweigh it.” (Or maybe the positives don’t outweigh the negatives, in which case you need to get back to work and do better.)

Challenge #3: Work with what and whom you have

If your first instinct is to ask the business to allocate a quantity of developers before starting, you’re not thinking small enough. Instead, identify those individuals who are already interested in the technologies underpinning your project. The prospect of being allowed to experiment and potentially becoming the go-to subject matter expert within your company can be a great motivator. They will make the time necessary in their busy schedules. And don’t forget, we’re only talking about a one-month period (for now.)

Contrary to popular belief, Google does not categorically allocate a 20% Project to every employee with no questions asked. Googlers first invest their own time into prototyping and documenting their ideas before asking for 20% time to pursue them in earnest.

Similarly, your investment can grow proportionately with the scope that you’ve successfully delivered. That’s why it’s so important to start with something you can deliver in one month. Once you’ve successfully accomplished that, deliver the next bigger thing in two months. And the next in four months, and so on.

As you invariably cross that threshold where you do need to ask the business for additional resources, the value you have already achieved serves as a “security deposit.” Your demonstrable ability to execute becomes your credit rating.

So where is a good place to start?

Pick a software solution that does not handle any sensitive data and won’t hurt the business if it goes down for an hour. Here are five examples of quick-hit projects that we’ve observed at some of our customers and how you might measure their value. They may or may not map to your business, but may give you some ideas about what we mean when we talk about taking small steps toward digital transformation:

1. Run one of your intranet websites (e.g. your cafeteria’s lunch menu) in a serverless container. Measure how many hours of toil you saved the company per month by removing that one instance of an operating system. Maybe you saved some OS licensing costs and some compute resources too? 
2. Replicate a tape-based backup strategy for one application and play around with the various storage classes. Measure how much quicker the retrieval time is and how much money you save per month on storage.
3. Offer secure developer environments to software engineers at the push of a button, using infrastructure as code. Measure the improved lead time, overall developer satisfaction, and money saved by consuming resources only during working hours.
4. Get rid of the need for a VPN connection for an internal application and authenticate your employees using context-aware access. Measure employee satisfaction and loading time/latency when they can access e.g. their department dashboards from any device at home (while simultaneously keeping your company even safer).
5. Set up a simple, serverless, continuous integration and deployment pipeline and see how many minutes you were able to shave off the build time and multiply that by the number of builds per month. Maybe you also managed to save money, since it need not run 24/7 when your developers are asleep?

Starting small doesn’t mean doing something trivial. It means validating your ability to execute, arming you with evidence of success to fend off skeptics, and netting you additional resources to invest in bigger projects moving forward.

And sometimes it can result in something truly innovative: just recently, a small team at a European insurance company used custom AI models and visual AI detection to rapidly prove within six weeks that a machine learning model for document analysis could double the speed of claims management, without compromising accuracy. With this early evidence of success in hand, they were then able to ask for the resources to roll it out in earnest. And if that’s not a great first step towards a larger digital transformation, we don’t know what is.

Save the date for Google Cloud Next ’21. Learn from our top developers on how to begin your digital transformation journey. Register here.