Web analytics: website conversion and visitors conversion

The goal of most sites is not just to attract the attention of a random person, but also to instill in him a constant interest in the site itself and, more importantly, in the company's products, and then turn it into a source of income. I think many would agree that conversion can be considered one of the most important criteria for the effectiveness of a commercial site. But how many gestures have to do to get conversion data, even using advanced web analytics systems? And it would seem that all that is required from the analyst or the owner of the site is to define the “goals”, the achievements of which he expects from the site visitors. Or even simpler: the owner should be enough to once indicate to the system what to consider as a purchase - the main goal of a commercial site.

But users of various web analytics systems often have to perform a number of additional, albeit significant, actions:

• In more advanced analytical systems, there is a tool that tracks the achievement of goals, but as a goal you can often set only a visitor’s visit to a specific page, rather than performing a specific action. Therefore, it is necessary to add additional pages to mark the end of the chain leading to the goal (“Thank you for your purchase!”, Etc.).
• When working with less advanced systems, you have to segment the data yourself, choosing the necessary data according to the pages that correspond to the goal achievement: the number of visits (i.e., “purchases” made) or the number of unique visitors (various “buyers”).
• If there are several equivalent goals, we cannot simply summarize the number of visitors, otherwise we will take into account twice (if not more) those who made several purchases. Therefore, we again have to resort to additional tools - to segment, combine, define complex logical conditions ...

In principle, with a simple calculation of the conversion, the ratio of the amount of goods made through the site to the number of visitors to the site does not cause any special problems. But for this and special analytics tools do not need: it would be enough to install a regular counter on the site, and take the number of purchases directly from the site database. The only problem is not this: for example, when converting to 3% and 1000 visitors, we will not understand whether 30 people made 1 purchase, or 6 people purchased 5 products per month. Agree, the difference is great.

How could these problems be avoided

There are many options for improving the web analytics system. In my opinion, for a start it would be enough to cover the following key points:
1) To give the opportunity to set as a goal not a specific page, but a user's action - an event or a sequence of events. After all, the majority of purchases and similar goals often consist in simply pressing a button (“Order”, “Download”, “Subscribe”) or fill out a form and send it (again by pressing a button).
2) After reaching the goal, the visitor will assign him the corresponding status. I think it is logical to distinguish the simple visitor / reader from the buyer. Using not only the conversion of goals, but also the conversion of visitor statuses, we will immediately see how many people were able to get interested in and push to buy.

Naturally, the ranking system can be expanded, not limited to just 2 statuses. Consider the simplest strategy for an online store or a similar site:

1. Attracting users to the site: users get the status of a visitor or reader
2. Converting users to customers: assign the status of the Buyer or Client
3. Maintaining user interest and re-sales: we track multiple (made several purchases) or permanent (continuing to make orders, for example, at least once every six months) buyers.

With this approach, it is possible to estimate not only the conversion of goals, but also the conversion of visitors, which in some cases will be more accurate, because with a general calculation of the conversion, it is not entirely clear what causes a change in the level of sales: thanks to attracting new customers through advertising campaigns or due to the level of service, product quality and other factors that encourage customers to return and buy something else.
')
When used wisely, ranking by category will help build long-term relationships with customers. The list of possible statuses can be made even more, having thought over into which categories essential for running a business, it is necessary to divide site visitors. The main thing, breaking all visitors into groups, the user can see which of them brings real economic benefits. Thanks to the study of the behavior of each group and the pages they visit, it may be easier to understand what prevents visitors from making purchases, and what, conversely, pushes to order, and this will improve the conversion in the future.

On the one hand, the above ranking method is most suitable for various online stores. But on the other hand, it may well be useful to non-commercial sites. The only question is what is considered an "order" or "purchase." It all depends on the direction of your site, the options are actually a lot:

• subscribe to a newsletter or post new blog posts,
• check in,
• commenting or just reading the topics,
• Likes and other assessments
• census
• adding your content, text or graphic (photo, avatars, etc.) to the site,
• file download

and all the other activities you expect from your visitors.

One of the advanced ranking options is implemented in our project (who cares about ranking visitors in the Grapery web analytics system ), but that’s another story.