Convert - Designing Websites to Increase Traffic and Conversion Part 5 doc

STEP TWO: ANALYZING YOUR FUNNELS  ❘  103 funnels. In the meantime, simply look at the exit rates for key pages as an indicator of where people are leaving the site. FIGURE 6-5  Viewing a Google Analytics report in the context of a goal Step Two: Analyzing Your Funnels e next task is to map out the actual ow of trac through your main pages to your main goals, looking at three categories of visitor behavior: Proceeding to the next intended step  Going somewhere else on your site  Quitting the site  e purpose of this exercise is simply to visualize where your site is leaking the most visitors. It makes sense to start your optimization process with xing the biggest leaks, before you can move on to progressive renements. Measuring Your Conversion Rate Once you have identied one or more goals, you will be able to measure your starting conversion rate (CR). If you have congured a goal in Google Analytics, you will also be able to view the CR for that goal graphically (as shown in Figure 6-6), including changes in conversion rate over time. NOTE  If you have not set up goals in Analytics, you can still figure out your conversion rates for each goal simply by dividing the number of times each goal was reached in a particular period by the number of relevant visits to your web site in that period. 104  ❘  CHAPTER 6: MAKING YOUR SITE SELL FIGURE 6-6  The Analytics report for a single goal Knowing your conversion rate may be interesting, but it does not usually give you actionable information. If your conversion rate goes up or down, you need to know why, and at what steps the leaks are occurring. e overall conversion rate of a fun- nel is a high-level indication of its eectiveness, but in order to optimize, you need to break the process down into more detail, which is where funnels come in. Attrition Rate At every step through a funnel, some of the visitors who entered the funnel will leak away. e number of visitors who successfully move forward can only diminish with each step. It can never go up. e conversion rate of any funnel is a product of the cumulative attrition rate at each step along the path from an entry point to a measurable goal. Your objective is to deliver as many as possible to that goal, by minimizing the attrition rate at each step. Figure 6-7 shows a simplied view of a funnel. 100% of visitors enter the funnel at the top. 68% of those drop o at the rst step, then a further 47% are lost at the second step, and nally there is a 41% attrition at the nal step. e result is that only 10% of the original visitors make it all the way to the goal. Once you can visualize your conversion funnels like the example in Figure 6-7, you can start to identify the steps that demonstrate the greatest opportunity for improvement. STEP TWO: ANALYZING YOUR FUNNELS  ❘  105 In this case, the rst step (with the greatest attrition rate of 68%) would seem to oer the most potential for improvement, but all three steps merit attention. –41% –47% –68% 100% 17 10 32 FIGURE 6-7  Visualizing a funnel in terms of attrition rate Visualizing Your Funnels in Google Analytics e funnel visualization view in Google Analytics makes it quite straightforward to see the click-through rate from step to step, as Figure 6-8 shows. Analytics lets you dene the steps in your site’s funnel, and will give you a diagram showing how many visitors you convert or lose at each step. Once you have set up your goals and given them some time to run, you will be able to view many of the reports in Analytics within the context of your goals, as well as ana- lyze in more detail the paths people take to achieve the goal. e attrition rates in the funnel in Figure 6-8 are 89%, 63%, and 97%, respectively. Each score needs to be interpreted in the context of the page in question. From this site’s home page, common sense tells me that a click-through rate of 11% to the About page is not bad. I would be more concerned about the later steps. e horizontal colored bars at each step show the proportion of people who proceed to the next step or do not proceed. Of those that do not proceed, some may go some- where else on the site, and some may leave the site altogether. e lists shown on the right-hand side of each step indicate where those who do not proceed go instead. You should be most concerned about the exit rate. 106  ❘  CHAPTER 6: MAKING YOUR SITE SELL FIGURE 6-8  The funnel visualization view in Google Analytics The Law of Multiples Imagine the path through your web site, from home page to goal, is a 300-mile train line. If the track is strong, at, and clear for 299 miles, then heads across a bridge that has a section missing, how many trains are going to reach their destination? It doesn’t matter how well maintained the rest of the track is, how many trains begin the journey, or the value of goods they carry. If a bit of track is missing your success rate will be zero. Fix that section of track and the improvement in success rate will be astronomical. e Law of Diminishing Returns dictates that the closer you get to optimal, the harder it gets to make further improvements. STEP TWO: ANALYZING YOUR FUNNELS  ❘  107 e Law of Multiples tells us that the eects of the conversion rate of individual steps are cumulative, that is, any impact you have on any step along the path will apply to the whole conversion rate. Double the conversions at any step, and overall conversion will also double. Because it is practically impossible to double a click-through rate of 50%, you can much more easily double a click-through rate of only 2%. Either one would double the nal conversion rate, which is why we focus rst on the easier wins. As soon as you model your funnel, you will see exactly where to start focusing your optimization eorts. Unless you know the attrition rate from each step to the next, you