The explosion of digital data leaves many retailers unsure of where to start in creating a data strategy.

Michael Ross, co-founder and chief scientist, DynamicAction

Michael Ross, co-founder and chief scientist, DynamicAction

Digital technologies are catalysing a massive shift in consumer-facing retail.   Consumers have almost unlimited choice and information, they are using a range of devices, being bombarded by digital marketing and are shopping in new and complex ways.

Behind the scenes, the digital world is powered by a breathtaking array of technologies, each with its own set of operating rules.  Many also require completely new types of decisions—such as programmatic marketing, bids on Google/Facebook, personalised landing pages and website merchandising. All of this activity creates a vast “digital exhaust” —the big data trail of every decision, impression, click, search, view and transaction.

Any retailer, whatever its size or focus, has to work with this data in new and smarter ways. For data-led businesses [Google, Amazon, Zalando], data science is simply how they have always run their businesses—it is part of their corporate DNA. The challenge now is for traditional retailers to embrace data science in the same way. While most retail leaders recognise the need to think differently, they struggle to know where to start, what needs to change and what success looks like.

The enterprise must see data as a catalyst for thinking differently.

Data-led innovation creates the opportunity to think differently—to improve processes, save money and drive new revenue. But driving change in an organization is hard. As we have seen from the ongoing struggles of traditional retailers vs. Amazon, retailers must innovate or get left behind.


To reach the transformational power of data, a clear understanding and mindset must be implemented throughout the entire retail organization. The data manifesto suggests five principles that are core to evolving to become a leading data-driven, agile and innovating retailer.


  • Data needs to be on the CEO’s agenda. If it’s not, any data-led initiatives are highly likely to fail.
  • Data is a source of innovation. The enterprise must see data as a catalyst for thinking differently. If you end up making the same decisions on the same frequency, in the same silos, with the same data, same logic, by the same people with the same incentives…nothing will change.


  • Data is the new oil. Data is a critical corporate asset that needs to be discovered, mined, extracted and refined to turn into something useful, and then be joined across systems. In particular, data cannot reside in a system or organizational silo.
  • ‘Good enough” data is hugely valuable. Data does not need to be perfect or complete, but it does need to be insightful and studied to develop a clear path to action.
  • Make sense of the digital data tsunami. The plethora of digital data can be overwhelming, but it is critical to understand customer intent and exposure. Understanding what’s useful is not straightforward.


  • Not all decisions are equal. “Decision agility” is needed to take advantage of new insight and opportunities. The challenge is that many retailers still apply a one-size-fits-all approach to decision making.
  • Make “good enough” decisions. The objective is not to achieve perfection, but should be “how do we make the best decision possible given the available data?”
  • Celebrate mistakes as an opportunity to learn. It is impossible to improve if the retailer conceals failure and hides waste. The key is to minimize the loss, learn quickly and avoid making the same mistakes again.


  • Data is not one role. A variety of skills are required to make sense of data: data architects, analysts, algorithm designers, mathematicians and statisticians bring different skills to the table. A Chief Data Officer—or similar—is critical to act as a conductor of this new data-driven world.
  • Rethink silos. Many of the critical decisions of digital commerce can no longer be optimized in the organization silo.
  • Introduce decision product managers as a new role. They intermediate between the production and logistics owner, and the data scientists. Often called a “decision engineer” or “business performance manager,” they translate a business question into a math problem, and then interpret the results.
  • Plan for change management. Managers may not immediately encourage new approaches to data, as these new approaches can be unfamiliar or threatening. They could feel like their jobs are at risk.
  • Be clear on who the decision maker is. Everyone has an opinion about data. It is typical that any discussion around data ends up with 20+ people in the room with varying opinions, subtly conflicting objectives, and no obvious decision maker.


  • Everything is an algorithm in digital commerce. This results in the application of logic to data.
  • Averages are the enemy. They are often misleading and rarely representative. Outliers, deciles, dimensions and stratification are critical tools for unraveling averages. Whenever you are presented with an average/ratio/percentage, a good question to ask is “what’s the distribution”?
  • Instill a culture of analytical curiosity and constructive challenge. Many people think they are “good at data.” In practice, the best data scientists are the most humble. Data science is hard, messy, easy to get wrong and easy to misinterpret. Beware of certainty and defensiveness.
  • The devil is in the detail. Leadership teams often become detached from the detail. In the digital world, the aggregated or simple story will often muddle the real story.

Where does a retailer truly begin to effect change through empowering data analytics? The critical business challenge is to understand which techniques should be applied to which problems. Data needs to be de-siloed, aggregate metrics need to be de-averaged and aggregated analysis decomposed. Traditional approaches to analysis based on purely transactional data may be subject to reversal when additional dimensions are added.

It is easy for business managers to ask unanswerable questions, and easy for data scientists to do clever analysis that does not drive any decisions. When it comes to data and retail transformation, the key principle is to establish a clear direction and process across everyone in the retail organization.

DynamicAction provides analytics technology built for retail merchandising teams.