Tips on Creating Customer Value from your existing Data Science practice…

For years, companies have been monetizing the data by selling it as such. We have numerous examples of this from market research firms such as Gartner, and IDC to technology and data companies such as HERE, INRIX,, etc. However, it is becoming clear that with data commoditization, the art of business scale is not in selling the data but using it as an underlying platform on top of which more valuable products and services can be offered.

Of course, this is easier said than done and over the last 5 years, with the rise of ‘Data as a Service’ jargon many companies have tried and 80%+ of them have failed to monetize is as such.

However, some companies have been able to identify clear use cases to drive the business scale. These companies have a keen understanding and continue to harvest the options of data science helping them provide actual customer-facing or customer-enabling value within their existing products.

For example, MOOCs use data science to provide better service to their customers by recommending most applicable courses, monitor how students are progressing and optimizing the lecture content and pace, etc. E-commerce companies use analytical engine that uses historical data together with matching learning techniques to provide high-quality, personalized, in-product recommendations. Another example is an anti-spam product that uses deep learning to categorize email.

In this case data science augments a company’s engineering capability by providing statistical and big-data functionality that becomes part of the actual product experience. Or perhaps, Amazon is the best example here.

The best examples of intrinsic data monetization that has reached the scale through platform are Google, Facebook, Amazon (

I see Twitter as a company that has a platform scale potential but it is still experimenting with best monetization approach.

I really like what I am seeing from Foursquare.

On the other spectrum, think about UBER, HERE, INRIX all using their data to provide value added service either as an underlying structure for autonomous driving, city/infrastructure planning, or simply traffic navigation.

If your business is sitting on the data that has been monetized as a data feed today and you are thinking of scaling that business further, consider the following items prior to making the leap?

  1. Do you have a data set that is generated by the intrinsic operation of your business? In other words, do you possess and ongoing, easy and controllable access to the unique data set that is otherwise difficult to obtain? — Think about how you are currently obtaining this data. If it is obtained externally, think about the acquisition costs as well as can the source turn off the data access at any time? If data is not generated by your existing business, the business scale in a long run may be questionable.
  2. What are the problems this data and platform can address? What are the Use Cases? — Think about the problems your current customers are talking about. Can you think of any use case/scenario in which access to your data platform on an ongoing basis would address those problems?
  3. Who has those problems? Who are the potential Customers? — If you existing customers have challenges, think about who else would benefit from access to the data and insight? Try thinking outside of the box, data is used for Financial, Buy/Sell Transaction, Advertising, Operational Optimization, etc. purposes. Can you think of any business in these areas that would find the access to your data valuable?
  4. How big are those problems? Are they nice to have? Or a must have? — Think about the use cases and value they bring to your customer. What is it worth? What is the size of these opportunities in $? Does it appeal to masses like FB/Google/Amazon or to specific industries like HERE/INRIX, etc?
  5. How are they solving those problems today? Are there Substitutes? — Think about the problem you are solving and how your potential customers are doing it today. Do you potential customers have a easily accessible and affordable substitute? Like it or not, Excel or Google Sheets are the most common substitute for easier functions or small data set challenges. In recent years, Tabelau/Qlik/Domo are also considered natural substitutes especially if visualization, more advanced analysis, and access to data from various sources is required.
  6. Is there Competition? — Anyone else offering such service? If yes, who are they? Can you differentiation and how? If not, why do you think that is?
  7. Do you have in-house Subject Matter Experts who can analyze this data through customer lens and work on platform definition? If not, can you easily access them? — This is a key! Unlike many product management roles, this product management role requires deep understanding of the business for each use case.
  8. Does your Technology Stack scale with this business? — This is another very important aspect of ensuring your team and platform will scale and support the business needs. I have seen many companies try to build their Data Platforms using traditional BI infrastructure and tools. If you are serious about your Platform business this is a BIG NO, NO, NO! There is a difference between Data Lake of stationary and well defined structured data and a real time process of analyzing data using deep learning and delivering insight in a SaaS model. I will write more about this in the follow up posts.
  9. Do you have Analytical brain power to understand your use cases and can select appropriate Analytics Model(s) to support those? — There is a difference between Data Engineer and Data Scientist. These are hard to come by resources. There are many traditional analytics platform and as of late open source solutions from Google, Amazon, Microsoft and Facebook available as an open source, but one vs. other could make a big difference for your problem. I will write more about these in the follow up posts.
  10. Do you have a GTM/Distribution/Sales Team that can help create the ecosystem of buyers to monetize this platform. Platform monetization is much more complex than traditional B2B sales approach. It is what I call a spider web approach to selling. You need to go back and think about your use cases, your potential buyers and create a grid approach to capturing the market — use case by use case, data point by data point, API call by API call, etc.