There are infinite ways to carry out a digital transformation.
How do you know which way is the ‘right’ way for your business?
(Preferably without getting caught in endless analysis paralysis).
There is one methodology that we have been using with our clients to help them assess the feasibility of different use cases, technologies and approaches for their business: the digital incubator.
The digital incubator (DI) is a methodology for rapidly testing hypothetical solutions to key business problems with minimal investment of time and resources.
In other words: the digital incubator solves the business problem of determining the best approach to solve business problems!
For example, if you want to test a new technology, but there are 3 different tools available, you might use a digital incubator to assess and test each one or all of them quickly and see if they meet your needs.
We’ve developed our own Mesh-AI approach to running Digital Incubators, providing bespoke teams to clients to help them rapidly determine the next-best-action and make the highest-value business decisions in complex environments.
In this blog, I’ll explore why enterprises struggle to run digital incubators or similar methods, how it works, the business benefits and to close I'll give some real-life examples of how they can be used.
We see that our enterprise clients have a whole heap of different priorities and goals, but there are often so many different ways of getting the job done that they struggle to take the next step.
And, most importantly, with their tech teams already under a lot of strain and pressure, finding people with the time (and willingness!) to run these experiments is challenging.
There are a number of constraints and challenges that most enterprises face when trying to do this kind of rapid testing:
1) Lack of clear responsibility
Most teams within a large enterprise already have a long list of priorities that they are striving to get done.
There is no specific team that has responsibility for testing these kinds of things, particularly when these tests might run across business departments (e.g. data discovery tooling that affects the whole business).
2) Internal bureaucracy and overburdened teams
Even if there were a specific team, there is often a high degree of internal bureaucracy, delays and handovers means that it takes a long time to get anything moving.
3) Lack of specialised skills
When testing more advanced use cases, such as data science or machine learning, there can be a very steep learning curve.
This delays time-to-value even further, because your people need time to learn the technology and best practice approaches first, before they can even start to test its feasibility as a solution for your business.
The end result of these challenges is that, under normal circumstances, it can take an enterprise business six months or longer to validate prototypes and hypotheses, which is simply too long when you’re trying to accelerate your transformation and demonstrate clear business value to defend your budget from cuts.
That’s why we developed our DI approach: to help enterprises skip the need to hire a team, upskill them and wait 6 months to find out if their idea will work or not.
We help them to cut straight to the answer. So how does the process work?
A digital incubator is more a mindset than a specific technology or environment.
It is an agile approach that focuses on starting small and failing fast. It takes one (ideally well-defined) use case and provides a container for testing different approaches.
The goal is not necessarily to build out a small-scale technology as a prototype before scaling, but rather to rapidly determine the best course of action from a range of potential ideas.
For example, many enterprises want to move towards an event-driven architecture. But the first step might be to ask what tool you need? Is Kafka the right one? Or what about RabbitMQ? Do we use a managed service? Which is the best for our use case?
You could guess and then end up realising you’ve made a mistake further down the line. Or you could test the waters with a digital incubator.
As I mentioned above, enterprises struggle to focus their resources on this problem, which is why we developed our own Mesh-AI digital incubator approach.
We put together a cross-functional squad of subject matter experts whose sole responsibility is executing the digital incubator.
This solves the skill and responsibility issues and provides a dedicated, project-ready task force to carry out the project.
We have found that teams of four to five work best, with two to three experts from our side teamed up with an executive sponsor and a subject matter expert from the client side.
When we set things up in this way, we find that, rather than taking six months, a valuable digital incubation can take only six weeks.
During the first week we define and qualify the use case. The second to fifth weeks encompass building and running the tests. The sixth week would be for review and knowledge transfer.
Let’s get concrete about the business value of the DI.
We have found the following three to be the main sources of value for our clients:
1) Rapid priority and feasibility testing
The DI can give you an idea of how feasible a given technology or approach is for your needs and how quickly you will be able to execute it.
2) Rapidly demonstrate business value
If you have a hunch that a specific technology or method should be valuable to the business, then you can test that hypothesis most quickly in a DI environment.
If we think that a certain tool should save costs then we should see evidence of that in the DI, which is useful for demonstrating the value to the rest of the business before scaling it further.
3) Develop a tested, repeatable methodology
If you have tested a given technology in an incubator (e.g. event-driven architecture with Kafka) and it works, then you can scale that technology much faster because the knowledge on how to use it is already there.
Here are some real-world examples of digital incubator work that we are doing with enterprise clients:
Change Data Capture (CDC) Using Kafka
The client had a problem syncing two databases that were being updated via ETL. They wanted to test the feasibility of using change data capture via Kafka to synchronise the databases, as part of a potential larger move towards an event-driven architecture.
We set up a 6-week DI to test CDC on Kafka in a small area of the business, customer operations, which was suffering from data integrity issues leading to incorrect customer invoices.
This DI is currently ongoing but results are currently pointing towards an increase in data integrity which serves a couple of purposes: solving the invoice problem, leading to increased customer satisfaction, as well as opening the door to more advanced data science approaches, such as machine learning.
Data Discovery Tooling
Another client had a lot of data stored in different places and wanted to move towards a data mesh architecture that would help them make the data more trustworthy and accessible.
But they didn’t know which data discovery tool to use as part of the mesh. They turned to Mesh-AI to conduct a feasibility test across a range of tools across different categories, such as enterprise versus open source, for example.
By themselves, it would have taken 3-6 months just to find, hire and upskill the people they would need to get started give the challenges in finding talent. Then it would have taken a further 9-12 months to see significant business value.
By working with Mesh-AI, we were able to build the test environment in the cloud, carry out all the tests and deliver a clear outcome in 6 weeks. They now know which is best suited for making their data as easy to find as a Google search.
If you recognise that your business doesn’t have the scope to test potentially cost-saving and value-creating new hypotheses, then consider taking the digital incubation approach.
It’s a rapid, low-cost approach to making high-quality business decisions that have been shown to deliver value.
If you want to be competitive, you need to sort your data constraints, and that's where Mesh-AI can help. Identify the areas in your organisation that require the most attention and solve your most crucial data bottlenecks. Download our Data Maturity Assessment & Strategy Accelerator eBook.
Interested in seeing our latest blogs as soon as they get released? Sign up for our newsletter using the form below, and also follow us on LinkedIn.
