A Shopaholic’s Guide to Analytics II.B


The Rule of Three is a writing principle that suggests that things that come in threes are inherently funnier, more satisfying, or more effective than other numbers of things – Wikipedia.

3 Ps of success, Blind Mice, Little Pigs, Stooges, Musketeers, The Matrix, The Lord of the Rings, rings, pairs of shoes, 3 year memberships… Everything is better in 3s – including this shopaholic series!

  1. A Shopaholic’s Guide to Analytics
  2. A Shopaholic’s Guide to Analytics II.A: Half a shopping bag of useful techniques in Analytics
  3. In this last hoorah on the topic of Retail (therapy) Analytics we’ll empty our bag of some the most useful analytics techniques for keeping our customers happy and loyal. From the customer’s perspective, these are “A feeling that I got a good deal” and “Convenience”.

Today I am referring to any entity that transacts as a customer – individual, household, business etc. and any item for sale or service as a product.

A feeling that I got a good deal – what offers, when and how often?

Whether it’s the word SALE or finding a rare collectable, we all want to feel that we had a fair transaction. But as a retailer or service provider, we need a balance between being competitive and fair to our customers and staying in business.

What is the impact of price on demand?
How long should a promotion run before it’s unprofitable?

shopaholics-guide-to-analytics-1Price does not affect demand of all products the same way – electricity versus floor cushions, burgers versus Porsches. The economics 101 method to understand this impact is price elasticity / sensitivity – the ratio of the percentage change in demand over the percentage change in price. “Elastic” products (ratio greater than 1) are sensitive to price changes.

However, price and demand change over time, sometimes seasonally but not always consistently – affected by economic and other environmental factors. In this case, time series forecasting “causal models” (described in “The right product”, II.A) can be used to model the relationship between price and demand directly. From this model, price elasticity can be calculated, or what-if scenarios can be run to measure the direct impact of price, taking other factors into account, at points in the future.

These techniques can also be used to quantify expected impacts of promotional activity, length and frequency and avoid over promotion.

Which creative is more appealing to customers?
Which product offer is more profitable?

Predictive models and optimisation techniques (described in “Good service”, II.A) can be used to best allocate competing offers or where similar offers have been given in the past. If there is no history, we need to test the effectiveness of our offers through experiments on small samples of customers and extrapolate these to what is likely in real-life. This is known as a choice experiment. To derive statistically viable decisions, we use experimental design to make sure we are capturing sufficient information across the different choices. A simplistic form of this is an A/B test.

Convenience – what is relevant and where?

If shopping was a sport, then as an elite athlete, I expect towels to be stocked in the locker room and the showers to be functioning. Basically, there’s enough going on in our lives – and often too many other competitive options – for customers to deal with difficult or restrictive processes.

Yes, I realise I sound like a brat. But as the e-tailer market grows, people continue to work longer hours and globalisation is a reality, it is even more important for retailers and service providers to make transacting easy. There are operational considerations – integrated systems, web design, accessibility, etc. – but there is also the need for detailed profiling to understand the viability of the target market.

What products are the most relevant?
What is the best store layout and window dressing?
What are the most effective channels?

Demographic – a profile of the different types or segments of customers and how they are likely to behave under various circumstances e.g. during lunch breaks, with young children, in retirement, etc. Using statistical segmentation techniques such as clustering or self-organising maps are useful for creating segments but profiling is the process of differentiating these segments and is done through slicing and dicing and visual exploration[1].shopaholics-guide-to-analytics-2

Where should we build the next store?
Where should we locate the distribution centre?

Geospatial and location – a profile of the geography and terrain overlayed with hotspots of activity e.g. industrial, commercial, residential, thoroughfares etc. and, to optimise decision making, demographics and economics. Geospatial visualisations and network maps are helpful to highlight and differentiate between these areas of interest.

BUT convenience is underpinned by how well we understand our customers’ needs.

Do we have enough of the right products?
Are we proving exceptional customer service?
Is there sufficient value and choice?

I hope that you have picked up a pair or two of comfy shoes to help you on your analytics journey. If you ever feel lost in the sale crowds, as with the sport of shopping, focus on one thing at a time – an “aha” moment you can make a reality. Set your well-articulated goals and invest in the right-fitting solution of people, process and technology for the relationship you want to have with your customers.

Learn more about how you can quickly get started with exploring your data in the cloud with this on-demand video, Insights in Seconds.

Happy shopping!

[1] SAS Enterprise Miner has the out-of-the-box ability to profile segments statistically using comparative graphs.


About Author

Business Solutions Manager

Annelies believes that there is potential for Analytics everywhere. She works as an evangelist, enabler and execution strategist to empower individuals and organizations with good Analytics practices. Annelies is the Advanced Analytics technology lead at SAS Australia and New Zealand responsible for product management and enablement. During her career, Annelies has held various positions supporting the customer lifecycle from strategy and requirements to implementation and adoption. This experience gives her a practical view of the end-to-end process of data analysis across government and industry, including engagements in several customer analytics, demand forecasting, text analysis and allocations optimization projects. Annelies is the current co-chair of the NSW Chapter of the Institute of Analytics Professionals of Australia, the largest analytics community in Australia. She has a research Masters in Mathematical Statistics and guest lectures at several tertiary institutes in Australia.

Comments are closed.

Back to Top