The atmosphere of the meeting has always been lively. During the website analysis portion of my session as a guest, I answered a total of four questions from the audience. These four questions are very classic and representative, and I am eager to share them with my friends. However, the answer to each question is not that simple. I am going to divide it into two serialized articles. Please wait patiently for the next article after reading this article.
First question:
A friend who works on an e-commerce website said that he often looks at data, but data is just data. Is there any way to find something from the data that can further help his business. He further asked, are there any models and methods for website analysis that can be used directly?
This question is simply awesome. This is the common sentiment of many website owners - we need data, but what should we do if we have the data? And this is also the value and role of website analysis!
I have always maintained that tools are important, but the mind that uses them is more important. In order for the tool to play its due value, we should master some of the most basic thinking methods - methods do exist. That's what the picture on the right represents. This diagram is so simple yet so profound that I even regard it as the Yin Yang Tai Chi diagram in website analysis. That's right, there are three eternal things we need to do after we get the data - check trends, check segmentation, and check conversions. Then, we need to combine these three to view segmented trends, segmented conversions, and view conversion trends. It seems like there are only 6 situations, but in actual applications they are ever-changing and can tell us a lot of valuable information. With this diagram and operating according to this thinking model, website analysis can really help business.
For example, e-commerce is most concerned with the sale of goods, and the core lies in the goods themselves. Then through website analysis and the use of segmented conversion methods, we can know some special attributes of the product that we may not have noticed before. The specific method is to classify the product according to two dimensions - the amount of attention the product has received (which can be easily obtained through website analysis tools) and the sales conversion of the product (that is, the ratio of sales volume to the amount of attention), through the e-commerce backend and website analysis tools. can be calculated) to subdivide. After simple calculations, we can get the following picture:
There is no complicated mathematical calculation required to get this graph, and you only need to use one metric of the website analysis tool - the Page View of the product page to get this graph. But the value this picture brings to us is huge.
For the upper right quadrant - products with high sales conversion and high attention, of course, you can keep the status quo, and even give them some promotions to further stimulate sales.
For the lower right quadrant - products that have received insufficient attention but have good sales conversions are obviously some potential profit growth points. With this situation, you can immediately do two things: 1. Immediately study why they can convert so well. Is it because of some factors that have not been paid attention to? Everyone knows that this situation often happens. 2. Immediately give these low-attention products more exposure opportunities, so that they can increase their attention and enter the list of star products (upper right quadrant).
For the upper left quadrant - products with high attention but poor sales conversion, it obviously indicates that there is a problem of insufficient motivation to promote transactions. The product manager should immediately start to study how to promote sales conversion - is it a problem with the shopping cart or inconvenient payment? Or are competitors' prices more attractive, or are their promotional offers not enough? Wait, wait, etc. This quadrant, like the previous quadrant (lower right quadrant), is a place where great achievements can be made.
The last quadrant in the lower left corner is disliked by everyone. No one pays attention to it, and the sales situation is not ideal. So, see if there is an opportunity to increase exposure, or do some promotions? Or, buy less goods. If it's a skimming product, it doesn't matter if it stays there.
Isn’t it interesting? In fact, just two simple measurements - Page View and sales volume, can tell us so many strategic things. This is the power of conversion and segmentation.
If we add another measure, which is the size of the pie in the picture - the price of the commodity, then it can tell us even more information. In the picture, the product in the lower left corner has the highest price. So, is the price an important factor that makes it sink? - It is worth digging into by the product manager. Likewise, trends can also give us a lot of stories and discoveries. I'm going to save it for my first blog meeting to share it with you. Stay tuned!
Second question:
This question was raised by a friend of an e-commerce website who does SEM promotion. The story is this: His website attracts traffic through SEM, and then there are 400 toll-free numbers on the website to accept sales intentions (Leads); but he found that his SEM was very expensive, and the conversion of the resulting Leads was particularly unsatisfactory - —Spent a lot of money and actually had poor results. He wanted to ask if website analytics could help him.
Classic question, I can’t help but praise this dear friend! Isn’t this just another question we ask ourselves every day?
My answer is: Website analysis can help you so much!
Of course, your problem is actually not one problem anymore. We need multiple methods to work together.
First of all, this website uses SEM, but it is very expensive, so the issue of SEM return on investment (ROI) must also be solved.
Secondly, this website is all about conversions. Then this is to solve the problem of poor traffic conversion rate.
Finally, is there any relationship between SEM strategy and final conversion? It is actually the intersection of the first two questions.
You see, such a question is actually three classic situations. How great! Then we will answer them one by one.
I.
Many, many websites spend money on SEM. How do you measure the effectiveness of SEM? Is it right to use click-through rate (CTR)? In fact, CTR is very important, but it is only the first step in several conversions. We also have two jumps, three jumps... and n jumps that are often referred to in the industry. With these jumps, we can calculate the conversion status of each SEM keyword during the jump process, and then measure which keywords jump well enough and which ones cannot. I believe many friends have already used this method.
Another model is that you can measure the effectiveness of different keywords by the cost of completing key jump (conversion) rates. For example, if you jump to the third jump, keyword A is 3 yuan each, and keyword B is 10 yuan each. Generally speaking, you know which keyword to invest in more.
