Introduction to e-commerce website optimization

Transcription

1 Introduction to e-commerce website optimization 01 You re reading this because you run, own or have some responsibility for an e-commerce site. You want your website and mobile site to work harder, please more people and generate more cash. You know that around 95 per cent or more of the visitors who come to your site don t buy. That doesn t change when you throw more money at getting more eyes on the site. Whatever changes you have made to try and improve things have delivered mixed results. You keep bumping your head against these walls: how can you get more of your visitors to place an order? how can you get those who bought once to buy again and spend more when they do? The answer is in this book. It s not a silver bullet. You won t find a list of tactics that will magically make you more money. Alas, there are no latest User Experience (UX) best-practice tips that will do the trick. What you will get is a step-by-step guide to the structured approach used by companies all over the world. It s a proven and foolproof approach that s not based on the latest fads but on scientific principles that enable you to do what s best for you. Follow it and you, like others before you, will have a guaranteed way to continually improve your online sales. E-commerce website optimization You may know about website optimization, or Conversion Rate Optimization (CRO). Perhaps you ve dabbled in it. Over the last decade or so, it has gone from being the latest buzzword to becoming integral to the way e-commerce M01_ KAPF7538_01_C01.indd 1

2 2 E-commerce Website Optimization companies do business. Over 90 per cent of marketers now say that CRO is important if not crucial to their organization. 1 So what exactly is it? In the e-commerce context, CRO can be defined as a system for influencing (or converting ) a higher proportion of your visitors to buy more and spend more. It has nothing to do with bringing additional people to the site it s all about generating more sales from the visitors you already have. How does it work? The first step in the process is identifying ideas to improve the site s ability to generate revenue. How you come up with those ideas and how you decide which ones to focus on, is largely what determines success. A big part of this book is therefore devoted to this concept. You then test those ideas on actual visitors in real time to see whether they work. If they do, the tests also give you an indication of just how effective they really are. In just a few weeks from now, you could have a bank of valuable ideas with definite potential not hunches or rules about best practice that may or may not be relevant to you. Soon you will start to discover what makes your visitors take out their credit card, as well as some things that make them keep it firmly in their wallet. Overview of the process Most CRO systems use the tried and tested Scientific Method. Although different firms give their method different names, the principles are the same. This method has helped scientists in fields as diverse as social sciences, geology and chemistry to come up with new ways of understanding the world and, ultimately, improving aspects of our life. Central to The Scientific Method is a reliance on measurable evidence, hypotheses and theories, 2 as illustrated below: Step 1: Define the question One question drives all CRO efforts in e-commerce: How do we make more money? Everything you do on your optimization programme is aimed at answering this question. Often the unit of measurement for the extra sales generated is Revenue Per Visitor (RPV). M01_ KAPF7538_01_C01.indd 2

3 Introduction 3 Figure 1.1 The scientific method used in CRO Define the question Gather evidence Interpret the result and its implication for redefining the hypothesis Form a hypothesis Analyse the result Devise an experiment to test the hypothesis Step 2: Gather evidence Public results SOURCE Adapted from Lars Christensen, The Hands-On Guide for Science Communicators, Springer, 2007 Data from a variety of sources hold the key to finding out how to improve your average spend per visitor. Relying on guess work, opinions and even so-called best practice is like playing the lottery. You may win a bit of money here and there but it s all down to chance. By following an evidence-led approach, you remove luck from the scene and bet only on things that you know are likely to make a big difference. In the next chapter, we show you the types of tools you can use to get your hands on the data you need. Chapters 4 6 cover a range of research methods in detail. From this, you ll gather a huge amount of data which you need to go through and tie all the strands together to create a plan or roadmap for the months ahead. The detailed process of prioritizing what to work on first and what to ignore and then creating an Optimization Plan is covered in Chapter 7. M01_ KAPF7538_01_C01.indd 3