could waste resources on steps that perform better. Goal Value, CPV, and EPV ere is no target conversion rate at this point. e rst target you should be inter- ested in is protability. To measure protability, it helps to know your Cost per Visitor (CPV) and Earnings per Visitor (EPV). If your web site’s EPV is more than its CPV, you should be in prot. In other words, the value your site delivers for your business is greater than the cost of running the site and attracting trac. (Later, your objective will be to increase protability—the margin by which your EPV exceeds CPV.) Only you will be able to provide these numbers. It is relatively easy to measure prot for an e-commerce site, but a web site’s indirect value is more subjective. For example, if your site generates sales leads for your business, you may want to estimate the aver- age prot earned from each lead, and assign that value to a lead-generation goal. If your site gathers a list of e-mail address, can you estimate the lifetime value of each name on that list? Consider the business value generated by each goal. Analytics lets you assign real or nominal values to any goal. Many e-commerce services will pass back an actual value that will tell Analytics the sale value of each transaction. (For more info, search Google for “How do I track e-commerce transactions?”) You can also give any goal an arbitrary value on the conguration screen (see Figure 6-3). Setting goal values lets you view the trac sources and paths through your site that are most protable, which will also help you target areas to optimize rst. If you know the real cost of the time and services spent on building and maintaining your site, you may be able to put a gure on your CPV by dividing that cost spread over 108  ❘  CHAPTER 6: MAKING YOUR SITE SELL time by the total number of visitors. A simple way to reduce your CPV and to increase your EPV-to-CPV ratio is to drive more trac to the site. However, bear in mind that while accurately targeted trac is likely to increase your average EPV, more of the wrong trac could reduce EPV, so raw visitor numbers are not always meaningful. NOTE  If a likely goal from your site is for visitors to pick up the phone and call you, you need a way to keep track of phone calls generated by the web site. A great way to do this is to use a telephone script that includes the question “How did you hear about us?” Also consider doing this for any face-to-face leads. When you have modeled your web site’s goals and funnels, you will soon be able to see each funnel’s performance in Google Analytics. Look rst for pages with high exit rates or low click-through rates. ese should indicate where you can achieve the most signicant improvements. Step Three: Optimizing Conversions through the Funnel In the “rst best guess” approach to web design, we built web pages from our own internal perspective, which we hoped would attract trac and help people nd what they were looking for. e new approach takes a far more direct and analytical approach. When you work with multiple funnels, you can spot how visitors are progressing with greater pre- cision, and take specic steps to optimize every step in each funnel to maximize conversions. Here is the simple and straightforward process for optimizing any funnel: 1. Identify leaks. 2. Generate alternative ideas. 3. Test to see what worked. 4. Repeat. e rst step is straightforward. Just look for the worst leaks in your most important funnels. STEP THREE: OPTIMIZING CONVERSIONS THROUGH THE FUNNEL  ❘  109 e second and third steps work together. You need to apply all your best knowledge and ideas to the problem. Creativity alone will not give you solid progress. You must test each change you make so that you know whether it had a positive or negative impact. Chapters 7, 8, and 9 take you through a comprehensive system for generating  ideas for how to make your pages work better. Chapter 10 applies all the principles to every step on the Awareness Ladder.  Chapter 11 introduces the methods you can use to test your ideas and to know  definitively what works best. Finally, optimization delivers best results when you do it as an iterative process. When you switch from the “rst best guess” approach to a culture of optimization, when- ever a change is made and tested, you get to see—and then choose to keep—only the things that are proven to work better. Every cycle builds on previous successes, creat- ing a culture of continual improvement. Optimization Combines Creativity and Analysis Analysis lets you measure performance with quantiable results. It will tell you if design option A converts better than design option B. Without these numbers, you can’t be sure of selecting the best-performing option. When you test dierent options iteratively, your numbers will ensure you only select the best-performing option, guaranteeing forward progress is achieved. However, testing and measuring can only describe what is—the past performance of A and B. Simply measuring performance can’t tell you why A is better than B, nor can it reveal what might work better. Consider that the process of optimizing your design is similar to the process a scien- tist follows when trying to gure out how things work. A scientist will start with an idea or hypothesis, then construct an experiment to test that idea. e experiment should deliver empirical evidence, which may indicate whether the idea is valid or invalid. at result may require a new hypothesis to be formed, or even lead the sci- entist to question the theories upon which the hypothesis was based. e scientic method has made possible huge advances in technology, despite the fact that it rarely gives a denitive answer. Because any theory is only as valid as the 110  ❘  CHAPTER 6: MAKING YOUR SITE SELL evidence that supports it, and the process is never complete, scientists always remain open to the