What are the most important products on a retail website?
19/11/2010 | Written by | Categories: Analysis and Insight

Retailers have used various methods over the years of identifying the most important products in their range.  This varies from most units sold, most revenue generated, (getting smarter) most profit generated, highest ROI, (smarter still) products that generate high basket values (through incremental customer purchases), etc.  The information is used for merchandising, designing store layouts and other marketing efforts around products.

In the online world, all of these methods still apply.  But as with everything else, there is more information available in the online world and so retailers can get even smarter with how they merchandise their products.  Beyond looking at what products customers actually purchased, online retailers can collect and use information on what products visitors were interested in based on what they were looking at.

The most popular products are not necessarily the ones that people purchased the most of, they could instead be the products that had the most visits.  But then, if these were the most popular products, which ones are the best performing?  Which products do visitors need an extra push on before they commit to the purchase?  Which products could you get incremental sales out of through making some sort of change?

A merchandising report that provides an online retailer with all the information they require may end up looking something like this (click on report for a larger image).

The key metric/column for me is the Product Page Success Rate.  This is ideally defined as the number of visits in which that product was added to basket divided by the number of visits in which that product was viewed.  It refers to a product page as products may have more than one page on a website, in different departments for example.

So what to look out for within this merchandising report?  Key products to identify are both popular products (high visits) with a low Product Page Success Rate and less popular products with a high Product Page Success Rate.  I generally filter on visits to only include products with more than X visits and then sort on the Product Page Success Rate to easily identify these products.  A few examples are highlighted:

  • A (product 1) – the most popular product based on the number of visits that viewed that product and also the number of units sold
  • B (product 5) – the most valuable product based on revenue generated, even if not the most popular or with that many units sold
  • C (product 6) – this is a fairly popular product but it is not being added to basket, some suggestions are provided below as to why and what can be done to improve this
  • D (product 7) – for some reason, people just weren’t convinced on this product – they were viewing it, adding to basket but only a low proportion ended up buying it
  • E (product 9) – this was viewed in less than half the number of visits as product 1 – but people loved it, it was added to basket nearly half the time

Popular products that aren’t being added to basket need to be reviewed urgently.  There could be an easy/quick change you can make which leads to incremental sales for that product.  The issue could be any of a variety of things:

  • Availability in certain sizes/colours – can you get more stock from stores/warehouses?
  • Price – should you be reassessing the price on this product as visitors feel it is not good value for money?
  • Image/video – does the product look unappealing visually (particularly important for fashion items) and can you reshoot the photo/video to improve this situation?
  • Product description – have you sold the product well enough?

The flip-side of this are products that not many people are finding but they commonly add to basket when they do view it (product 9 in the example above).  It might be they only appeal to a small niche market but for some of these products, if they were exposed to a wider market via promotion on the site or a paid search ad, their sales could be significantly boosted.

The merchandising report ideally would include the department, category and brand of the products.  Through roll-up reports, you can then identify the most popular and best converting categories and brands or identify issues with your merchandise, e.g. a brand that is popular but overpriced.  This moves the marketing and merchandising efforts from product level to a larger scale with a potentially bigger impact on your business bottom line.

To answer one last question, how often should this sort of report be looked at?  As changes can have an immediate impact on performance, would it hurt to have each member of a merchandising team look at five to ten key products each morning based on the data from the previous day?

Let me know any questions regarding this suggested report or make alternative suggestions for what people may find useful.  If you think a report like this may be just what you need to boost sales so that you hit targets for this Christmas, please do get in contact.

4 responses to “What are the most important products on a retail website?”

  1. […] This post was mentioned on Twitter by Tim Leighton-Boyce and Peter O'Neill, Kelly McClean. Kelly McClean said: RT @peter_oneill: The key products on a retail website aren't always the most viewed/purchased – http://bit.ly/adrL1I #measure […]

  2. I know it sounds cliche, but this was a very thought-provoking post. I had always thought that high margin, high volume products were the ideal products, the ones in the top right of the quads.

    Now you have opened my eyes to a higher level of understanding. While those are good products to watch for (and the anti-thesis, the ‘dogs’ with low-volume and low-profit), there’s also the consideration of which ones contribute to a high margin basket. I guess a good analogy would be the steakhouse product mix: Where you don’t make much on the steak, but your high margins come from the extras.

    It reminds me of a feature we have in our product that is not well publicized. The operation is called setmath on the analyze tab, if you can find it. If not, please contact me :). It can analyze two sets (of products in this case). It can tell which ones were in set 1 but not set 2, vice versa, and both. Another way to think of this is joiners and leavers. This kind of thing helps you to do basket analysis on an analytics level. For example, envision a report which has a list of products, and the columns are period 1 and period 2. You want to derive three reports out of that. Which products were high in threshold in period 1 (and only period 1), period 2 (and only period 2), and both periods. The first are the leavers, the second are the joiners, and the third are in both periods. If you compared products that were in these lists over a series of periods, you would find out which ones were your stars period over period.

    I think you already mentioned our product in a Twitter – that’s how I realized your blog exists. So I hope this isn’t considered to be a spammy comment. I just thought that since you already mentioned our product in your tweet, you might want to know how to do what I described. I’d be happy to walk you through an example perhaps in a blog article.

    • Peter O'Neill says:

      @Ward Thanks for the feedback and the kind words, I guess what anyone writing is trying to achieve is to make others think.

      I tried out various data extract tools with GA over a year ago now and at the time, I didn’t feel that any was at the standard I was after. It appears that it is time I relooked at this and tested out the leading tools to see how they go with creating dashboards or exactly this kind of report. I will let you know once/if I find a company with good data to work with on this project, so that you can give your feedback on the results.

Leave a Reply

Your email address will not be published. Required fields are marked *

Looking to improve your digital analytics or conversion rate optimisation?
Contact us today on +44 (0)20 7148 5970

Alternatively, come visit us in London and we’ll buy the coffee!
Get in touch
Leave us a message
Use the form below to leave us a message.

Lindsey Street