How to write an executable web Analytics report Essentials

Intermediary transaction SEO diagnosis Taobao guest Cloud host technology Hall

This article translates from Avinash Kaushik's latest blog post: The "Action Dashboard" (an alternative to crappy Dashboards), please click on the link to view the original. Thank Avinash generously agree with me to translate his article, you know, he is the master of the Web Analytics world! Here is the full text of this article:

Do you know the difference between a crappy report and an elite report?

One is simply providing data, and the other is providing real information.

That's why most of the summary reports are the "crappy" core reasons, that is, they are a stack of data, but do not provide a slightest context for the data, nor do they give a slightest possibility.

Here are some examples of less-than-good summary reports, saying that they are not good because I always think of the feasibility of the value to measure them.

The above is perhaps the most common kind of summary report, there are a lot of data, even dig deep, but you can not look at him and said: "Wow, it seems we should be good in ... strengthened. " No, you can't see it at all.

I took out the other report above simply because it represents the characteristics of the 80% Web Analytics Summary report. In Excel, there are countless tags (tabs) that allow you to click. It's not a brief report, it's a lot of money. Ask a counselor to show off his/her Excel skills, but in fact--the report doesn't tell you anything worthwhile.

Let's keep looking. You often walk behind someone and Ponder, "well ... His/her back is really beautiful. So you step up, catch up and then look at him/her with your eyes (yes, you're married, but see there's nothing wrong with it!) and then you're extremely disappointed-he/she is not pretty at all. The above brief report is the feeling. It looks sexy, has web 2.0 style, and there's quite a bit of data out there, but unfortunately, there's not a bit of an enforceable strategy you'd like to get.

Why is that? These reports have good intentions, they cost a lot of hard work, and it may take months to organize the material together, but why is it a "crappy" report?

Here are some of the potential (but destructive) reasons why most of the essentials reports get away with giving advice to policymakers:

They leave the interpretation of the report to management (or client/questioner/other users who use the report). This is a fatal flaw because most briefings are highly aggregated for each KPI, ignoring all nuances and analysis.

Most management actually want real insights and executive advice, but they don't trust the people who do the research, so they'll ask them what the data is. So we faithfully put all this data in a very small number 3rd font (the smaller the size of the English font, the smaller the text, the translator) fill in a A4 size of paper, and then with the report to our report readers with a magnifying glass.

Most of the people who do the analysis live in a closed circle, gathering information, analyzing what's going on behind the data, and making recommendations based on the data is not something we do or encouraged by our management and our organizational structure. This has also led to the direct dumping of large amounts of data in the report.

Often, the creators of a brief report are "outsiders" (such as consultants or experts) who often do not have much experience to help them understand what is beneath the surface, as in the three cases mentioned above. This leads to the "non-practitioner" usually making some mistakes and making three types of summary reports.

If you want the management/customer to take action, then you have to give them information, not data. It takes a bit of effort to reach this point, and you should take advantage of your charisma (not to violate the human resources management guidelines, of course) and be patient.

The first step, please always remember the above three types of problems;

The second step is to get a workable solution from this blog.

My effort to solve this problem is to try to break through the philosophy of humanity: how can I create a "essentials report" to motivate site analysts to provide management with the information they need to make decisions?

Recommendation 1th: Remember the philosophy of reporting only for management's "few key figures": only 35 (five are already up to the top!) Determine the key metrics for your business success (Metric), and then remove any minor metrics.

Recommendation 2nd: Here is a small attempt at my "humble", even somewhat "ugly" "Essentials Report":

Each square represents a solution to the problem of poor reporting on each of the above human causes. (Sorry to cover some of the above text to protect some privacy) Let's take a look at the report.

First, at the top of the report (called the top, the translation should be "headline") that identifies the most critical metrics, and who is responsible for the business part of these metrics (obviously, each metric corresponds to a business), and who is responsible for interpreting the metrics. In this report, the Business section is responsible for the Sidney Song, the analysis of the value of the measurement data is Florian Pihs (translator: Oh, is not too narcissistic?).

You may also notice the red dots in the upper-right corner, which means the overall picture shown in this report is bad. It can be changed to two other colors, yellow--it means no need to fire anyone, but it's ready to fire (i.e. it's OK, but not ideal), green--it means giving someone a big hug and a box of chocolates (that's pretty good). And then......

The first square in the upper-left corner (that is, the chart) shows the trend of the key metrics and makes a very desirable subdivision (in this case, the "end up shopping" metric is counted according to 4 key customer classifications).

The purpose of this square is to satisfy the management's curiosity and show them that you know what you are doing. However, sooner or later this part will be forgotten (this is not a bad thing!).

The square in the upper right-hand corner, and the second box (key Trends & Insights, Critical trends and insights) is designed to add more value to the report by interpreting trends and the context associated with the data. The square explains what has gone up, what has been lowered, and has alerted the poor data. Your work really started here.

The contents of this box will be the most viewed by the management at the beginning, and after a while they will have confidence in you and will be glad that you have provided contextual information, but they are also going to forget this part (which is not a bad thing).

The third square of clockwise numbers, the square in the lower-right corner (Actions/steps to Take), is the driving force for shy web analysts to talk to marketers, site owners, VP, and everyone who knows the market and the situation, To help site analysts identify the root causes behind the data trend and to make credible recommendations for action. Web analytics are rarely able to do these things on their own. This requires contacting others, needing to talk, and means collaborating to identify the solution. This is an excellent opportunity to gain a keen understanding of the business.

This square is the key to making the summary report work. You will not leave the readings of data to others, or blame others, but to suggest that it is worthwhile to do so. Your management will thank you and this is part of what they will continue to focus on for some time to come. This part also means that the monthly meeting will not become a meeting to blame each other, but to decide who should do what. Amen!

The fourth square, which is the impact on Company/customer (impact on the customer/company) in the lower left corner, is in case the management doesn't quite understand why they should take action (let them listen to the idea of your poor old analyst). I feel like this is a missing part of any of the essentials reports, which are generally short of the final touches, and this box is to make up for it. It is the answer to "if this trend continues, how will it affect companies and customers". It also urges marketers/analysts to do some hard work to anticipate the impact of these trends and put them in the report.

This is a killer square, and if there are no other drivers for action, then the contents of this square are. It will tell us how much we lost, how many customers we missed and how many opportunities we wasted. Now if they ignore you again, they'll be asking for trouble. Believe me, your suggestion will be accepted.

See, it's that simple.

You make up for people's problems, make up for the flaws in the whole system, and you have a better understanding of the company's business (thanks to the third and fourth squares).

The rest is the constant victory.

After a while, you will certainly get more trust from the management level. All the crappy summary reports were gone and replaced by the following "Essentials report":

Now that you are asking the management to follow your advice in their own judgment and are helping the company to take action, who should study these figures? They're paying you. I want us to provide an executive opinion.

I would also like to stress that this is not going to happen overnight, but it is time to take steps to that end. May the strength be with you. Now it's your turn to share some of your experiences and opinions with us.

Related articles:

SNS report Download: Domestic and foreign SNS website Analysis report