possibility that their current working theory may be proven wrong. Apply exactly the same discipline to increasing the conversion rate of your web site. Testing will give you measurable results, but no nal answers. To nd an option that improves on your current best option requires insight into why one approach works better than another, and the ability to visualize an even better solution. e numbers may help point in the right direction, but they can’t take you there. In marketing and design, just as in science, previous theories exist that describe how things should work. It is foolish to ignore the lessons of the past that teach us how people tend to respond in dierent situations, and how to persuade them to be willing to take the action you want. Ecient optimization requires not only being familiar with this body of knowledge, but also having the insight to apply that knowledge to each new problem with sensitivity. Like a scientist, you need to be able to generate new ideas—hopefully even brilliant ashes of inspiration—and that is the domain of creativity. So what do we do if neither analysis nor creativity can give us the answers we want? e solution lies in unifying both methods into a single process, in which creative possibilities are explored and then validated through analysis. I’ll describe how this approach plays to the strengths of both mindsets and also provides a rewarding expe- rience for all stakeholders. The Limits of Creativity ough, in theory, creative endeavor is boundless, the process needs to be channeled in order to be ecient. It is not possible to test an innite number of creative possi- bilities without an innite budget. e optimal design solution is out there, but it can only be discovered through a disciplined, methodical approach. e challenge you face when optimizing your web site is how to harness the strengths of the creative process in the pursuit of continual improvement. If not every idea can be put to the test, you have to use heuristics (rules-of-thumb) from received wisdom to help direct your creative energy. Originality is a dangerous frontier land. At once, it is the place where great original advances can be discovered, but it is also full of opportunities to fail. e saying, “e pioneers take the arrows, the settlers take the land,” applies perfectly to the quest for the optimal web site. STEP THREE: OPTIMIZING CONVERSIONS THROUGH THE FUNNEL  ❘  111 Biology provides the perfect analogy to guide our discipline. Life constantly gener- ates mutations, most of which fail. Occasionally, a new solution emerges that is more successful than previous solutions, and which can become established in its environ- ment. is evolution works extremely well over a very long timeline (millions of new tests), and your resources will certainly not stretch that far. Use Conventions A design solution that has become established in its environment is called a conven- tion, and conventions provide the ideal starting point for optimization, oering the best probability of success. ey work because they have worked before, and because they are familiar. You don’t have to think how to use a log-in panel that requests “user name” and “password,” because you have seen and used thousands of similar panels before. e familiar is oen unconsciously perceived as safe and easy; the unfamiliar as risky or dicult. Is it possible that there is a more eective solution than the user name/ password log-in box? Certainly there is, but to nd it you may risk trying many poorly performing alternatives. e rule I always apply is to use conventions wherever the design problem is familiar (such as where to place a logo or the “log out” link) and where a conventional solution already exists. at leaves more creative energy to apply to the truly unique problems that every design challenge presents, which include the big questions of audience and proposition that actually most greatly impact success. So apply creative thought from start to nish to every aspect of your web design, take advantage of conventions that have been proven to work, but rely nally on the results of your own tests, because every web site’s design challenge is unique. Test Your Best Combine your creativity, insight, common sense, intuition, and experience to each new problem. But only try ideas that you have reason to think will work, avoiding random guesses. I would never advise testing any old dumb idea; you’ll probably waste time and money. For the best results, always strive to create a better solution, applying all the resources at your disposal (with just a dash of the unpredictable). Of course, you should never assume that your own best idea is any good at all. You must always test, but test things you have reason to think will be eective. 112  ❘  CHAPTER 6: MAKING YOUR SITE SELL In our team, we’ve coined the term “CICO” (which stands for “crap in, crap out”) to describe the eect of testing poor ideas. Especially when you start running tests on your site, it’s tempting to test a weak option for the thrill of seeing a result. When doing tests, your objective is to improve on the original (control) solution. ere’s barely any value in showing an idea that doesn’t perform as well as what you had before. It will most likely provide a negative learning experience, which may point in the direction of a better solution, but hasn’t carried you any nearer to that solution. At the same time, don’t be afraid to experiment. Because of the way split-testing works, you will always be able to observe performance, and respond quickly by kill- ing o poorly performing options and moving forward what works. Provided you are not practicing CICO, any detrimental eects will be short-lived but the gains you can make will stay with you. Attention and Momentum Attention is the currency of your web site. When new visitors arrive, they have a cer- tain amount of