
Digitalisation is at the peak of the hype curve, but most companies struggle to get going.
Some companies hire an army of data scientists hoping for magic to happen.
Other companies hire consultants hoping for a strategy that sets them apart from the competition.
Everyone is hoping to become the next Airbnb, Uber, Netflix, Amazon, etc, but obviously this will only be for the happy few.
Most companies, however, will benefit from improving their existing business, e.g. through lower cost, higher output, higher efficiency, better quality and improved safety.
This implies that the focus of Digitalisation should be on existing employees and the work processes in companies.
You can hire new people like data scientists, but they often lack the domain knowledge (i.e. in-depth knowledge of the work processes). You can hire strategy consultants, but they often are too far removed from the shop floor.
You should focus on your own people and develop them to transform the company for you.
In my experience, there are eight practices that give you a better chance of succeeding in starting a Digital Transformation and they focus on employees. I’m not a business consultant (and not planning to become one) and I’m not sure if these will work in your situation. There are most likely more practices that help. Anyway, I do think these eight improve your chances significantly.
The first practice is taken from LEAN and is called:
- Pull, not push
When starting something new, you’ll be faced with resistance. Yet there are always people around eager to try new things. These people you need to identify and you need to give them all your attention and energy.
Start initiatives in the areas they work in, be it a location or a process, doesn’t matter. Give them your time (by dropping other things when they call on you for support or advice). Make their lives easy, by breaking through barriers and ask them to escalate bottlenecks. Offer them training opportunities and let them present their work to your senior managers and executives.

Sometimes the bigger opportunities are in areas where there is most resistance. Still, don’t burn up your energy there, work around the resistance first. Only spend energy on areas where people ask for support, hence the name pull, not push. People should be pulling for help rather than you pushing it on to them. In the short run, pushing may work, but in my experience, the results won’t sustain.
So, pull, not push.
- Sponsorship
You need people high up in the organisation to sponsor digital, otherwise, people won’t start moving. This doesn’t mean the initiative should start from the top (although that would help), you can also influence top leaders by showing early results of bottoms up initiatives.
Sponsorship can be as simple as mentioning Digital in communication to staff. Easy to do, great effect. Other things that help are releasing budget for initiatives, freeing up people’s time, facilitating intra company events by bringing thought leaders together and organising visits to other companies (often other industries).
You can start without a sponsor, but the effort will not be sustainable.
So, find yourself a sponsor.
- Small core
Less is more. If you want to move fast, better start with a small team. If you have worked in teams of more than 10 people, you know what I’m talking about. The more people on a team, the more overhead in terms of communication and often that doesn’t help to speed things up. Price’s square root law is also applicable to Digital: 50% of output is generated by the square root of people involved. If you have say 100 people in your organisation working on Digital, roughly 10 of them will generate 50% of the output.
To get started in Digital, you need a small full-time core team (~5 people) that paves the road for the rest. Select only the best people for this group, but also pay attention to the overall composition (i.e. complementary skills).

You need to look for people with domain knowledge, coding experience (no need for full stack software developers though) and communication skills. These people are hard to find, but don’t make the mistake to settle for less, you will regret it as things won’t move.
So, get a small core team in place.
- Engineers turned data scientist
Data science is the most sexy job of the 20th century and this attracts loads of young people to computer science studies. You definitely need data scientists in your digitalisation effort, but don’t underestimate the need for domain knowledge. No matter how smart you are, you will not be able to catch up with years or sometimes decades of domain knowledge. Most data scientists do pick up a minimum level of domain knowledge but in some areas this is very difficult, for example in engineering.
I’ve always worked with engineers and most engineers did some form of coding in university. When they start working the need for these skills reduces as they start using commercial software for their work. That’s a shame as they bring the domain knowledge and have the skills to code (but stuck away somewhere). I find that when you get engineers to code again, they become super productive. They won’t have the in-depth coding skills of data scientists, but if you explain to them that (Scikit learn) machine learning mainly involves linear regression, they quickly will get going as they understand what you are talking about. The other way around, I’ve seen smart data scientists completely miss the physics behind the models they were building ending up with completely useless outcomes.
So, for a transformation, get help from a few data scientists, but better focus on getting your engineers to pick up coding again.
- Online + offline training
No shortage of online training these days. You can learn anything digital in Massive Open Online Courses (MOOC). You can even earn nanodegrees in any topic you like. This is great of course. However, training is only useful if you apply it in practice.
To make it practical, you can complement online training with offline (classroom) training and let the participants work on real-life problems in your company. This way you get a win-win, where people develop themselves and solve problems that exist in companies.
The main benefit of classroom training is not what you think it is. In companies, people are always pulled towards day-to-day work (urgent, not important) and this makes it difficult to devote time to learning. If you let people only do online training, only the most dedicated people will complete the training. By ‘forcing’ them into a classroom, they will more likely block their agenda’s and focus on learning. By having people in a room together they will also get a broader perspective of how the material could help solve business problems. This isn’t easy in (individual) online training.
As part of the classroom training, you can ask people to form groups to work together on the business problems. By working together, they learn more, they are more likely to stick to it and they create value for the company.
So, don’t keep it to online training only, complement it with offline training.

