【Foreword】
Tools are an indispensable element of website analysis. But how to choose a satisfactory tool is not that simple. How to choose the right tool according to the needs of different websites? This has become a question that many friends are concerned about. In the first part of this topic, we discussed the choice of tools for small e-commerce websites. In the second part, we look at how medium-sized and large e-commerce websites choose website analysis tools.
Correspondence between needs and tools (continued 1)
Type 2: Large and medium-sized e-commerce websites
In my opinion, the website analysis needs of large and medium-sized e-commerce websites are divided into two categories. One category is: no, no, no category (very interesting, I just watched Lao Luo’s speech that Tenly forwarded to me, and I immediately thought of this category); the other category is: I want everything to be category.
E-commerce websites of this type either do not need website analysis, or do not need additional tools for website analysis. I often hear: "What is website analysis? I don't need it. What I need most is traffic flow traffic!" It is true that these websites have their own log files for analysis, but generally speaking, website analysis is not enough for these websites. Not urgent. Even though the site isn't perfect, it's been running like this for years and there's nothing bad about its performance. Moreover, modifying the website will affect the effectiveness of my SEO, and people in various executive departments will also be very unhappy. The workload will increase and there will be more rules to measure them.
It is difficult for me to comment on this concept. After all, e-commerce in China is still in an era of extensive competition. E-commerce is low-profit or even negative-profit, but it doesn’t matter. As long as I have scale, I will have investment. If I have scale, I can continue to expand the scale. This makes e-commerce in China less intensive. , grabbing resources is the last word. In addition to e-commerce websites, group buying websites show this feature more strongly. I definitely don’t dare to laugh at this approach. What kind of business environment will have what kind of business rules, what kind of business rules will produce what kind of business strategies. In this magical land of China, it would be embarrassing not to do this. Not normal.
But I firmly believe that at some point, this will all change quickly.
Another school of thought is that I want everything to be similar, hoping that website analysis can solve all website optimization problems. This hope is not very specific, but it can convey a clear desire for my website to be optimized. This idea cannot be said to be wrong, but website analysis is not a panacea and cannot solve all website problems. The starting point of website analysis is user behavior. If the website has few users and the users are not active, then there will not be many areas where the website can perform website analysis. Similarly, if you hope that the SEO ranking of your website can be quickly improved by relying on website analysis, this is not the right tool.
Okay, no more gossip. For large and medium-sized e-commerce websites, the real areas where website analysis shines are as follows:
Locate website structure and page problems in the shopping conversion process; provide possible optimization solutions; test and fix optimization solutions.
Positioning, segmentation and analysis of traffic; providing comparison of traffic effects based on segmentation.
In addition to directly analyzing the traffic effect from the perspective of end-to-end ROI (investment return), it can also provide effect analysis from the perspective of Engagement (i.e., the various behaviors and degrees of traffic participation in the website); or in other words, not only can we see the traffic Input and output, you can also see the process of flow input and output on the website. Only by being able to see the process can we have the opportunity to optimize.
Product category analysis: Which categories have good conversions, which ones do not, and what are the possible reasons.
Product analysis: Which products have good conversions and which products need to be added (products that consumers demand but are not provided).
Visitor analysis: Segment visitors, for example, according to the number of website visits into new visitors, repeat visitors and loyal visitors, or according to the purchase of goods by customers (for example, new customers, repeat customers and loyal customers); or according to geography Segments and other rules segment visitors and segment them into different behaviors, conversion patterns, and more.
The requirements listed above are usually things that a large and medium-sized e-commerce website must do when conducting website analysis. In the previous two articles, the requirements for e-commerce website analysis were systematically explained from a broader perspective: What should we pay attention to when analyzing e-commerce websites (Part 1); What should we pay attention to when analyzing e-commerce websites (Part 2). It is recommended that everyone can refer to it.
As you can see, the above analysis and optimization content are difficult to achieve simply by relying on free website analysis tools. Tools such as CNZZ mentioned by friends in the previous article have no problems in monitoring traffic, but when it comes to the segmentation level combined with business, they cannot provide solutions. After all, it is a free traffic tool, so you can’t ask for too much. Therefore, a friend left a message and said well, "Website analysis should be more than just traffic analysis." This is indeed the case. This is a difference in demand. If you just look at the changes in traffic without doing more in-depth analysis, then CNZZ and 51la are enough. Using more complex tools is a waste. But for website analysis, especially for large and medium-sized e-commerce websites, good tools are very important.