attention. ey use that attention to scan the landing page, looking for one thing. If they nd that one thing, it generates more attention. If they do not nd the one thing, their attention may run out. If the visitors get to the point that they believe they are more likely to nd what they want by going back to the search results than by persevering with your web site, you have lost them. e exact avor of the one thing depends on the visitors’ current goal. Do they want to nd the product they’re looking for, to get the answer to a question, or to be entertained? (If you have built your web site around the new multiplicity model, you should have a pretty good idea, because you know which funnel they arrived on.) Whatever the visitor’s precise goal may be, the one thing is always to answer one simple question: “Am I in the right place?” at’s it. Answering that one tiny question is the only thing your web page needs to do. e more quickly and decisively you can answer the question, the more attention your pages will generate, and the more people you will keep interested so that they continue to move up the Awareness Ladder. e challenge, of course, is how to answer the question most eectively. at’s what the rest of the book is about. [...]... The two factors tend to go hand-in-hand Writing you-oriented copy requires stepping into the visitor’s point of view, where features appear as benefits Figure  7-4  ​​ Rewriting this introduction to be more you-oriented and benefit-rich increased click through by 10.8% Match the Promise of the Link One of the early warning signs of low conversion rates is question marks floating above visitors’ heads... can refer to his landing matrix and craft a message specifically for a particular market (middle-aged men who are starting to go bald) and a particular level of awareness (know of his product, but do not trust its effectiveness) Joe might write, “How Miracle Hair-Gro Stopped Me Looking Old Before My Time.” Self-Interest The second critical factor in creating appeals to capture attention is to make it... looking to you to give them something to get their interest, in order to be encouraged to read on Remember that your target visitors (that is, the undecided ones) do want to be in the right place, and they are open to being interested All your page needs to do is to provide enough interest to persuade them there is value on offer Because we look for noticeable text to confirm where we are, and that... Home—Risk-Free” uu “10 uu Simple Tips That Could Double the Traffic to Your Web Site” “The uu Best Hair-Loss Treatments Revealed” I look for three major factors in crafting an appeal This chapter addresses each of these factors and shows you when and how to use them to best effect: Relevance—This page should be appropriate to my present situation uu Self-interest—It should offer me a quick solution to my... site’s highest-level offering, low-cost towels All landing pages should seek to be as specific as they can be, while still including the whole relevant audience A diverse landing page strategy makes it far easier to be specific, which means less work and better results Target a Specific Market For the towels site, the market is at Step 2 They know what they want: to save money buying towels and linens... they need to see to know they’re on track If you already have a matrix of landing pages targeted to specific search terms, you may even know the precise language they used to find you, so it’s great to use that language To have any chance at converting visitors to a goal, you must first convince them that they are likely to get what they want from your web site Unless they are compelled to look further,... n Similarly, don’t expect your initial message to convince All it needs to do is engage Trying to achieve too much too early could be counter-productive A good rule of thumb for headlines is if your reader will think, “I can skim-read that, and come away with something useful.” Examples include: “How uu to Attract The Perfect Partner” 5 uu Proven Ways To Succeed at Interviews” “How uu I Saved $4000... segments, the challenge is to get as many visitors as possible to proceed up the ladder The Three Elements of Conversion For a page to get visitors to proceed to the next step, it must let them know they are in the right place To do that, three things need to happen These are the three critical elements of conversion: Get uu their Attention Keep uu them Engaged Call uu them to Action All three elements... awareness Joe’s job at that point is not to get those people to buy straightaway! That would be unrealistic and unsuccessful Remember, every prospect has to pass at some point Th e Th r e e E l e m e n t s o f Co n v e r s i o n  ❘   1 1 5 from Step 0 to Step 5 on the Awareness Ladder, and through every step along the way in sequence All Joe has to do is to get folks to take the next step up the ladder... gain—Offer me a solution to what I want now uu Each stage of your visitors’ journey starts with an appealing proposition that seems to speak to them and to offer something they want The next chapter is packed with examples and ideas for transforming your propositions into attention magnets Keep Them Engaged Now that you have someone’s attention, you need to get them to proceed toward their next goal . visitors. A simple way to reduce your CPV and to increase your EPV -to- CPV ratio is to drive more trac to the site. However, bear in mind that while accurately targeted trac is likely to increase. more segments, the challenge is to get as many visitors as possible to proceed up the ladder. The Three Elements of Conversion For a page to get visitors to proceed to the next step, it must let. landing pages targeted to speci c search terms, you may even know the precise language they used to  nd you, so it’s great to use that language. To have any chance at converting visitors to