eCommerce, but With the Charm of a Shopkeeper’s Memory

What’s your favourite software? I was recently asked this question and it made me ponder. Despite working for a software company, I found myself slightly stumped as to what software I really enjoy. After a few minutes I decided that Spotify was my favourite piece of software and is something I use pretty much all day, every day (including right now).

Product discovery, your friend with the great record collection.

Aside from ease of use and low cost, it’s the great Radio feature in which songs are recommended based on a song or artist of my choice. In the old days, to discover new music I would have relied on a DJ whose judgement I trusted, or a friend who knew my taste. But now the same application that plays my music does this for me. Product discovery is crucial in almost every type of online business, and recommendations are an essential part of this, especially in instances where there is a large product catalogue.

Emarsys presented How to use Big Data to Implement Recommendation Campaigns at this year’s TFM&A exhibition.


Register your interest to visit TFM&A 2015 now.

When: 25 – 26 February, 2015
Where: Olympia, London

Spotify is a great example of how predictive analytics can use past customer behaviour to trigger further engagement. The aim is to highlight products that are very relevant, right now, to each and every customer. This enhances the shopping experience by:

  • Reducing the time needed to find a suitable product – reducing the customer struggle.
  • Showing more of the right stuff and less of the irrelevant, helping you to build brand affinity and that warm and fuzzy feeling that validates their choice of retailer.
  • Optimizing everything your marketing team ever does!

From an engagement perspective, product recommendations are proven to boost conversions, reduce cart abandonment, increase average order value, improve customer retention and create new purchase opportunities. That pretty much covers every angle in terms of eCommerce. And the beauty of it is that the algorithms behind the recommendations are constantly learning, so the more you use them, and across more channels, the greater chance they have of getting it right.

So, how can you implement recommendations?

If you have enough visits on your website, the right technology will crunch this data and match it to your product catalogue, seamlessly and in real time. Across your website and emails, you can then use intelligent recommendations such as ‘Recommended for you’, ‘People who bought this item, also bought…’, ‘Similar/related items’ and ‘May also interest you’. And whilst these are designed to sound simple…the reality is anything but.

Example of Spotify recommendations

So Spotify clearly have this technology, but they’re missing a trick. The app does a great job, but I still get regular emails from Spotify promoting music that just isn’t right for me.

Example of a badly targeted Spotify email

This email is titled ‘New Music for You’. But it isn’t new music for me; it’s new music. Considering the plethora of data that Spotify must have it seems like a huge opportunity lost, not using the intelligence to create a consistent experience both onsite and within email campaigns.

Automated recommendations are an intelligent way of managing a large product catalogue. Marketers can spend hours running intelligent segmentation models with Recency, Frequency, Monetary Value (RFM/eRFM) and so forth, but ultimately which products are you actually going to show those customers? Recommendations make that choice easy and allow marketers the option of combining intelligent product discovery with ‘Top sellers’ and ‘New products’ so they are able to keep customers engaged whilst still promoting the products they want to.

Update: while I was writing this blog, Spotify must have read my mind. Yesterday I received an early Christmas present:

Example of Metronomy being recommended on Spotify

Within a couple minutes of receiving the email I had added the track to my ‘new music’ playlist. Recommendations work!

I’d love to hear your thoughts on this. Share comments below or tweet @emarsysUK.

TFMA Insights recently hosted a webinar with Emarsys on The Science of Segmentation – to learn more about this topic you can access the archive of this webinar here.

Free Download: The 2015 Ecommerce Trend ReportWP_ecommerce

A review of 2014 and what to expect in 2015

TFM&A Insights reached out to some of the keynotes from eCommerce Expo 2014 to ask for their opinions regarding all that has happened in 2014 and what they are anticipating on the horizon. Questions include:

  • What were the major events of 2014 to impact the ecommerce industry?
  • What are the major challenges that the industry faces?
  • Do you have any advice for businesses facing these challenges?
  • What do you expect the major trends and developments of 2015 to be?

Download Now >>

Daniel Hagos

Daniel Hagos

Customer Solutions Manager, Emarsys UK

Having studied Business Computing and working as a Technology Consultant for a multichannel blue-chip, Daniel joined Emarsys London. After working in the UK Sales and the growing New Markets Sales teams, Daniel now works alongside the Emarsys Client Services team as Client Solutions Manager. As the Emarsys product set grows, Daniel’s role is to provide existing clients with advice on their goals and future direction so they can plan effective marketing strategies for growth, whilst utilising the full range of Emarsys solutions.

February 6, 2014

Warning: Invalid argument supplied for foreach() in /home/ubmemea/webapps/tfmainsights/wp-content/themes/enso-content/sidebars/post.php on line 44

Get the TFM Insights newsletter

The latest marketing technology Insights


Managing Digital Marketing in 2017 Report [Download]