Insights from April’s Retail AI, Robotics, and Data Conferences (Part 2)
Last week we published part 1 of our insights from April’s two events focused on technological revolution in retail – the Retail Robotics and AI Conference and the Retail and Consumer Goods Analytics Summit. Our key takeaways from part 1:
- The ubiquity of “Data is Oil” may drive retailers to hastily design and invest in data strategies.
- A retailer must know what they want to stand for prior to initiating a broad data transformation – or they risk investing resources and time into projects that will not support their positioning.
- Technologies using machine learning gather consumer insights from millions of reviews, posts, and blogs that are published on the internet. These fast and relatively affordable machine learning insights can help refine the retailer’s strategic positioning.
- Once positioning is clarified, it is easier to define what your data strategy will be: what problem should you solve first, what steps should you follow, what investments do you need to make?
Our insights in Part 2 focus on a successful execution of a new data strategy:
Step 1: Recognize that not all problems can be solved by data analytics.
The sniff test for exploring where a data strategy could apply was summarized by Kristian Hammond, a Professor of Computer Science at Northwestern. Is your question or task truly data driven? Can you access the data that would support answering the question? Do you (or will you) have the volume of data required for analytics to be meaningful?
Step 2: Recognize that an internal analytics group isn’t enough.
A small group of data analysts creating insights and sharing them with a few at the top will provide limited benefit to your company. A key part of using data analytics is getting buy-in from your employees at all levels and building a data-culture. At the Retail and Consumer Goods Analytics Summit, Speakers from Kimberly-Clark and Mars both advocated creating workshops that educate senior leaders and then the broader organization on data analytics and their applications. An understanding of “What’s possible” leads to adoption as terms like AI and Machine Learning are defined and explained. Rather than building a siloed division, a culture of data spreads through the organization.
Step 3: Manage Expectations
Transforming an organization takes more time and effort than most will appreciate. Speaking at the Summit, Colin Reid of SAS suggested “current timelines are often unrealistic and we should likely add 40% additional time on to any project.” Why? Getting back to Hammond, “Clean Data” is difficult to find. The data an organization can gather must be identified consistently – think of column headers in an excel spreadsheet that reflect the objectives of the questions that need to be answered – as well as identifiers that are input as consistently as a password. This is not a 7-day cleanse. The volume of data required for effective machine learning goes well beyond the row limits within Excel.
Step 4: Find the talent.
Data analytics requires a particular set of skills. It requires individuals to have strong knowledge of qualitative and quantitative techniques and understanding of various technology platforms. If that wasn’t enough, data analysts and data scientists need to be able to distil statistics into a story that ultimately leads to a business decision. Statistics knowledge is a must, but not all statisticians can be data analysts.
Step 5: Define your personal privacy expectations and data security requirements. Steps 1-4 in building a data strategy were internally focused. Step 5 is focused on your customer – and key to getting their buy-in.
If data is oil then “AI and machine learning are the new oil derrick” (another Humby-ism – circa 2018). Retailers need to press the accelerator. But first strategies and competitive positioning must be defined. Leaders must identify the questions that machine learning and AI can address in order to support strategic execution. And organizations must take steps to prepare for change. To use another phrase coined by a Brit – “this is a brave new world.”