BI trends: What to expect in retail in 2019?
To help retailers and brands plan for 2019, Researcher Claudia Tajima and Fiona Swerdlow are interviewing experts within Forrester for their series, ´Applying 2019 Predictions To Retail´. This week, Claudia interviewed Jennifer Belissent, Ph.D. and principal analyst on Forrester’s consumer insights team, on their 2019 BI predictions report. Here’s what Jennifer thinks retailers and brands can expect and should focus on regarding BI for the rest of 2019.
Claudia:Your first BI prediction for 2019 states that companies cannot be successful simply selling raw data and that self-service data marketplaces will struggle. BI tools need to start delivering data insights and services. How does this shift affect retailers?
Jennifer: For retailers today, there is a vast amount of data that you could use to improve business and better understand your customers. Many retailers already use their data to understand their customers and forecast trends. But today is a rapidly evolving landscape of new, alternative data sources. Opportunities to enrich data with new sources are appealing. However, retailers must evaluate those opportunities carefully. Why? The time to value is longer when buying raw data instead of buying data insights. For example, buying a customer’s credit score would be faster time to value than buying the raw customer data to ultimately find their credit score. My recommendation is that retailers should not rush to buy data or expect to be able to buy data from a marketplace and get all the answers they’re searching for. In some cases, retailers need insights service providers to interpret this data.
Claudia:How will the demand for data storytelling skills impact retailers’ talent acquisition strategies?
Jennifer: This demand exists because there is a gap between technology users and data scientists. They don’t always speak the same language, but a storyteller can bridge the gap. Organizations need a storyteller who can talk to the business team, data team, and the technology team and help them reach a common understanding. This balance is critical for BI teams to be able to both organize data and deliver the data in a compelling way. Forrester’s research suggests that more mature companies, those that are more ´insights-driven´, have these skills.
Claudia:Organizations are predicted to abandon unactionable BI reporting and dashboards. How should retailers respond to growing derelict dashboard graveyards?
Jennifer: Retailers are seeking answers to questions such as: How does one store compare to others? How does it compare to regional sales? However, retailer leaders’ interest in specific reports or dashboards eventually goes down over time. Creating a data center of excellence increases more data awareness, but it also brings about a frenzy of requests for new dashboards and reports. Ultimately, many of these requests end up as orphaned dashboards. It is important for retailers to be careful of how they embrace data democratization. Take time to step back and rationalize, prioritize, and determine which data from reports and dashboards you need and don’t need.
Claudia:Why should retailers consider adopting data fabrics in place of data lakes?
Jennifer: In the past, many organizations chose to put their data into massive data lakes. However, these organizations did not fully think through how their data lakes should be organized and used. Today, organizations are starting to realize that there is no major benefit to putting all of their data into one centralized data lake. The new trend is to create a data fabric of woven data from across different parts of the organization that sits somewhere central. Data fabric stores maintain their own individual data, but there is a central data point where it can all be accessed.
Claudia:What recommendations would you give to retail leaders looking into investing in BI tools in the coming 12 to 18 months?
Jennifer: Data catalogues, which serve as a knowledge repository, are becoming very popular. Organizations typically have one centralized data catalogue. Interesting data catalogue outputs include: use cases, algorithms, as well as which reports the data has been used for and where has it been tested in sales. In terms of ambient data governance tools, retailers should look for BI tools that have data governance built directly into them.
Author: Fiona Swerdlow