Data and analytics play increasingly vital roles in firms of all sizes and in nearly every industry. But can mid-sized firms that aren’t built around data still compete on data and analytics? That’s what data strategy is all about.

Keeping Up Is Not Enough


Many middle market firms are getting serious about data and analytics because they fear (reasonably) that competitors will eat their lunch if they don't. Some of the biggest threats come from digital native firms like fintech and insuretech startups. Soon, leveraging advanced analytics effectively will be a survival requirement in virtually every major industry.


But keeping up is not enough. Business strategy is fundamentally about being different from the competition—whether by offering a better product, a lower price, or defining a new niche. Middle market firms need to think deeply about how data and analytics can set them apart. To determine how they can do this, companies need to consider where on the data value chain they can add value.


The Data Value Chain


From acquiring data to communicating insights from it, there are several links in the data value chain. Let’s consider each in turn.


Data acquisition. Elite tech firms stand out as ravenous acquirers of data: Google, Facebook, Amazon, and others have built business models around services that are woven into consumers’ daily lives in ways that allow these firms to collect vast personal data on their users. Middle market firms typically collect a much more limited range of data about their customers—and have a much smaller customer base.


But this doesn’t mean a handful of firms are destined to capture all the value from data acquisition. Just because Amazon knows exactly how much you'll pay for socks on a rainy Tuesday afternoon doesn't mean they have the data to answer the crucial questions in your industry. In fact, Amazon, Microsoft, and Google are now building AI business models on the assumption that less tech-centric firms have valuable data—and will need these giants’ technology and expertise to leverage it.


The predictive value of data is highly domain specific. Industry incumbents that figure out how to acquire the most relevant data from their existing customer base can use that data to build a moat and secure their incumbent advantage. Others will lose market share to upstarts who master the data game.


Data aggregation. Data aggregators and data brokers like Acxiom and LexisNexis get data on consumers and businesses from a vast array of public and proprietary sources.  These firms build pipelines to ingest and clean data and package it for use by other businesses. Middle market firms that need data on consumers and businesses beyond what they can collect from their own customers are usually best served by buying it from data aggregators. Data aggregation requires massive resources; it benefits from returns to scale. This is not the place for most middle market firms to add value.


The cost of data aggregation will decline as a new generation of tools streamlines and automates many aspects of data engineering. As a result, commoditization may gradually replace barriers to entry as the reason most middle market firms cannot compete on data aggregation.


Data science. This is the exciting part of data because it’s where the insights come from. There is a major skills shortage in data science; recruiting presents serious challenges. And retaining talent is not just about salary. Data scientists want opportunities to grow their skills and work on the most interesting problems—and with the most advanced technology. This gives elite tech firms like Google and Facebook a huge recruiting advantage.


But middle market firms in legacy industries can still get ahead by using advanced analytics where their competitors have not yet done so. If the competition’s analytics are still limited to business rules—or even to traditional statistical models such as linear regression—it’s possible to get an edge with more sophisticated machine learning techniques. This can be accomplished by building an in-house data science team, outsourcing, or a hybrid approach.


Industry leaders like Google have open-sourced some of their most powerful algorithms. Meanwhile, by automating many data science tasks, analytics software designed for non-coders is making it easier for business users and BI team members to dabble in data science. As a result, much of what we consider cutting edge data science today will become standard practice. This trend will make it harder for all but a few elite firms to set themselves apart with superior analytics.  Advanced analytics will be a must-have, not a differentiator.


Data storytelling. For the insights from data to be valuable, they have to be explained to humans, by humans. Or do they? AI is challenging both sides of that assumption. “Black box” AI models make recommendations that cannot be explained easily—if at all. Meanwhile, AI products are automating the process of generating a natural language narrative—telling a story—from data. Developments like these threaten to one day commoditize the art of data storytelling.


But for now, firms that learn to tell compelling stories with data will have an advantage. It's not enough to hire good data storytellers—there also need to be C-suite leaders willing and able to hear them. This is a challenge for many middle market firms that have not traditionally been datacentric—and an opportunity for those that can figure it out.


Threats and Opportunities


The rise of data science poses a major threat to legacy firms that don’t keep up. But there are also opportunities for firms whose leadership is prepared to seize them. Middle market firms should ensure that data strategy is integral to their business strategy and that data analytics has strong champions in the C-suite. These firms should aggressively explore ways to capture data that will help them build a moat, while hiring or contracting the expertise they need to leverage their data.