What kind of website analysis tools do large and medium-sized e-commerce websites need?
As mentioned before, tools are not good or bad, only the tools that meet your needs are good. When choosing a website analysis tool for large and medium-sized e-commerce websites, you need to ask the tool provider about the following functional requirements:
Traffic location and monitoring
This function was first implemented by Omniture SiteCatalyst, that is, through the SAINT Classification method, unlimited marking of traffic sources can be achieved. This method was later simplified by Google Analytics into the form of UTM external links. The segmentation attributes of the traffic were limited to no more than 6, but it was enough. Regarding this function, please refer to this article: Use Google Analytics’ Link Tag to gain an in-depth understanding of the quality of traffic sources (advertising).
This feature is very powerful for segmenting traffic sources.
Full path and custom path
As mentioned in the capabilities and incompetences of Google Analytics, Google Analytics actually does not have full path analysis and post-customized path analysis functions. It only has predefined path and conversion analysis. Relatively speaking, for large and medium-sized e-commerce websites where path and conversion analysis are very complicated. As far as I'm concerned, it's very ineffective. Omniture SiteCatalyst is very powerful here, and path analysis is its specialty. Other commercial tools should also be good in this area.
Flexible segmentation capabilities
Flexible segmentation means that both predefined segments and custom segments are powerful. Google Analytics’ predefined segments are awesome, one of the best tools I’ve ever seen, bar none. In terms of customization, Omniture SiteCatalyst can achieve unlimited custom segmentation, but you need to think about it before implementing the tool, and then define it when the tool is deployed. If you miss a place you want to subdivide, you can always upgrade the tool to make it up in the future. Therefore, I often tell my friends that using Omniture’s tools without customization is absolutely a waste. While up-front customization requires strategic thinking and implementation effort, it's worth it in the future if the tool is specific to your business.
However, currently there is no tool that can fulfill all analysis needs through default predefined functions. I think that due to the uniqueness of each business, such a tool may never exist unless - a tool that can monitor and interpret all variables on the web page, and record all the startup and data changes of these variables and press Tools for logical separation appear (it's so confusing and complicated that I think it's difficult for such a tool to come out).
Integrate with business dynamics
This is relatively difficult, but commercial tools should be able to do it, otherwise you have reasons not to choose it. The so-called integration with business dynamics means that website analysis tools not only monitor data from the perspective of traffic (traffic process monitoring), but also provide data monitoring from the perspective of business flows caused by the flow of traffic within the website (business process monitoring) . It's more abstract, let's give an example. For example, when you purchase products on Jingdong Mall, your visits to each page will be recorded by website analysis. This is the flow process of traffic, using the page as the carrier. If you complete the purchase of a product, the flow process can be recorded as: "View product information" -> "Add to shopping cart" -> "Fill in receipt information" -> "Confirm purchase" -> " "Payment" -> "Payment Successful", this is a dynamic business process (as shown in the figure below). Commercial tools describe this process not only from a traffic perspective, but also from a business perspective. The difference between the traffic process and the business process is that the traffic process uses the page as the carrier. Due to the different URLs or page names of different product pages, it is too complicated to use the traffic process to describe the overall business process. By directly defining the business process, business flow analysis can be easily implemented. Free tools are all for traffic process monitoring, and some paid commercial tools have business process monitoring. This is why Omniture SiteCatalyst distinguishes between Traffic PRops and conversion evars+successful events, because the former is the traffic process, and the latter describes the business process (friends who have not used Omniture do not need to delve into this point).
Product tagging, segmentation and conversion tracking
The attributes of the product itself are very complex, including product name, type, SKU, price, inventory status... Some e-commerce websites have as many as 30 product attributes. Free tools make it too difficult to record this stuff. If it cannot be recorded well, it cannot be segmented well. In addition, product conversion tracking is a business process, not just a traffic process.
customer segmentation
Free tools generally only provide segmentation of new and old visitors. Commercial tools consider the business process perspective, extract the customers among the visitors (visitors who have purchased goods), and classify them according to new customers (those who can only buy goods once) , distinguish repeat customers (those who purchase goods twice) and loyal customers (those who make three or more purchases). In addition, for some other attributes of visitors, such as whether they are registered customers, whether they are silver, gold, or diamond card customers, as well as the personal attributes filled in by customers when registering or shopping, using this website analysis tool, you should add monitoring variables Implement monitoring. It seems that there are very few websites in China that use website analysis tools to conduct such detailed website analysis from the customer's perspective. However, using website analysis tools to segment customers and combine them with traffic processes and business processes is absolutely powerful in analyzing and mining customers. . Other customer segmentation such as geographical segmentation, browser version, operating system version, javascript support, etc. Needless to say these basic ones.
