Skip to Content

Affinity Analysis: What’s up with the drag-along effect?

Part 2 of a 4-part series

In my last blog, I introduced the idea of affinity analysis. Now I’m going to talk about the drag-along effect.

To refresh – we’re using a simple definition of affinity analysis based on the retail experience where companies try to gain insight into what products customers buy together.

If you really want to dig into the weeds of affinity analysis as more general topic, Wikipedia is a good place to start. But for this space, let’s start by diving into one of the more concrete areas of affinity analysis – namely the drag along effect. (Again, thanks to Shantanu Goswami at SAP for the input here. I’m sharing what he explained to me.)

It’s often beneficial to see purchasing behavior in terms of primary products and associated products. When a man buys a suit (primary product) he may also buy a tie (associated product). In a nutshell, the tie here describes the drag along effect. Sometimes it’s hard to see what’s dragging what – as when customers buy socks and underwear. But you get the point.

OK, so let’s say that you detect a drag along affect for batteries associated with high-end cameras you may sell. This correlation is obvious enough to detect – with or without sophisticated business intelligence tools. But do you know which brand of batteries people are buying? Is there a correlation with factors such as time of day or the age or gender of the customer? Do some stores sell more batteries than others? If so, why?

 

Insights like these take analysis and in-memory technology to crunch massive amounts of data. My expert, Shantanu, will explain in a bit more detail. 

 

What other examples of the drag-along effect are out there? Are you getting this insight? How are you getting there?

Get started with the Discovery Service for Affinity Insight v2.0.

Learn more about affinity insight for retail.

Blog 1 – Affinity Analysis and the “man aisle”

To report this post you need to login first.

5 Comments

You must be Logged on to comment or reply to a post.

  1. Shantanu Goswami

    Retailers have been tracking ‘trends’ for ages. They base their promotions and merchandizing decisions on them. What has changed with this innovation are essentially 3 things: Firstly the retailer can now analyse their data realtime and at a transaction level. Not aggregate data, not patterns, not trends. They could not do this before due to the sheer size of the data and the speed of its creation. SAP HANA enables this paradigm shift. What else is new, is the statistical functionality built into this tool which allows the Retailer to get a realtime sense of how their market baskets are performing and judge the effectiveness of their promotions. Lastly, but not the least, this tool is a starter pack which lays open a new, secure roadmap into realtime behaviour analysis. Enabling businesses to run better, serve better and be more profitable.

    (0) 
  2. Thomas Hepler

    I think there is a lot to this, but there are others who have been mining social data to create affinity graphs like Colligent.com based on consumer likes within social data. This seems like a much more natural expression of “affinity”.  Just guessing, but SAP using purchase data is quite a bit different than following emotional insight.

    (0) 
    1. Gina Keeler Post author

      Hi Thomas, thanks for the comment. I’ve invited one of my colleagues/experts to provide some fruther insight. Stay tuned…

      (0) 
        1. Gina Keeler Post author

          Hi Thomas – We have a demo video that shows some of the functionality of the solution. You can access that here. I’m also still tracking down an expert who can comment as well 🙂

          (0) 

Leave a Reply