When the subject of conversion optimisation comes up, most of the talk is about on-page optimisation – utilising A/B or multi-variant tests to determine which headlines, calls to action, images or page layout persuades your visitor to take the necessary steps to reach their goal (and yours).
It’s not the only way, mind you. There are ways to increase your conversions using channels outside of your website.
Let me paint you a picture.
Your company sells car insurance and I’m shopping around for a new policy. I dread this point every year and in fact it’s often the reason why I just renew my policy even though I could get a better deal elsewhere. But this year is different. I’m going to find the best deal and it just so happens I’ve found it on your website.
I’m pretty pleased, that didn’t take long. So I start to fill out your online application form. The smile on my face slowly begins to fade. Within five minutes it’s a full on scowl. So many questions! You’re asking for dates for this, policy numbers for that and… oh my word, is this really worth it to save £40???
Thing is, I know that every insurance website will ask me this. If I want the saving, I’m just going to have to go dig out the old paperwork so I can answer the questions. But I don’t have time right now, I’ll do it tomorrow…
Or not, as so often is the case.
It’s not just insurance, it happens everywhere on the web where some kind of transaction is occurring. It’s typically referred to as an abandoned shopping cart, but it equally applies to non-ecommerce processes, such as lead generation forms, job applications and account registrations, for instance.
As optimisers, this pains us. It’s our job to convert as many visitors to customers as possible. We could just accept these as lost or we could figure out how to get them back.
Bringing them back
Remarketing, as the name suggests, is the activity of re-engaging with visitors who did not convert first time round. The most popular form of remarketing is via display advertising. A cookie is set when a user visits a site and if they do not convert then personalised display ads are targeted at them on their subsequent travels around the web. Major players of this technology include Struq, Criteo and Google.
That kind of remarketing can be very effective, but I want to talk to you about another type of remarketing that can do wonders for your conversion rate. Email remarketing.
Urgh, email marketing. Boring. With all the sexy new technology available, people would rather not talk about email. But you know what? It’s still one of the most effective marketing channels you can use.
Why? Because there’s infinitely less competition for attention within the private walls of their inbox than there is out there on the wide open web. Especially if the message is one to one, relevant and of value. For instance…
It looks like you weren’t able to complete your insurance application today, so we’ve saved the information you entered. When you’re ready, just click the link below to pick up where you left off and you’ll be insured in just a few minutes.
How I recovered 52% of abandoned shopping carts
Lets look at a real example. In a previous role I worked for Jobsite.co.uk, a leading job board in the UK. Advertising recruiters have a choice of buying job ad credits over the phone from a salesperson or they can do self-service via an ecommerce system, complete with shopping basket.
Tasked with increasing revenue, I analysed the performance data to identify optimisation opportunities. I spotted one in the basket abandonment data. These people were interested enough to place a product in the shopping basket and start to fill out their billing details… but for some reason just couldn’t bring themselves to hit the ‘Buy’ button. Whilst there were plenty of on-page optimisations to do, I felt these people were so very close to a conversion that we had to try something a little different.
So what did we do?
We decided to utilise the Conversion Manager product from SeeWhy, which provides real time personalised messages to visitors that abandon shopping carts. With the software integrated with our system we were able to send triggered emails based on the behaviour of the visitor and the information they provided on the site.
Each visitor could receive up to 3 emails depending on their actions, or inaction, all written in a non-sales tone and featuring a link to their shopping basket, containing their partially completed data.
The result? 52% of those visitors who received a remarketing email from us returned to the shopping basket and completed their purchase.
So, a little more about the emails…
The first email needs to be sent immediately. There are 2 reasons – 1) not all abandonments are deliberate, sometimes people get kicked out of systems or lose internet connectivity and 2) the visitor is in a buying frame of mind if they were in your checkout, so now is a great time to reinforce your message and offering.
If the customer does not purchase (or register, or download, etc.) within 23 hours then send the 2nd email. People are creatures of habit and your visitors are likely to be back online the same time next day (i.e. lunch break). Why 23 hours and not 24? Well, that allows for any delay in the ISP delivering the email and for it to be ready waiting for them when they next login in.
If they’ve still not purchased after 6 days and 23 hours send them the 3rd and final email. Any more than this and it starts getting spammy. You might want to trial a discount voucher code in this final email (we did), as one last attempt to get them to complete their order.
Be careful of offering the discount any earlier than this. If your promotion occurs in the first email, you’ll train your visitors to abandon the shopping cart immediately, just to get the discount email. It’s a very patient customer who will wait a week to get their third email just to game the system.
The tone of the emails is important. A friendly customer service tone is what you’re after. An overly aggressive sales message will come across as spam – and remember, some of the abandonments may be because of genuine difficulties completing the process.
Make sure you capture the contents of a field as the data is entered. This makes it possible to return the user to the point they left off when they click the link in the email. This is a big win for the visitor – tell them in the email copy that they can pick right up where they left off. This’ll increase the likelihood of them finishing the transaction.
One of the advantages of using a product like SeeWhy is that it synchronises with your purchase data in realtime. This ensures that you don’t send out a remarketing email if the visitor has recently returned to the site and made a purchase. Failure to stop the scheduled email would result in a confusing experience for the customer – and a frustrating one if you’ve sent them a discount code moments after they purchased the product at full price.
So how did each email perform?
|Timing||Shared of Recovered Baskets|
|Email 2||23 hours later||16%|
|Email 3||6 days 23 hours later||6%|
As you can see, the vast majority of recovered shopping carts came from the first email. If need be, you could just implement a single email if you have limited resources and you’ll still likely experience an uplift. However, you may feel that it’s worth the additional work to recover the additional 22% from emails 2 and 3. It’s also worth noting that you can experiment with the timing of the emails. The ideal phasing of the three emails could be different for your product and audience.
So, if you’re looking to optimise your shopping basket (or your lead generation form, your sign-up process, or download page), think about all the channels at your disposal. To encourage visitors through the buying funnel, you could A/B or multi-variant test the content of your web pages. To pull browsing visitors who did not convert back to the site try remarketing via display adverting. But to get those people who were so close to buying that they almost completed your shopping basket process, give email remarketing a go. Guessing a 52% recovery would look pretty nice on your stats sheet?