Of course, SEM itself is already a profession. How to choose keywords and how to determine the placement strategy requires a lot of skills and knowledge. I am not an expert in this field. You can ask Tony (Liao Chenqing) for advice. However, I want to say that website analysis can definitely help you understand the actual effect behind each keyword. This is actually the fundamental reason why Google launched Google Analytics and why Baidu launched Baidu Statistics - it is to let you know the effect of the money you spend, so that you can spend your money in a safe way.
Once you know the effect, you can target it. Doesn't this save money? Doesn't this increase ROI? From this perspective, website analysis can directly help you spend your money wisely.
II.
Next, your problem is poor conversion, and traffic from SEM generally has the problem of "not being able to jump".
This is really a classic question.
Failure to jump means poor conversion. Poor conversions are often caused by the following reasons (note that the first reason is of particular concern):
1. The keywords (or advertising creative content) do not match the landing page.
It's so common. Bai Yao even coincidentally gave a random example from reality in the following meeting.
The keyword says, "I have discounted products here", but after logging in to the page, it is just the homepage of an ordinary mall, and you can't see where the discounted products are - and then the visitor immediately leaves in anger and disappointment. The reason why this happens is that when buying keywords, you only pay attention to attracting visitors, but ignore that they will actually "leave".
Never let your visitors (each one of them your potential customers) down. Match, be sure to remember to match!!
There are many ways to solve this problem. One is, why not set the landing page of this keyword as your promotion page? Isn't this a double benefit? Or, you can change the keyword to something that can attract people on your landing page.
How to discover the existence of this mismatch? Sometimes it is obvious at a glance (expectation mismatch), and sometimes it is not so easy (logic mismatch). My suggestion is to use the Bounce Rate value to measure the "angry and disappointed" users of the landing page, and then find out specifically whether there is a mismatch or other reasons (discussed below). Bounce Rate is another extremely simple metric that is provided by almost every website analysis tool.
2. The design of the key conversion page is unreasonable, unclear, and difficult to use.
Also an all too common question.
Every page is beautiful? Unfortunately, visitors don’t know what to do next. The most typical example occurs on the landing page - no clear instructions, no call to action, no navigation... and so on. This will also cause the bounce rate to be very high, even if the content of your landing page matches the content of the keywords (or ads) placed.
Key pages refer to node pages that must be passed through in the conversion process, such as add to shopping cart, payment page, order page, etc. The quality of the design of these key pages is too important for the final conversion of the website.
Using the Pathing tool of the website analysis tool, the Funnel tool or the Next Page Flow tool can tell us the pages and page conversion information we need above. There are many methods and reports on website analysis tools, all with the same purpose, which is to complete the "conversion" method we mentioned above. The purpose of these tools is to help you discover where there are conversion barriers, which is what we call leak points (where is the leak point, haha:)), and then read UCD China’s article to learn how to make it more reasonable. Design the page to cover the leak now!
3. The website conversion logic is unreasonable.
Every website designer loves every web page they create, but real product designers pay more attention to the macro structure.
If the macro structure is unreasonable, no matter how reasonable each page is, what's the use!
For example, if something that can be done in 5 steps has to be divided into 2 steps, and it turns out to be 10 steps, visitors will lose patience. Originally, the main process could complete the task very well, but some interfering side processes had to be placed. As a result, the visitors felt like they were in a maze. There is a mismatch between the logical next step that the user expects and the actual next step, etc. etc. - ugh, there are too many. These situations really shouldn't happen.
Bai Ya gave an example at the meeting, and everyone was happy. Joyo.com takes more than 20 pages to complete from confirming the product to final payment! We all know from website analysis that if there is no more link, there will be more losses. Is it necessary to complete the ordering task in 20 pages?
White Crow emphasizes - only leave the most critical processes to the most critical tasks, retaining customers is the first priority, and discard all other obstacles! One word comes to mind: Occam's razor.
How to identify this situation? Usually we will use usability testing (Usability Test) to find these problems. Website analytics can certainly help - as is the Pathing report. If the logical structure is unreasonable, the visitor's Path will be confused and the leak points will be scattered. If these two characteristics are present, then there is a high possibility that there is a problem with the website structure itself.
If your website looks like this, it’s time to hire a professional product designer.
III.
Finally, your question is whether there is any relationship between SEM strategies and final conversions.
Of course, I have already answered it before - it must match. However, SEM's strategy does not necessarily determine the final conversion, because (what I say below is extremely important) - SEM determines whether the website can obtain enough target traffic, but whether these target traffic can achieve your goals as you expect. , depends on the power of your website.
How to make your website powerful? - You must analyze the behavior of the traffic on your website and judge their interest, expectations, anger and even disappointment through behavior; you must also test the design (or Hands and feet) to see whether visitors can act according to your expectations (or even fall into the trap you set).
Once you have website analysis, you have periscopes, radars and sonar, otherwise you will always be lurking in the deep sea - anyway, the sea is very big and deep, and I can run rampant until I hit the rocks. Or, conversely, you can move toward your stated goals.
Once you have website analysis, you will know how you should design - because it is the actual visitors who are giving you guidance, and they are telling you what to do correctly. ——Isn’t this significant enough?
Finally, after reading the first half of my article, you can also conclude - it turns out that website analysis is preceded by SEM (Internet promotion) and followed by product design (UCD). It turns out that its use is so wide!
It's time to pay attention to her. [End of previous article]