We are frequently approached by many startups and established small businesses who want to get data science help. They would have collected (or obtained access to) large amounts of data, which they rightly believe would be instrumental in the success of their business. Obviously, the hype about data science has carried wide and deep and they immediately start looking for data scientists to hire. The savvy ones realize that not only acquiring the right person is quite challenging, but maintaining one on payroll for expected or anticipated needs is very expensive.
The advise we provide to all of these SMBs is the same: develop a data strategy first. (What is meant by data strategy?) Below are three examples of businesses which are varying stages of maturity, their analytics motivation and best course of action.
Company 1 (very early stage start up):
This company has no revenues, only an alpha stage product which relies on analyzing multiple public data sources to deliver value to customers. The company has access to dozens of data sources which it plans to leverage to provide answers to some common marketplace questions. The founder recognized correctly that data science will play a key role and wanted to "hire" one. Finding one and keeping one on staff for an early stage start up is pretty much impossible. But the fact is that at such an early point in their development, they do not need a full time person for this role. To understand how much data science effort is involved and to chart out a clear roadmap for the future, this company needs to invest a very tiny amount of money into data strategy.
Company 2 (start up with several customers):
This company has a product that has gained good traction in the marketplace and now wants to expand the scope of its offerings. Once again, the executive team has realized that data and analytics are the right way to take this forward. They have also recognized that predictive analytics will be a differentiator for their product and now wants to move ahead with developing, validating and deploying models built using their data. But they have questions about whether they have all the data they need, what are the performance targets for their models, what IT infrastructure is necessary to support their predictive model deployment and so on. These are more mature questions than company 1, but still are questions that are best answered with a well defined data strategy.
Company 3 (a mid sized, well established business in a "traditional" vertical):
This company has been around for more than 50 years and is no startup! However they are totally new to analytics, although they have been collecting and utilizing data all along. What has changed is the availability of high performance technology at an affordable price: big data, cloud and machine learning. Here too, the executives realize that analytics will help their company lead their vertical as an innovative problem solver. They are sitting on terabytes of data but need to determine which are the low hanging fruit that analytics can address in the short term. They also need to understand what pieces of the technology stack they should invest in before they can hire a team of data scientists. While obviously more mature that the earlier two examples, they too need a clear data strategy and a roadmap.
The bottom line is that all businesses (small or otherwise) need to take a step back, evaluate their data assets and reexamine their business objectives before jumping on the latest hype bandwagon.
Our affordable CLEAR (Comprehensive Level of Enterprise Analytics Readiness) package allows any business to get to that point quickly. Download our brochure for timing and pricing information.