4 4 E-commerce Website Optimization Step 3: Formulate a hypothesis Based on what you ve observed, you create a hypothesis which predicts the outcome of a proposed change on the website. Make your prediction rationally, as a sensible estimate based on evidence, rather than an optimistic hope, built on your emotional desires. The prediction is tied to an appropriate metric, such as RPV, which is a number you can track in order to measure the effect, if any, of introducing something different. This process is explained in detail in Chapter 8. Step 4: Experiment You may be spot on with your hypothesis or totally wide of the mark. No-one knows until you put it to the test and run a scientific experiment. The results will either support or refute your hypothesis but either way, you ll learn something about what makes your customers buy more and spend more on your site. In the world of CRO, the experiment usually takes the form of an a/b split test. Half your visitors see your current page, and half see the new variation which incorporates changes that you predict will improve RPV. The test group has no idea they are seeing something different, as it appears to them seamlessly integrated into the main website. Yet behind the scenes, the technology is carefully monitoring who has seen which version, how they are behaving and, crucially, how much they are spending. After a short period of time, you compare results to see how the variation has altered their behaviour and which test group has spent more. Step 5: Analyse the result The experiment you run could have one of three outcomes: The new variation is a winner and your hypothesis is substantiated. The new variation does not win and the hypothesis is refuted. There is no difference between the new variation and control; it is inconclusive. On face value, it looks like you should have your fingers crossed for a win all the time. Yay! Crack open the campagne, make the test page live on your main site and watch all those visitors start spending. In fact, you should do your best to avoid this kind of trumph and diaster thinking and treat wins and negative uplifts in the same way. Long-term success comes out of being truly scientific. Hoping for a positive result means you ve emotionally invested in the result, which introduces a form of bias. M01_ KAPF7538_01_C01.indd 4

5 Introduction 5 Whatever the outcome, make it your aim to bank the learning and move forward. A win can often point the way to further tests around the same hypothesis that could give you even bigger uplifts. Tests that don t win are also valuable. There is usually a lot you can learn from them, insight you can use to design another variation that really rakes in the cash. Step 6: Interpret the result The purpose of a scientific study is often to contribute new knowledge: CRO is no different. Whether you have substantiated or rejected the hypothesis, with every concluded test you know more about your customers than you did before. Gradually, you will learn vast amounts about what works for your customers and what doesn t. Step 7: Record the result In the scientific community, the contribution of new knowledge is formalized by publishing the details of the experiment. Similarly, your work is not complete until you have documented your result. This includes recording the insights from each experiment to share with your colleagues in other parts of the business. Remember your central question: How do we make more money? With each experiment you collect new knowledge about your customers. Over time, this leads you to make long-term improvements in the site and generate increasing amounts of money from them. The rest of this book tells you exactly how to implement those seven steps into your own organization, with practical examples, step-bystep instructions and case studies from real e-commerce optimization projects. Summary A huge majority of your visitors don t buy what you ve got to offer. This book talks you through a proven system to help you to change that for the better and achieve substantially more revenue. It is based on The Scientific Method, which has been used for years by scientists in many disciplines, to develop new ways of understanding and improving aspects of our lives. It involves seven steps, anchored by the central question: How do we make more money? You start with research and analysis to gather M01_ KAPF7538_01_C01.indd 5

6 6 E-commerce Website Optimization evidence in order to identify the most valuable ways to improve, or optimize, your site. This gives you a host of ideas and, to make best use of your time, you rely on a system to prioritize which ones to work on first. In a scientific experiment, ideas are then tested in real time on the visitors to your mobile and web sites. Some of these ideas will generate higher revenues in the experiment and can go live on your site. Others may not give you the result you expected but will give you unique insights into your customers which could be the start of the next big winner. Notes 1 E-consultancy (2015), Conversion Rate Optimization Report econsultancy.com/reports/conversion-rate-optimization-report/ (Accessed: 1 July 2016). 2 Gauch, Hugh G (2012) Scientific Method in Brief, Cambridge University Press. M01_ KAPF7538_01_C01.indd 6