These are some of the features that are necessary for a website analytics tool that serves large and medium-sized e-commerce websites.
In addition, even if some tools can achieve the above functions, you should pay attention to some other places (which may cause you pain in the future).
Be careful to avoid areas that will hurt your balls in the future
Since commercial tools must take care of business processes, they either have strong customizability capabilities or provide segmented customization solutions for a certain field (such as the Xiaoai Statistics we mentioned earlier). We won’t take a look at the latter, as it is not a solution for large and medium-sized e-commerce websites. For the former, customization capabilities can be a double-edged sword.
It is no exaggeration to say that good customization requires effort. On the one hand, you must have a very good understanding of the business and process of your own website. On the other hand, you must have a very clear understanding of your own website results, URL structure, and SEO structure. On the other hand, you must have a very clear understanding of your own monitoring expectations. All three are indispensable.
If you don't want to spend this effort and let outsiders do it for you, you will definitely end up applying powerful tools in a nondescript manner. This kind of cup is actually common.
I think that the choice of website analysis tools is only one aspect. The function is powerful, but the analysis team does not have it or is very weak, then the more powerful the tool, the worse things may be, and it is not as happy as free tools. I have this experience. I can take pretty good photos with my Pentax SLR camera, but if I hand it over to a girl, the photos will not be as good as the cards in her hand. This is not to discriminate against girls, this is a powerful tool that requires strong ability to match, the so-called good horse with a good saddle.
Let’s not go too far, it’s getting late. If you don’t pay attention at first, the things that will cause you pain in the future include:
The definition of the page is too arbitrary, for example, forgetting to make a unified identification for all pages of a category, or forgetting to attach a business process to the page process.
Incomplete definition of segmentation requirements - if you forget to submit some important segmentation data requirements that need to be provided by the tool, subsequent analysis will definitely be compromised.
In the initial implementation, the pursuit of comprehensive monitoring resulted in difficulties in implementation. There will be too many data reports in the future, which will greatly increase the difficulty of analysis. ——This point and the previous point illustrate that it is very necessary to grasp the initial needs and the degree of segmentation.
If traffic identification and product identification are not carefully done, the final analysis will have to be done all over again, or the data will not be obtained.
Once you use commercial tools, if you are afraid of discomfort in the future, you can talk to me first. I have a lot of very depressing experiences here, but if you notice the failures of others, you can at least avoid making some detours.
Even with tools like Google Analytics, if you don't do it right at the beginning, you will inevitably suffer from pain in the future. Take a look at these two articles, they may be able to help you: be wary of the pitfalls of website analysis and monitoring implementation (Part 1); be wary of the pitfalls of website analysis and monitoring implementation (Part 2).
Serve
If you use paid tools to analyze large and medium-sized e-commerce websites, you must pay special attention to the service. Because these tools are not that easy to use. I recommend purchasing from the tool manufacturer first or requesting three types of services.
Category 1: Use help. You don’t necessarily need someone to be present, but you must have someone official to help you when you are unsure.
Category 2: Let official people help to do an analysis together, and reproduce the specific tools and methods used in the analysis results. This is a great way to master the tool very quickly.
Category 3: Complete knowledge base. In the middle of the night, I really couldn’t figure it out, so I could read the article and figure it out on my own.
Don’t choose tools that don’t have any services, I won’t tell you why.
Summary of this article
Because large and medium-sized e-commerce websites are too complex, this article is too much to say at once. It is possible that some friends find it difficult to understand, so please leave me a message to ask. There are several points to pay attention to when selecting website analysis tools for large and medium-sized e-commerce websites: 1) customization capabilities; 2) conversion and path capabilities; 3) business process monitoring; 4) service capabilities. As for whether there are any success stories about this tool, I don’t think it’s even important. Many tool cases are unsuccessful, which is not the fault of the tool at all. The customer is also responsible. After all, the cases of applying these tools in China's business environment cannot be said to be successful - they are too new.
Finally, I wish you all good luck. If you have any questions, leave me a message. If you have problems when you start using commercial tools, please also contact me! A new week has begun, and I wish everyone a happy new week.
Original link: http://www.chinawebanalytics.cn/how-to-choose-web-analytics-tools-2/
Please indicate the source when reprinting