Share this content

Is Barclays right to recommend cart abandonment as a way to shop online?

19th May 2015
Share this content

In the new Barclays television ads, members of the public can be seen offering snippets of practical advice to help people when they are online.

The most recent would suggest that shoppers should place items in their basket but not checkout and instead wait for a coupon to arrive. It is, in effect, advising cart abandonment as a way of shopping. If you haven’t seen it, here it is:

Etailers, marketers and ecommerce professionals have often debated whether recovery incentives actually work. So, we were very interested when a client came to us who had been doing cart recovery with another provider and including a 15% coupon in their recovery email. This produced a lot of recoveries and generated a massive 26.2% sales uplift!

Surely a 15% incentive must have been generating loads of extra sales?

With permission, we let the existing programme run for several weeks to get a baseline (the "15% Incentive" line), then we removed the coupon and continued for a further period (the "No incentive" line). The sales uplift immediately plummeted from 26.2% to 6.4%. Had we made a horrific mistake? Well, here's what happened:

The most obvious feature is that the recovery numbers were all much bigger originally

All of these had been boosted by the 15% incentive, so they fell when we removed it:

  • Sales Uplift:  Value Recovered / direct sales (i.e. Value Purchased - Value Recovered ) fell from 34.0% to 6.0%.
  • Recovered Rate: Recovered Value / (Value Carted) fell from 13.0% to 4.0%.

But the value purchased hardly changed – it increased from $79,620 to $80,184.

Customers were still buying at the same rate!

Clearly, when we stopped the incentive, a lot of people were buying through the normal checkout process, instead of getting diverted into abandonment recovery. This shows in two numbers:

  • Immediate Purchase Rate: Purchased Value / Carted Value, which rose by 10% when we stopped the incentive.
  • Percent Abandoned: Number of Carts Abandoned / Number Carted, which fell by about 9%.

Even better, our client had been allocating about $3,000 per month to the incentive, without increasing the value of sales at all.

What was happening? Here's our theory

  • Absolute price was never much of an issue with our client's shoppers, but nobody likes to pay more than they have to, so when our client put a 15% coupon on the cart abandonment email they changed the behaviour of a lot of people. Shoppers who would have bought immediately were diverted (cannibalised in the marketing jargon) into becoming delayed buyers.
  • Instead of just putting products in the cart and buying, these shoppers would leave, wait until they got a cart abandonment email, then return to the site and copy the coupon from the email. A much more complicated process!
  • It's well known that simplifying the checkout process improves sales (here are some great tips), so complicating it will reduce sales.
  • Basically, some of the would-be savers tried to save 15%, found it was all too much trouble, and dropped out. Maybe the recovery email went to the spam folder, or maybe the coupons expired before they got around to using them and they left in a huff. The precise reason doesn't really matter - what does is that the added complexity seems to have significantly reduced sales, cancelling out any beneficial effect from the offer.


Be careful with incentives. Whenever you introduce a powerful reason to buy in a particular way, some of your clients will switch to it. But if your new way is complicated, some of those clients will drop out, and you will actually lose some sales.

Mike Austin is CEO of Fresh Relevance.

Replies (2)

Please login or register to join the discussion.

By nick-o-lied
19th May 2015 11:59

Could you please review and perhaps QA the figures, formats and the formulae on the table and try explaining the results again.

It's an interesting topic but for a forum for data minded people the presented results reek of illiteracy.



Nick o'Lie


Thanks (0)
By Mike Austin
19th May 2015 16:27

Hi Nick,

Thanks for the comment. Interesting role-reversal for me, because I'm usually the one critiquing online maths. I'll see if I can clarify.

Apologies for the % characters in the Value Purchased column: these were typos. 

I've recreated the underlying numbers (the original data was in our UI and is sadly long gone, because we have to comply with the data protection act and delete data when it's no longer needed)...

Carted = $178,000; Immediately Purchased=$60.000; Abandoned=$118,000; Recovered=$20,000; Total Purchase=$80,000

Carted = $173,000; Immediately Purchased=$76.000; Abandoned=$97,000; Recovered=$4,000; Total Purchase=$80,000

As you can see, the Abandoned figure dropped from $118k to $97k when the incentive coupon was no longer used. The Recovered figure also dropped but overall our client made more money (especially because their profit margin rose due to selling at a higher average price). Presumably this was because shoppers were not being incentivized to abandon.

Thanks (0)