- Invest in self-service
Self-service means you let people from the business develop digital solutions themselves, rather than IT. This can be as simple as a dashboard, but can also be a simple application or automated workflow. Self-service doesn’t mean people have to turn into full blown software developers. The key is to let people create a minimum viable product that they can show to IT, so that they can create a solution that is sustainable without spending too much time in the design phase. It’s worth noting that a lot of digital solutions never make it to production as the development process on its own solved the root cause of the problem.
Often the most difficult part of digitalisation is getting access to data. Ensure you give people access to the core data sources in your company. Also give them the tools to work with the data and platforms to publish their products. Give people space to develop their own solutions, but support them when they’re stuck.
So, invest in self-service.

- Never wait for next/better/shinier tool/initiative/promise, never
Digital moves so fast that when you’re halfway implementing something, the next better version is out to replace it. It is tempting to wait for the next, better, shinier version as it may give more value. This is a problem many companies face. The problem with waiting is that you never get started. And if you don’t start, you don’t solve problems.
Many data scientists/developers experience this as well, with new libraries and frameworks coming out at record speed. Learning a new library, framework or even programming language requires a serious investment of time. This time you could also spend on developing a solution. Quite some data scientists burn-out as they cannot keep up with developments. For a business however, it doesn’t matter which framework was used, as long as it works. You’ll add more value by solving problems, even if the solution uses ‘old’ technology. In digitalisation time ‘old’ is relative anyway.
So, never wait, never.

- Bend the rules in favour of speed
The last practice is called ‘bend the rules in favour of speed’. I don’t mean breaking the rules here, although sometimes that may be required.
Bending the rules means that you accept things not to be entirely as per the book, but you know that leaving it like this for a limited time won’t do much harm. This requires some bravery of your staff, so don’t come down on them if you discover the book wasn’t followed.
In large companies, inevitably there is red tape and red tape slows you down. Red tape was put in place because in the past something went wrong and controls were put in place to avoid this in the future. Nothing wrong with that, however, red tape often has undesirable side effects.
When you develop digital solutions, you most definitely will encounter red tape. For example, you want to test something with production data, but you need to adhere to all sorts of policies and need to get approvals from all sorts of departments. Boom, a couple of weeks delay. Or, you could ask a friend in IT to make an ad hoc connection and voila, test done, but in 5 minutes. Make sure you protect the people that help you bend rules, often support teams lower in the organisation.
So, bend the rules to keep you going.

Summary
If you can’t get started with digital in your own organisation or when you are stuck, consider the eight practices that I described above as they will increase your chances to get unstuck again or start going. They aren’t rocket science and not all of them may be applicable to your situation. If you feel stuck, at least try a few to see if you can get going again. Like any transformation it is most important to keep moving.
In summary, these are the eight practices that I described:
- Pull, not push
- Sponsorship
- Small core
- Engineers turned data scientist
- Online + offline training
- Invest in self-service
- Never wait for the next/better/shinier tool/initiative/promise, never
- Bend the rules in favour of speed
Good luck with your digital transformation!
Opinions are my own and not the views of my employer