One of the models that I use frequently in my consulting assignments is the 10 types of innovation model by Doblin/Monitor Group. As the name suggests, the model identifies 10 different forms of innovation arranged around the business ecosystem, the offering and the customer experience. This model is important as most people think about improving the product when they hear the word innovation, but in practice, as the authors of the model show, investment in product innovation has the lowest return on investment (RoI) of all the 10 types. Instead, innovating the business model, customer experience and surrounding processes have significantly higher RoI.
One of the innovations that is not necessarily concerned with the product but with the other dimensions of innovation is DevOps. Although the term DevOps is used frequently, I still notice that many fail to identify how deep the impact of the adoption of DevOps is. It fundamentally changes, among others, the interface to the customer, the product in the field and, frequently, the business model. As we have discussed throughout this blogpost series, digitalization is concerned with continuous value delivery to customers which is predominantly driven by releasing of new software to products in the field. As the customer often is the owner of the product, there frequently is some involvement by the customer in order to ensure that the update happens when it is convenient for the customer.
The relation to the product in the field changes in at least two ways. First, traditionally software was extensively tested as part of the integration in the product before being shipped to the customer and this included several manual steps and activities. Now, the software is released and the product is expected to update itself without causing any issues or concerns. Also, the product needs to be able to roll back if an issue is detected in order to ensure availability.
Second, with the continuous connection to the product, we also want to collect performance and quality data and bring it back to the company for analysis. This connection forms the basis for all subsequent technology innovations, such as AB testing and the use of artificial intelligence.
Third, accepting a cost structure associated with continuous deployment of software to the field but failing to capture the additional value that you are providing is suboptimal in that it decreases margins. So, the adoption of DevOps often causes a change in the business model as well where the transactional business model is replaced or complemented with a continuous business model, e.g. subscription, usage or performance based.
In my experience, there are at least two challenges that many embedded systems industries are experiencing when digitalizing, i.e. certification and the operating models in the business ecosystem. First, most certification approaches assume a “one-off” certification of a system, including all its technologies. This certification activity is very labor intensive and expensive and consequently many companies are very careful to conduct a certification as infrequent as possible. This is diametrically opposed to DevOps as we want to update our systems in the field as often as possible. There are several approaches that address this and offer continuous models for certification, but in practice these have not yet seen major adoption in many industries yet. Depending on the class of certification needed, this requires process innovation not just at the company level, but at the business ecosystem level.
Second, the operating models in the business ecosystem often complicate the adoption of DevOps. For instance, in the defense industry, the customer, typically the military of a nation state, assumes a transactional model where all requirements for a new system are defined before a tender and the winner is selected based on price while satisfying all requirements. Again, this way of operating again severely complicates the adoption of DevOps and requires changes at the business ecosystem level before the full benefits of DevOps can be captured by everyone involved.
Concluding, although the term DevOps is used widely and many assume that it’s simply the combination of development and operations, in practice DevOps significantly changes the relationship with the customer and the products in the field as well as the business model. DevOps is a keystone technology that drives many changes in the company as well as in the business ecosystem. The adoption of DevOps is often complicated by certification needs and existing work practices in the business ecosystem which requires a multi-pronged change management approach in order to realize it. But, as we all know, the best way to get better is through continuous improvement. As Mark Twain said, continuous improvement is better than delayed perfection!