Web analytics: analyze it! Part 1. Introduction

Part 1. Introduction.
Part 2. Data collection.
Part 3. Basic metrics.
Part 4. From statistics to analytics
Part 5. Divide and think
Part 6. Action!
Attendance counters have been familiar to all Internet users for many years. Many site owners put several icons of different counters at once, and day after day the numbers on them grow and more and more delight the host.



Not everyone knows what information can be extracted from site statistics, except for numbers of visitors and, sometimes, sources of visitors. However, modern statistical systems provide much more opportunities for counting, processing and analyzing attendance data. They even began to call them differently: “web analytics tools”!
')
Now the majority of successful commercial sites pay great attention to the collection and analysis of statistical data. In particular, the largest international online store Amazon.com has created a cult of analytics. One of Amazon’s first employees, James Marcus, says:

“The time of inaccurate approximations and focus groups, anecdotal adjustments and smoke rising from the marketing department has passed. A company like Amazon could record every step of a visitor, every click and mouse movement, and did it.

As data fell into virtual heaps, hills, and mountain ranges, you could draw a wide variety of conclusions about this unknown creation — the consumer. In this sense, Amazon was not just a store, but also a colossal repository of facts. All we needed to do was connect the right formulas to these facts. ”

Advanced Russian online retailers also realized the importance of analyzing statistics: for example, Ozon.ru has an analytical department that analyzes and optimizes key indicators of the site. No wonder: as soon as it comes to efficiency, analytics are indispensable.

Thanks to the results of site statistics analysis, you can increase profits, choose the right direction of development or find new sources of customers. At that moment, when a novice web businessman is happy about the end of the website development, an experienced marketer is just starting a real job: analysis, optimization, testing, again analysis ... Still: analyzing user behavior on the website allows you to “squeeze” more out of the advertising budget, even more increasing the main indicator of the site: profitability, or return on investment.

Profit growth can be achieved through several factors. If you do not go into the psychology of decision-making, then any online seller must solve three problems in order for his business to be successful:

1. Attracting users to the site (advertising)
2. Converting users to customers (sales)
3. User retention (repeat sales).




These tasks are most often displayed in the form of a funnel. The analogy is popular, but not quite right: if nothing is lost in a regular funnel, the narrowing of the sales funnel symbolizes the “leakage” of users as they pass through it. Only a small proportion of users become customers.

The task of web analytics is to find opportunities:

1. reduce the cost of attracting a visitor to the funnel and increase its relevance.
2. “Extensions” of the funnel - an increase in the percentage of conversion of users in the client.

The result will be a decrease in the cost of the client!

How to achieve this? Well, let's go slowly, covering all stages of learning. I conditionally distinguish three such stages.

1. "School". Understanding the basic numbers, the difference between "unique" from "page views", knowledge of the concept of "referrer".
2. "Institutskaya". The ability to highlight problem pages and transitions on the site, "quality" and "low-quality" traffic sources.
3. "Business". Streamline business and decision making with analytics data. Testing, calculating ROI for each traffic source, e-merchandising. The ability to calculate the accuracy of the sample data.

We will deal with passing through all the stages in turn, while I wait in the comments for interesting, non-trivial questions, cases and tips :)