Digital Service ProductThis is part two in a three-part blog series from enterprise software thought leader Dr. Timothy Chou. Part one can be found here. Look for part  three in an upcoming edition of the AEM Industry Advisor


A digital service product has the potential to double the revenues and quadruple the margins of any company that makes construction or agricultural machines. If you assume a digital service product is priced at 1% of the purchase price of machine per month, if your company sells a machine for $200,000 and you had an installed base of 4,000 connected machines, you could generate $100 million of high-margin, recurring revenue annually.

So, with that fact in mind, how would you build a digital service product?

Getting Started

The first step is to have your newly appointed product manager define the digital service product. Let’s start with what it's not. Digital service is not break-fix support. Digital service is not based on human labor. Instead digital service is information that is personal and relevant information. But what kind of information? Digital service is personal and relevant information about how to maintain or optimize the availability, performance and security of the machine. Whether you make machines or you use machines, you've dealt with the frustrations associated with disconnected service based on human labor.

What’s the serial number of the machine?”

“What rev level are you on?”

“Can you tell me whether the light is green or red?”

“Let me find a time for Mary to call you back.”

“We’ll be there from 1–5.”

Learn more about the impact of disruptive trends and technologies by signing up for an AEM Thinking Forward event in a city near you.

Defining 'Digital Service Product'

Let's take a step back for a moment. What, exactly, is a digital service product?

  • Digital service is personal and relevant information about how to maintain or optimize the availability, performance and security of a machine.
  • Digital service products serve the worker, not the software developer or business analysts. The worker might be a pediatric cardiologist or a construction site manager.
  • Digital service products have millennial UIs and are built for mobile devices, augmented reality and voice interaction.
  • Digital service products use historical data. Most enterprise workflow applications eliminate data once the workflow or the transaction completes.
  • Digital service products use lots of data. Jeff Dean of Google Brain has taught us that, with more data and more compute, we can achieve near linear accuracy improvements.
  • Digital service products use many heterogeneous data sources inside and outside the enterprise to discover deeper insights, make predictions, or generate recommendations and learn from experience.

A good example of a consumer digital service product is Google Search. It’s an application focused on the worker, not the developer, with a millennial UI and uses many heterogeneous data sources. Open the hood and you’ll see a ton of software and hardware technology inside.

Once your digital service product R&D team has established the definition and requirements of a digital service product, the next step is to inventory the data you already have. As the product provider, you already have access to information in your document management, CRM, service management, parts inventory and call center applications.

The third step is to connect the machines and in particular develop a way to connect any of your legacy machines. Monetizing your installed base is the fastest way to double your revenues and quadruple your margins. When you connect the machines, make sure you have a way to segregate the data and put that control in your customer’s hands. There are four types of data:

1. Static machine data -- This is data about the machine, which never changes, or changes infrequently. A prime example would be the serial number.

2. Environmental machine data -- Machines exist in the physical world, so location, humidity, temperature and altitude all fall into this group.

3. Dynamic machine data -- Machines have a large number of sensors, which can provide data frequently. Wind turbines have over 500 sensors. Vibration sensors can sample at 10,000 cycles per second. 

4. Nomic (I made the name up) data -- A gene-sequencing machine produces genomic data. Agronomic data includes the nitrogen level at a particular location in the farm. I’ve used the term nomic data to refer to this type of data across any kind of machine.

Just as you do on your phone, you choose which applications are allowed to see your location or access your photos. You should build a similar model for the owners of your machines.

Now that we have established the data infrastructure, the next step is to identify the workers. We’ve said a digital service product delivers information, which is personal and relevant. This of course could vary from person to person. Using cars as an example, personal and relevant information is different depending on whether I’m the driver of the Mercedes or a member of the design team. In general, workers could include the operator of the machine, the customer’s service people, the OEMs, the process expert (assuming the machine is part of a production process) and even the R&D team that designed the machine.

Finally it’s time to invest and build the digital service product. You’re in luck. In the past 10 years there have been major innovations in cloud computing, AI, 5G and the edge. Today you can get 1,000 servers for 48 hours for $1,000 And the costs are only going down. Meanwhile, on the software side, there are more and more choices both as open source and as cloud services. Check out my article to learn about 16 classes of software that will be part of your digital service product.

While one day I hope there will be packaged applications -- as we’ve seen in back office ERP -- today your best alternative is to create a team mixed between internal and external people. While some of you may be able to afford a totally in-house team, the practical solution in a fast-changing world is to partner with one or more external suppliers.

While the task of building a successful digital product is far from easy, there's a pot of gold at the end of the rainbow for those who do.

Dr. Timothy Chou is a leader in the third generation of enterprise software, a computer science lecturer at Stanford University, the former President of Oracle on Demand, a technology consultant and an investor in emerging technology. Many AEM members may already be familiar with Dr. Chou, who partnered with AEM to present an executive business model workshop and was featured in an episode of the AEM Thinking Forward Podcast.

Stay plugged in to AEM’s Thinking Forward initiative for events, articles, podcasts and insights related to big data and other relevant industry topics to keep your business on the cutting edge.

Subscribe to the AEM Thinking Forward podcast on your favorite podcast app.

Subscribe to the AEM Industry Advisor for more perspectives from thought leaders.