• Data Science
  • Derek Russell
  • DEC 11, 2017

Pour Another Coffee! - Why ALL Businesses Will Become Data Businesses

Ready to learn Data Science? Browse courses like Data Science Training and Certification developed by industry thought leaders and Experfy in Harvard Innovation Lab.

In many of my conversations with CTO's and CIO's, throughout a breadth of business verticals, the most common themes I have been taking away sound something like: "our organization is behind and we need to revitalize - we now want to become a data driven business". If this sounds familiar, know that you are not alone, and there are plenty of great resources you can partner with to provide a platform for which to drive towards this transformation (think Microsoft, IBM, Amazon, and Google), and a breadth of smart systems integrators (PWC, Accenture, etc.) to help deploy. But this isn't a race to stuff products and solutions down your throats, this is the first stage beyond a courtship, and the beginning of a symbiotic partnership. Understanding your business, how it operates, and what's most important to your customers has become the center of an elegant technology conversation. Where an organization's data lives, what it's used for, how it's analyzed, and who needs to make decisions based on trended insights in the data are not just in IT's wheelhouse any longer – this now belongs in the c-suite. 

According to a research press release by the International Data Corporation (IDC) just this past March of 2017, > 47% of the IT spend in an organization will now belong within core operating business units, clocking in at a 5.9% compound annual growth rate for line of business (LoB) spending. (See the chart below as to how this spending is allocated). Which business areas is this most prevalent in? According to the same research, the top 5 LoB's include: "discrete manufacturing, healthcare, media, personal and consumer services, and securities and investment services". 


(Areas with most LoB spending in 2017: applications, project oriented services, and outsourcing)

Alright, so lets not start picking on IT, they aren't going to be replaced by robots anytime soon and they still carry the most important business functions in a data driven organization: security (they help keep our data secure from cyber-attacks, remember Equifax?). IT is now more integrated than ever into how a business utilizes their data, and they can't/won't be doing this alone. But while the CTO & CIO work to build a new runway for this multi-node technology airport, you, the VP of <Insert Marketing, Finance, Operations, Sales> will need to land the aircraft.

"C-Suite, your DATA needs your help!"  

Ginni Rometty, the CEO of IBM, said something at her keynote speech for CES 2016that was very sticky for me:

"Data is the world's most precious

natural resource...80% of data is

dark...the challenge is understanding it".

The idea is that dark data, or data that is available for collection in every moment in our lives, is a rich resource just waiting to be collected by businesses who are strategically positioned and enabled to do so. 

Think of a typical journey map of your week (or even more relevant to our convo, think of a customer journey map in a retail center: see below). Monday morning you drive to work, you eat lunch with your colleagues, you send some emails. Sure, we know how many miles we drove and how long lunch was, but there is some valuable telemetry that is missing here: how many minutes did we listen to that podcast on our commute, and how did it affect your productivity on a project from 9-11am? Did eating a specific meal at a certain time impact the number of emails that were sent, the length, or the success at getting a fast response? Does the way we drive and the route we take correlate with how well we start or end our day, measured by heart rate, sleeping habits, etc. Setting aside privacy for the sake of these ideas, imagine how telemetry behind just the moment to moment way we live our day could help trucking fleets decrease accidents, HR improve productivity or increase retention in the workplace, and empower university's to construct curriculums that intelligently contribute more to prime learning environments. 


(A journey map of the typical customers experience when purchasing an athletic clothing good at a retail center: Much thanks to my fellow "Happiness by Design" team members, whom I presented with in a "Design Thinking" business case challenge in Austin - our prototype won first place.)

According to some predictions by Gartner, a leading global technology research and advisory company, IoT (including the collection of ambient data) will save consumers and businesses $1Trillion per year in consumables, services, and maintenance costs by 2022. The point is, ambient information in our daily lives can be captured and utilized to draw patterns, so that there is a deeper understanding (think FitBit allowing us to measure sleep patterns and walking activity) of who we are and how we can behave to meet/exceed optimal benchmarks. This is an invaluable business opportunity, and collecting structured and unstructured data into a safe repository, so that it can be extracted, modeled, and understood, will allow further penetration into the white space of serving your customers. AI of course, the hottest topic right now and in tech for the past 20 years (see the persistence of AI in companies' earnings calls below), is now shifting to the hottest topic in business, but AI cannot exist without the massive curriculum of data that it must sit on top of and learn from.


Microsoft has done something mind blowing with the "connected cow".  SCR Dairy, a herd management company, partnered with Microsoft to help assess when female cattle were in heat, so that they could be artificially inseminated at the correct time for optimal calve production. The data was gathered on IoT devices attached to female cows, correlations were determined with AI modeling around the number of steps female cows walked with their propensity to be in heat. It didn't end here, with all of the data that was collected, more insight were pulled to create "dark revenue" including patterns relating to the number of steps walked for highest probability of female calves born (which offer greater revenue generation), health of calves, etc. The system was also designed for mobile deployment to help herd managers scale while out in the field, decreasing the cost of labor and increasing harvested data. Imagine the use cases that can stem within your own organization?

Lengthening the Long Tail: Making $$ Out of the Data Story 

As businesses embrace becoming more digital, their ability to participate in generating additional revenue streams will become vastly more prolific due to our friend: The Long Tail. This theory follows that businesses, when they focus on specific strategic initiatives (services and products), incur maximum revenues at these focal points , and revenues drop as they move away from these points and into ones they are not familiar with. Although, as additional products and services are added, revenues are now aggregated across these additional product and service vectors as depicted in the chart below. Of course, this is assuming that these additional new points are not extremely cost prohibitive or asset heavy (enter data story). Amazon vs Walmart, before Walmart became a "brick and click", is a perfect example of the Long Tail in action. Walmart may only want to stock the "top 40 books" in its brick & mortar, while Amazon could theoretically stock as many titles as it wanted to in its vast network of connected warehouses. Although the niche "cat books" offered by Amazon aren't best sellers, the wide availability of titles cascade down the length of the tail, adding to a greater revenue sweep.


(Imagine this graphic, but aligned with "strategic initiatives" on the X axis, and revenue on the Y. Notice how revenue can aggregate indefinitely?) - Thank Martin Ford, for these ideas in his book: "Rise of the Robots".

Back to data. Let's take a car maker for instance. The focus area of a traditional car maker may include several different revenue streams, mainly, car sales and services (ie Onstar for GM). Now what if GM were to position itself to strategically utilize it's data estate to offer many more revenue generating services using the data, which individually aren't so impressive, but in aggregate accumulate to provide an impressive "long tail" of product/service streams and of course, revenues. Building a mobile platform for users to be contacted when their vehicle is up for maintenance, based on predictive analytics around tire wear and brake use, generated from years of historical and streaming (IoT) service data. Using this same data to provide forecasts on when customers would be coming in for particular maintenance items, providing for just-in-time delivery for a connected supply chain, intelligent labor scheduling, or prescriptively recommending sales teams have particular cars charged/gassed up and test drive ready for customers coming in for routine maintenance. Could the data then be used to calculate multi-tiered consumption model memberships for customers to pay monthly, and drive a "shared" fleet of vehicles?

We've asked each other a lot of questions here, and we've covered a lot of ground - the net of it all, there are so many different ways technology has crept out of the cost center bin, and into the growth center bucket. And it's happened very fast.

Now, what are some things you can start doing this quarter to make sure your business is more of an economic driver, with a data driven co-pilot?

The Harvard Innovation Lab

Made in Boston @

The Harvard Innovation Lab


Matching Providers

Matching providers 2
comments powered by Disqus.