What does Data Science mean for businesses

What does Data Science mean for businesses
Photo by Tyler Franta / Unsplash

Data Science (DS) has indeed captured the imagination of entire swathes of industries and domains. With the now famous statement that "software is eating the world" actually ringing true, businesses are undergoing digital transformation as a matter of routine.

As part of these digital initiatives, businesses are also looking to adopt intelligent systems, smart automation and data driven decision making. But the way this manifests in a business is different from one business to another and even from one industry to another.

Internet companies are famously considered to be innovating at a fast pace. This is perhaps true for many Internet companies but not all. But as one hops to other industries, the pace and appetite for innovation sees a marked difference.

The difference is a function of many reasons but I feel the most compelling reason that companies don't adopt tranformative processes and tools are the following;

  • Lack of awareness of what needs to be changed and why.
  • Limited knowledge of potential solutions and confusion around a myriad of options.
  • Change goes against established inertia and often the desire for change is unevenly distribured in an organisation with the average skwewing towards an apathy for change.
  • Lack of a plan that can describe the various stages necessary to undergo change.

The above reasons are typically the bread and butter of departments that are setup to effect change, but such departments aren't the norm in most companies.

Furthermore, unless these departments and groups also include representation from the business units they wish to influence, their guidance on affecting change will fall on infertile ground.

Changes that affect workflow and productivity can be seen as attemtps to automate away roles and jobs and not participating in these processes can lead to unnecessary malaise and stalling of any progress around what can truly improve business outcomes for the enterprise.

One way to start the process of digital transformation is to ask what is currently broken? This requires measuring the current process and identifying a metric that can be influenced. These metrics, often called KPIs can be useful if they truly measure outcomes without bias.

Using a contrived example, say your Sales is growing every month but your customer base is not. While measuring Sales (and seeing the growth) might point to a healthy business, you might end up blind to the fact that most of your sales come from a few customers.

A more robust metric might choose to measure the number of customers contributing to 20% of all sales, 50% of all sales, 90% of all sales. And what % of your customer base are contributing to 90% of your overall sales would yet be another metric to observe.

This allows you to track your dependency  on a few key customers and also identifies the opportunity to increase your customer base.

Similarly, if you have a process that can be digitized, it would be necessary to identify and measure the process by first breaking it down into its constituent steps and measuring the outcomes from each step.

If you are a Pharma company looking to improve drug discovery outcomes, you may need to identify the process from reading medical data sources, identifying useful excerpts all the way to the final step of recommending a potential lead.

Even if you automated one aspect, say identifying new data sources in a scalable way, the manual work of reading sources still remains a bottleneck. If you devised a funnel and identified where narrowing of the funnel takes place, then you have a very good place to start to make improvements.

If your excerption process can be improved, you can consider NLP techniques to help narrow down specific parts of the documemt for the expert to read reducing time necessary to locate relvant parts of a document (if this is found to be more time consuming that other steps).

Measuring your wokflow to its detail gives you a clear understanding of steps involved and where bottlenecks might be hiding. But more importantly it gives you the ability to think creatively about potential solutions to improve outcomes and to equip your teams with smarter and more delightful digital tools and solutions.