Two Design Patterns for Better Conversion Rates

Exit through the shopping cart – using persuaders and affordances to tell your customers why they should buy.

Written by Frank Gaine

Our clients have recently achieved real results when it comes to customer conversion and behavioural change, from medical drug adherence to online holiday retailing. We have helped them do this by applying critical thinking to Nudge Theory, Persuasion, Behavioural Economics, and an analysis of how industry leaders successfully implement customer journeys and conversion messaging. In this way, our designs not only solve user-focussed problems but also tackle business-focussed problems.

The Power of Research

Before you can innovate, you first need to understand the problem. The importance of research cannot be understated, but research can take many forms, from ethnographic studies with real users, to interviews with internal stakeholders within the business. An essential part of our research process is looking at design best practice within our client’s field of endeavour and across other industry environments where learning can be derived.

Travel and Beyond

Inspiration can come from a wide range of sources. So, for example, when working with a travel platform provider, we not only looked for inspiration within that sector, but cast our net further afield to include consumer electronics, telecoms, finance, and healthcare industries. Once we had gathered the insights, the learning was incorporated into prototype designs which we presented to real users and business stakeholders for evaluation. Let’s take a look at some of the conclusions from this research.

Customer Journeys

Online holiday retailing is where a provider bundles a number of products together and presents them to the user at a single price. Typically those individual products comprise a flight, a hotel room, and a rental car. In this situation, the user gets a discount compared with buying the bundle elements individually and the vendor gets a chance to offload its distressed inventory, for example, less popular flight times.

Our aim was to discover the most common customer journey types and to see which, if any, would be appropriate to our client. Our analysis showed that there were two predominant customer journeys applied by retailers, which we labelled Loop and Linear. After analysing the pros and cons of these existing models, we proposed a third model which we believe is more efficient and will lead to better conversion.


Virtually all retailers focus on the hotel as the first decision for the user to make in selecting their holiday package, primarily because hotels are the least commoditized product and the most anticipated part of the holiday by the user. The Loop customer journey allows the user to loop out of the process of selecting their Hotel in order to edit the specific flight or car hire that is also suggested to them at the same time. In this situation, there is too much going on. There are too many decisions for the user to make at the one time. It is overwhelming and increases the chances of the user abandoning their holiday shopping.


In the second predominant customer journey, which we have called Linear, the user is asked to consecutively choose a Hotel, then a Flight and finally the Car Hire. We observed that this lengthened the entire process and increased the chances of the user ignoring the distressed inventory items by being given too much time to focus on the multitude of options available.

The Anchor

Therefore, we proposed an all-new customer journey which none of the competitor retailers was using; the Anchor. This is where the user is allowed to focus solely on their Hotel selection, which is then placed into the cart alongside a suggested flight and car hire option, both of which are editable from that point.

In this situation, there is less mental load on the user when selecting the Hotel as they remain yet unburdened by their flight and rental car choice. Additionally, the time to cart is dramatically reduced and the cart works as a natural anchor point to go and confidently edit the flight or car suggested. Customers will be less likely to choose a flight or car hire other than those you recommend. In line with all good design practice, this hypothesis is currently undergoing testing and initial results look very positive indeed.

Statistics show that travel carts are abandoned 81.7% of the time. 27% of carts were dropped because the checkout process was too complex or because the site was slow.

— Grace Miller, Annex Cloud

Conversion Messaging

Conversion messaging is another area where we have applied critical thinking to nudge theory and industry best practice. Conversion messages are everywhere in online retailing and might be most familiar to you as these kinds of messages ‘Only One Left’”, ‘Free Cancellation’, ‘Great Value’ and such like. Indeed, how and when these messages appear might seem arbitrary or unconsidered at first glance. Some sites present a veritable maze of those messages.

Ticket sales for the Smithwick’s Experience Kilkenny are up 89% and overall online revenue has more than doubled.

— Mark McGovern of Smithwick’s speaking about the results forged by’s design

Conversion Categories

It is only when you have analysed enough of these messages that patterns emerge. Over the course of our research, we looked at thousands of conversion messages from dozens of cross-industry sites and found that any given message will inevitably fall into one of five categories;

Vendor Generated — for example, ‘Great Value’
User Generated — for example, ‘Average Reviews 9/10’
Price Specific — for example, ‘Save 10%’
Benefits Specific — for example, ‘Free Cancellation’
Urgency Related — for example, ‘Only 2 Left’

Building upon Dr Robert Cialdini’s Rules of Persuasion, we found that more than one Rule of Persuasion can be used to characterise any of the five conversion categories that we established. For instance, Vendor Generated messages can draw on both Cialdini’s principles of Authority and Liking, for example. ‘Top Hotel in Amsterdam’ and ‘Best for Your Family’ respectively.

An Intelligent Framework

Using this framework, we were able to show our client how they could intelligently apply conversion messaging into travel retailing. The nature and extent of the conversion messages employed will be determined by the data and APIs available. It also depends on the type of relationship the retailer wants to forge with their customer. The job of the visual designer now becomes even more important, to use colour, contrast, white space and directional cues to ensure the presentation focussed and uncluttered.

Hotel search result with conversion messaging applied.

Added to that, the more you know about your customer, the more you are in a position to change the behaviour of that customer, from being a one-time user to being a customer who really believes that you understand the kind of stay they want to have at a hotel. For example, if your data shows you that your customer has a propensity to book a hotel with a swimming pool, then you can serve up a more enticing image of the hotel on the search results page. A sunny pool area, instead of a photograph of the exterior of the hotel for example.

Hotel search result with personalised image applied.

Conversion 2.0

Conversion messages, while currently effective at helping people with their procrastination, inevitably introduce a level of stress into the relationship. We believe that stress is not the best basis for a relationship with the customer. It does not breed loyalty or commitment over the long run. This paradigm will only be successful until a better way of relating to the customer emerges, Conversion 2.0 as it were. That is where is headed next.

Man in airport using smartphone while awaiting his flight.

Partnering with OpenJaw for Online Travel Excellence

Answering the unique needs of travellers in today’s world with user-centred design.
People attending the Smithwicks Experience Kilkenny tour.

Transforming Online Sales for Smithwick’s

Merging brand and emotion for phenomenal online conversion.