PD fallacy #3: The start of production is the end of R&D

4 min readJan 15


Photo by Gerd Altmann on Unsplash

In the previous post, we focused on the role of platforms to accomplish a constant evolution of the products we manufacture. As such, we accept that the traditional premise of manufacturing — producing as many copies of the same item as possible for as long as possible — is no longer a viable way forward.

However, there’s an additional dimension: the evolution of individual product instances over time. One of the key differences between traditional and digital companies is their view on the start of production (SOP). Traditional companies view SOP as the end of R&D since manufacturing of the product starts and we can move on to other projects. Digital companies view SOP as the real start of R&D because we now have a feedback loop with the customer and deployed systems that we can use to inform our decisions concerning system evolution.

The key challenge many companies run into is the discussion on what needs to be present in the product or offering when it leaves the factory. The traditional view is that the product should have as many relevant and differentiating features as possible since many of these will commoditize over time. This means that from the point production starts, the product gets a little less relevant every day as competitors catch up, customers become less interested and markets shift. Consequently, to extend the lifetime of the product as much as possible, we need to push as much differentiation into it as we can during the R&D stage.

The view of digital companies is quite different: at SOP, the product should have all the necessary enablers in place as electronics and mechanics are often expensive to replace. However, in terms of functionality, we’re primarily interested in the minimal requirements customers need the day they get the product. From that point on, we can continuously deploy new functionality and features into the systems in the field to get to a continuously improving customer experience.

The challenge with the traditional approach is that we need to guess what will add value to customers. So, pushing for a large number of features without evidence that these indeed add the expected value easily ends up in featuritis and a very complex product. Complexity in products isn’t a problem in itself, especially for expert users, but only if all the functionality captured in the complexity is indeed used by the users. A complex product with hundreds or thousands of features where each user only uses a small fraction is likely to fail due to lack of usability.

Digital companies build the known, must-have features before SOP and then use the connected nature of modern products to iteratively add functionality. This lets us measure the actual impact of a slice of a feature and direct our R&D efforts toward the functionality and features that add customer value. It also allows us to not build features that fail to add value as well as remove features that aren’t used at all or only by a few customers.

Interestingly, the simple ability to measure feature usage is already incredibly helpful as it directs our R&D effort and avoids all the difficult opinion-based discussions in companies about features. Many people in companies have quite diverse opinions on what really matters to customers and using data instead of the opinions of people who may not even have met a customer during the last year or so provides a significant improvement in the effectiveness of R&D.

The main advantage of connected products is the feedback loop that allows us to experiment with new functionality, such as A/B testing. This facilitates ensuring that we only build features that are really used, but it also helps improve already implemented features to maximize the value delivered through each feature. Rather than adding feature after feature, in many industries, optimizing the features already present actually gives a higher return on investment.

We should view the start of production as the start of a continuous feedback cycle between us and our customers. This allows us to optimize the system functionality to minimize complexity, only add features that deliver value, optimize the way features are realized and remove features that aren’t or hardly used. Life starts at SOP, it doesn’t end there!

This week, I am leaving for a trip where I lack internet access for most of the time, meaning I will not post an article on January 22 and the one on January 29th will be later in the day.

Like what you read? Sign up for my newsletter at jan@janbosch.com or follow me on janbosch.com/blog, LinkedIn (linkedin.com/in/janbosch), Medium or Twitter (@JanBosch).




Academic, angel investor, board member and advisor working on the boundary of business and (software) technology