Dress to Impress

By August 21, 2019 August 27th, 2019 FIND Papers
604FDA - The FIND Predictive Retail AI Blog

Dress to Impress

August 21, 2019

Keeping it personal can lead to

Higher Open Rates
Higher Click Through Rates
Higher Conversion Rates
11 min read

Prioritizing resources to look your best.

Successful apparel retailers are beginning to crack the analytics code — and are seeing substantial benefits in several ways. From enhanced planning and allocation decisions to more appropriate and timely marketing, analytics increase top-line sales, increase margins, and improve the bottom line by 5 to 10 percent.

Additionally, new cloud-based systems streamline the time it takes to deploy to timeframes that are measured in weeks and deliver returns that are well above ten times the investment capital to add millions of unrealized profits to the P&L.  In short, advance predictive analytics are changing the game and in some cases, can make all the difference between having a winning year or facing dire consequences that are becoming more common as we read the headlines in the 2019 apparel retail market.

To take the first steps to deploying predictive data analytics, apparel players should review their business priorities and understand where analytics may be able to make the highest impact both to the bottom line and to the day-to-day workings of each team member

Given the ability to impact the bottom-line and prevent inventory management missteps, predictive analytics is therefore a key area of focus that must be on the top of any apparel retailer’s short list of priorities.  The question is how best to prioritize limited resources to get started and use your data to look your best. In this post, we recommend that retailers begin with a specific use case that shows value. Once value is demonstrated, projects can be scaled and just as important, if not more so, cultural change and acceptance of combining the art of retail with the science of predictive analytics can be realize.

The apparel industry has from its inception been a combination of creativity and data review.  Historically, the process of reviewing data was based on highly insightful and shrewd merchants using gut instinct and paper ledgers.   Today, spreadsheets have largely replaced pen and paper and enable retailers to analyze past data to do forward looking trend analysis. However, this process leads to reams of paper, a daunting amount of time for leadership to make sense of the reports only to be able to make reactive decisions about the business.

FIND AI Whitepapers, Case Studies and Retail Analytics Tools

Additional Complications

Further complicating the situation is that data is generally only readably accessible within each department— such as planning, finance, and marketing—with little coordination or information easily shared across teams.  Spreadsheets compound the issue as they do not lend themselves well to being shared even within the same teams. These limitations to data accessibility lead to poor planning, allocation and marketing decisions and limit the effectiveness and ability for advanced analytics. 

Getting started.

To take the first steps to deploying predictive data analytics, apparel players should review their business priorities and understand where analytics may be able to make the highest impact both to the bottom line and to the day-to-day workings of each team member.   Data analytics that support planning, buying and allocation generally deliver some of the highest ROI’s and most profitable impacts for apparel retailers.  

Specific goals and use cases should be selected.   Working across multiple teams that include members from the C-suite, finance, planning, buying, sales and operations, marketing and IT enable teams to create a prioritized list of business issues that can be empowered by data analytics.  These use cases can be enhanced by having a deep bench of data scientists and analysts in-house or by leveraging the expertise of a third-party partner. Whichever route is taken, teams should take care to ensure that the technical teams are well versed in AI and predictive analytics.  Traditional data analytics and business intelligence tools cannot provide the level of insights and move the needle to the degree being realized by modern predictive AI based approaches.  

The best use cases solve a specific business challenge while achieving a measurable upside both in terms of ROI and how team members spend their time. Therefore, instead of choosing a goal such as “improve planning,” a more specific goal of being able to “optimize all core products to the store and product attribute level” is recommended.  

Two areas should be considered when creating and prioritizing the list of use cases and goals.  In this example, the first issue to consider may be how to optimize the sales plan from size, color, and/or brand and what real-world constraints exist from the number of items in a pack and how much loose stock should be purchased.  The second item to consider is the impact on the business. How will streamlining and optimizing the planning for 80% of the business in core products be when you move spend to high-margin, high-demand products, how much time this will free up to enable teams to focus on the more creative 20% percent of fashion forward part of the business, how much will improved planning and allocation of core products minimize over and under buys and reduce the need for continual replenishment?

Use cases should first be prioritized based on the impact to the financials and team

Once sorted, the list should then be reviewed a second time with an eye towards how difficult a specific use case will be to implement both from an IT point of view as well as the cultural hurdles that may need to be overcome.  A project may be viewed to yield a huge payday, but may come with a correspondingly high amount of friction to implement while a project that may yield 70% of the impact can be quickly rolled out.  

It is recommended that projects then be ranked in such a way to deliver several short-term wins while larger, longer-term use cases that may make a bigger impact are being socialized and planned.  

A great place to start is with getting all data into one system that is online and viewable by multiple teams.  Modern cloud-based solutions including the FIND Predictive Retail Suite offer such an environment. FIND can take data from multiple sources within an apparel retail organization, clean it, and keep it up-to-date with very little effort on the part of retailers.  Once online, team members from all functional disciplines can view and collaborate on planning, buying, allocations, operations and marketing.  

FIND also offers Periscope.  A fast, effective and inexpensive way to have data analyzed for planning, buying, and allocation.  As part of the process, FIND’s data science teams will identify areas of opportunity and deliver quantifiable metrics to demonstrate the effectiveness of predictive AI for each specific merchant.