Striking Oil with Retail Data

Anyone ‘of a certain age’ will remember those old 5¼ inch floppy disks that could hold 1.2 Mb of data; barely enough for a minute of music in today’s terms.

I remember needing to swap in and out multiple disks to play games on my Commodore64, which had just 64kg of memory and zero onboard storage.

It’s a crazy contrast to my current mobile with 8gb of RAM that could swallow that C64 125,000 times over, with ridiculous data storage to boot.

Now that we can; data is collected everywhere and made available to us in various ways to use as we will, be that for general interest, personal development or to help our businesses survive and grow.

Data is increasingly generated in bigger quantities and at a higher pace; indeed, of all data available to us today, 90% was generated in the last 2 years[1]!

So we have access to massive amounts of better data, with better storage options and tools to manipulate and make that data work for us in intelligent ways. What do we do with it?

Clive Humby coined the phrase ‘data is the new oil’ around 15 years ago and while that’s typically understood to indicate the value of data, it also means that to extract value from data, it must be mined and refined and put to work intelligently.

 Key metrics for retail

In the retail environment there are a number of important metric indicators that are used to understand viability and success, as well as enable planning.

At a very basic level, all retail organisations should be collecting data via their POS tool, which provides raw data on items purchased, the value of each transaction and when each transaction occurred.

This data can easily be ‘refined’ into some of the basic key metrics:

  • Transactions per day/hour
  • Average transaction value/basket size
  • Total sales per day
  • Sales per square foot

Given the availability of other types of data, such as item and labour costs, we can establish cost of sale, gross margins and profitability down to department, product line and item level.

This data is clearly of value to retail for reporting and planning; however we add additional value by incorporating large amounts of historical data and using that to detect trends and outlying data.

This enables us to understand our business and capitalise on what we learn.

For example, if we were to take multiple years of POS data, looking at 15 minute intervals across day types (e,g. standard days of the week, holidays, sale days) we can easily detect trends across each day to see:

  • Which are my quiet and busy days?
  • When are my quiet and busy timesthroughout the day?
  • How far after a marketing campaign did we see the impact?
  • Which marketing campaigns were most effective contributed the most value?

We can also easily use this data to plan better:  improved schedules, more accurate budgets/targets, better use of time/task allocations and uplift traffic and revenue on quiet days.

The real oil is traffic data

Traffic data gives a real edge over standard POS data.

While collecting this data has traditionally been cost prohibitive and difficult to implement for retail operators, that has changed dramatically in recent times with better, smaller and more affordable cameras and processors, alongside solutions for processing camera images and making data available.

Key metrics from traffic data include conversion rate and sales per shopper, which add a whole new level of value to retail operations.

The conversion rate is a view of how many people were in store, versus the number of transactions over a period of time.

It’s a simple calculation and when broken down across the day, in 15 minute intervals, we can clearly see when customers coming in are making purchases, versus walking out empty handed.

What does this actually indicate though?  

Did you know that conversion rate is directly related to likelihood of customers returning to a store?

A study into conversion in retail[2] found that lower conversion rates result in lower rates of return, and vice versa for higher conversion.

So, if retailers want customers to return, understanding conversion rates is a great place to start and will allow a clear view on where and why issues are occurring that could be addressed to bump conversions in the right direction, and increase customer loyalty.

Use trend data to identify your optimal staff-customer ratio and increase conversion rates.

The question of when and why the conversion rate changes over a day, or different day types, is critical to addressing issues.  A time interval where we see increased foot traffic, but reduced conversion could indicate a customer service problem.

Using your traffic data to help plan schedules and tasks could significantly improve both customer service and conversion rates.

The reverse of this is of course understanding times of overstaffing to reduce or redistribute resourcing costs at those times.

Additional questions we can ask with traffic data might be:

  • When comparing stores or departments, are there differences in conversion?
  • How can one part of the business learn from another part of the business and have an overall impact on company customer service and conversion?
  • How does changing window displays impact traffic?
  • What is the impact of certain marketing activities on transactions vs traffic?
  • How does the weather impact traffic?

Tip of the iceberg of data analysis 

There’s a lot more we can analyse and do with data, including:

  • combining instore data with online and instore data to compare online traffic/conversion/sales
  • performing weighted calculations across multiple datapoints for optimal resource planning
  • ensuring instore activities are better planned, though monitoring of change rooms, choke points and register waiting times

Given our current situation, taking advantage of the data and tools that are readily available to us seems like a no brainer.

If we can refine that oil it’s possible to give retail operations a significant edge, all through really understanding our customers through data.

Using data, we can improve the customer experience and as well as the experience for our front-line workers; keeping customers happy and coming back, while ensuring staff are not overworked and are utilised effectively across every day.

There are lots of quick wins available to retail by ensuring access to accurate, quality data, and being able to refine it in ways that work for your operations.

My advice to retailers is to start tracking that traffic data. The value is there and accessible right now.

Are you ready to start mining the new oil for your business?

Malcolm Breen

By: Malcolm Breen

Solutions Consultant

1] “How Much Data Do We Create Every Day? The Mind-Blowing Stats Everyone Should Read” by Bernard Marr, May 21, 2018

[2] Perdikaki, Olga & Kesavan, Saravanan & Swaminathan, Jayashankar. (2011). Effect of Traffic on Sales and Conversion Rates of Retail Stores. Manuf. Serv. Oper. Manag.. 14. 10.1287/msom.1110.0356