Tuesday, 4 May 2010

Timing the trigger

When is a trigger email not a trigger? When you can’t remember why it was sent. Take this example from John Lewis: Last week I was shopping for some sunglasses. I fancied a pair spotted on one site and did a Google product search to compare prices and get more details. John Lewis had the pair in stock as well as information on UV protection, zoomable images, and in case they weren’t the right pair for me, some suggestions on what else I might like in different price points.

For one reason or another, I decided not to buy the sunglasses. I left JohnLewis.com and thought no more about it. That happened at the weekend. On Wednesday, I received an email from John Lewis titled: “Thank you for your interest. Free standard delivery on all orders over £30”. This left me puzzled. What interest? Is John Lewis referring to the sunglasses I looked at three days prior, or have I inadvertently visited the site and triggered some other kind of action? And why did it wait until Wednesday to send me this email?

In my opinion, a trigger emails should be almost instant—Amy Africa recommends waiting no more than two hours after the consumer has carried out the action (abandoned basket, or abandoned search, for instance). Three days later and it’s definitely too late. So whilst I appreciate that John Lewis wants my custom enough to bother sending me such an email, its trigger marketing ends up shooting itself in the foot.--MT


  1. Hi Miri

    An interesting and thought provoking post. The timing of when to send a trigger campaign is one of the easiest things to test so choosing the optimium time for your brand is something that should be based on hard science.

    Cleverly, John Lewis do not use offers to try and convert their abandoned baskets. This gives them the flexibility to test the timing from an hour through to days after without running the risk of any sales cannibalisation.

    At Velo//, we always implement a testing strategy when implementing abandon basket campaigns. There are three major things to test;


    The messaging and timing ideally should be different based on the type of product. High involvement products like furniture are not impulse buys and many people will add them to their basket as part of making their decision before perhaps visiting a store to see the item for themselves before returning online and placing an order.


    Varying the timing from hours to days getting a strong steer from your analytics.


    Varying the tone, the sender (customer services not marketing), length and more will change the effectiveness.

    There is no such thing as a industry standard right time to send a triggered campaign. The answer lies in a testing strategy using AB splits to find the perfect time for your brand. After all, with considerable conversion rates of typically at least 30% (our own experience), then this is a very good use of every online marketer's time. We've managed this process for a couple of clients, and always seen significant returns in very quick turnaround times.

    Keep writing the email marketing articles - I really enjoy reading them!


    Velo// - technology led marketing for retailers.

    Some case studies on this area:

    The abandon basket behavioural email campaign resulted in a 678% return on investment driven by a 70% increase in open rates, 36% click through rate and a 10% click to order conversion.

    Eddie Bauer
    Their remarketing emails outperformed their traditional email marketing by 21x (revenue per email) and 20x (transaction rates).

    Indium Web Management
    50% open rate of which 50% clicked through to the site and 53% of those who clicked through, went on to convert.

  2. Whilst your John Lewis experience has been less than outstanding, an immediate response to an abandoned cart is not for everyone; in fact I would suggest that it’s suitable for a very small minority of retailers.

    Here’s the crux, customers like to browse, add things to baskets, drop them off again, shop around, go off and make a cup of tea etc, etc, etc. Sometimes they might just want to think about their purchase, by sending them an immediate offer there is a good chance that all you’re achieving is eroding your margin.

    As ever what is right for one retailer is not necessarily right for another, and before you implement an abandoned basket program look at your history. For one retailer selling gifts it was taking on average 2 days for the customer to go from adding to basket to checking out, by using an immediate abandoned basket offer of 10% off they were giving away £££’s

    By sending an abandoned basket email 2 days later you can confuse the customer, because (and here is the tough love) shopping on your site is not the biggest event of your customers week, like Miri they will probably have forgotten they were even on your site, and all you do is confuse them. So don’t say thanks for your interest. Say something relevant.

    What JLP should have said is “Especially for you, get free delivery on all Sunglasses over £30” – this would have been far more targeted and allow Miri to choose some sunglasses she liked. No mention of a recent visit alluding to scary big brother is watching marketing, just a timely and relevant offer.