Why are sites that should know better still making the same old mistakes with their checkout process?

Why are big-name pure play sites like Amazon, Very and ASOS losing up to £420 million a month in lost sales?

According to a recent report by Moneybookers, some popular online stores have painfully slow checkouts. Reportedly, some of the worst offenders were “pure play” sites such as Very.co.uk (with a checkout time of 6 minutes 45 seconds), Amazon (5 minutes, 38 seconds) and ASOS (4 minutes, 36 seconds). By Moneybookers’ calculation, this potentially translates into £420 million a month in lost sales.

We know people will abandon an online purchase very easily (for example, around a quarter of potential customers will abandon their purchase if asked to register at the start of the checkout), so why are sites that should know better still making these mistakes?

As an aside, it is worth noting that the Moneybookers assessment was made from the perspective of new customers. Existing customers are unlikely to encounter such large checkout delays as they will use saved data (for addresses or payment cards) or features such as 1-click ordering on Amazon, thus streamlining their checkout journey. However, the lengthy checkout process is still a big issue if it is potentially driving away new customers. Attracting new customers, who can then be converted into loyal repeat customers, should be the number one goal of any online store.

Why is the checkout process so time consuming?

The main issue contributing to the length of time taken to complete the checkout process appears to be the sheer number of steps (allegedly 11 steps for Very.co.uk, but I had to give up after the site wouldn’t recognise my postcode for the address look-up and wouldn’t let me manually enter my address instead, but that’s another story). For some of the sites this is compounded by the lengthy page-load time between each step (note that Amazon have reportedly found that page load time increases of as little as 100ms result in a 1% drop in sales).

What does an optimized checkout process look like?

In our opinion, it should be possible to streamline the checkout process to take just a couple of minutes. The minimal set of data that needs to be collected is the customer’s name, their billing address, their email address (for sending order confirmation) and their payment details (card or bank details). Customers also need options for agreeing to T&Cs, for entering promotional codes or vouchers (if appropriate for the site) and for opting in or out of follow-up marketing activities.

To ensure the fastest possible checkout, the delivery address should default to the same as the billing address, with an option to input a new address if needed. Options for selecting the delivery method and for making gift purchases also need to be provided but if these default sensibly to the most commonly selected option for the site, they should add minimal time to the checkout process.

Entering this minimal set of purchase details should then be followed by an order summary and then an order confirmation page. This should not take the customer nearly seven minutes to complete. Similarly, using a facility such as Google checkout means that all the customer has to do is to enter their username and password and then select a couple of options (checkout.google.co.uk), taking less than a minute.

How do you balance the trade-off between conversion and quality of data?

Of course, we do understand that companies want to collect additional information from their customers during checkout, not just the bare minimum of name, address and bank details. This additional marketing data can be highly valuable to a company in several ways – for example, by selling the lists on or by generating advertising revenue. The assumption seems to be that customers will be so determined to make their purchase that they will answer any number of intrusive questions that you care to ask during checkout. As we’ve seen, that just is not the case – in the Moneybookers study, 74% of the 250 experienced online shoppers who they consulted were concerned at the amount of personal information collected during checkout.

Any attempts to collect information that is not seen as necessary to the purchase will meet with resistance from customers and potentially lead to abandonment of the checkout. If you’re buying in a store, it’s anonymous – people naturally expect the same privacy online. There is definitely a conversion trade-off to be managed. Our advice would be to keep the data collection to the minimum during checkout (to ensure that you make the sale!) and then to collect the marketing data afterwards (when you’re hopefully talking to a satisfied customer). A follow-up email survey, accompanied by a promotional offer for completing it would be one idea.

In conclusion

At Nomensa, we design retail sites with the conversion trade-off in mind from the outset. (It is also worth noting that these same challenges apply to any multi-step online transaction, such as registering for a university open day or completing an application form for a charity fundraising event). We recommend using A/B or multi-variant testing to measure the impacts on conversion of collecting additional data. If you calculate the monetary value of each data item, it is then possible to work out the optimal trade-off.

Our over-riding belief is that it is always better to attract new customers who can become long-term repeat customers with a high lifetime value. Collection of marketing data is then something that can be done at a later date – at this point it can be seen as of real value to the customer in terms of helping them to get an even better, more tailored service.

Share this post

About the Author

Juliet is a Principal UX Consultant at Nomensa, where she is responsible for implementing user experience strategy and services with a focus on the not-for-profit sector. When she's not considering the many facets of user research, Juliet can often be found explaining why cheese doesn't talk, and other life mysteries, to her small daughter.

Add a comment

Fields marked with an asterisk (*) are mandatory.

Comment details