More and more the concept of basket abandonment is mentioned as a method for increasing conversion - be it baskets in retail, quotes in insurance, bookings in travel or registrations in gambling.

Indeed, I am beginning to feel like it has been around forever. The real question for me is how to go from talking about it to actually producing the goods and enjoying the results.

Is anyone aware of organisations doing properly segmented basket abandonment at the moment (answer by return please!).

I’ve become bored of hearing what a good idea it is, and having personally been involved in doing it for the more forward thinking companies I decided it was time for an article that actually talked about what to do, rather than why.

Let’s start with the basics. Pick a conversion on your site that will drive the most business if it improves - quote not confirm, basket abandonment, register don’t deposit/purchase; there are many to choose from depending on the site in question.

That done (and yes it should be that easy) we are ready to build the lists and send out the emails daily to those matching rules.

Start by proving to your business it works and that doesn’t mean put it live and look at the results - it means build the list with 10% of the prospects removed to act as a control cell.

Compare these results for a period of time to show that those receiving the triggers are more likely to convert.

OK, so we have proved it works now let’s see what sort of creative message works best for the email.

The simple test (if you want more detail then contact me, this isn’t free consultancy you know!) would be to test between brand and offer, to find out which is most likely to convert.

Once you have your preferred message, start looking at email follow ups but remember people are different and some will be ready to purchase at different times. Follow ups to the initial email have accounted for over 20% of the new conversions in some campaigns. Test 3 or 4 days; 6 or 7 days; see which produces the best results.
Finally (for this example), start looking at the information you have on your site.

The web is a unique medium in that the prospect tells you what they are interested for you without you having to ask, so look at their visitor behaviour and split the creative message based on what they do on site, for example if you know they have looked at the delivery policy prior to dropping out, focus the message around a delivery offer.

So, without any complicated segmentation strategy meetings, big planning days or any of the other reasons that delay you actually getting a project started, we have gone from one good idea to 10 segments - along the way proving to your business the benefits and instigating a testing process, so you are ready to roll out to more areas.

Matthew Kelleher

Published 17 March, 2008 by Matthew Kelleher

Matthew Kelleher is commercial director as RedEye and a contributor to Econsultancy.

27 more posts from this author

You might be interested in

Comments (2)


Deri Jones, CEO at SciVisum Ltd

Nice practical start Matthew - but you've missed a big piece of the picture.

Before you can compare different campaigns (whether email or other) for basket drop-out: you first have to know whether the varying User Experience on your site caused the varying drop-outs.

By varying User Experience, I mean the speed of the pages from landing through to your chosen conversion point: how fast /slow they are, and what % of errors were users exposed to.

These metrics of user experience chanage in real time: as the traffic volumes increase all sites slow down to some extent: some start to throw sporadic errors at users ( errors that are not due to 100% outage, where all visitors hit a wall; but errors that impact 1 or 3 or 5% of users, statistically at random).

I had a client meeting before Xmas, where a 400,000 email campaign the night before had kicked off sales of a new product very well, they thought...when I showed the results of our 24/7 'mystery shopper' style monitoring of their key money making Journeys... they were gob-smacked: they'd exposed visitors to an error rate of over 10% for 3 solid hours, and there had been over 15 different types of errors that had been pushed at users.

Ranging from items added to the shopping basket that mysteriously were no longer there 2 steps into the name/address pages: through to gross techie error message gooblydegook in the middle of the page under the text ' you have the following items in your basket:'

Slow downs and sporadic errors are losing 3% of checkout journeys we find on average with our measuring projects - and in the majority of cases, the marketing folks and the tech folks are in the dark as to the scale of it, until the measurements come out.

I think too often, we forget just how much variation in User Experience there is in our sites: due to variation under traffic load, due to tech team tweaks and changes, and due to new feature roll-out.

If that isn't factored into your abandonment analysis, you're making big errors in your assumptions.


PS: Just today at a meeting where getting the best checkout ROI was the agenda, the client's tech representative didn't want to believe that 1% errors across the 24hr day, (3 % across the peak shopping hours) was something
that should be fixed.
'The site doesn't go down anymore' was as far as he'd got. 100% failures he understood, 3% was a stretch too far.
But by replaying the error pages from the past week, and proving that real users are seeing them, made it clear that sales were being lost.

over 10 years ago


Jeff Ince

Hello Matthew

The simple facts are that abandonment of shopping baskets or application forms etc will always happen as it is human nature. There are many tweaks that can be made to reduce the amount of abandonment, but this will never truly be eradicated as there are too many human interactions which drive abandonment.

That said, we have developed a solution which specifically addresses abandonment and our clients are achieving upto 50%+ additional, incremental sales from visitors who began the process and otherwise abandoned before completion. This is not analytical or statistics, but true actionable data.

We do not focus too deeply on why, more the when and put in place a solution to return results.

If you want to find out more or have clients looking to address abandonment, get in touch with me at and i will be happy to help.


almost 10 years ago

Save or Cancel

Enjoying this article?

Get more just like this, delivered to your inbox.

Keep up to date with the latest analysis, inspiration and learning from the Econsultancy blog with our free Digital Pulse newsletter. You will receive a hand-picked digest of the latest and greatest articles, as well as snippets of new market data, best practice guides and trends research.