Engagement (participation) is a special measure, Avinash Engagement should not be a measure as the site analysis, it was more like do not want to look for an excuse to use in the measuring site is a true measure of He cited the the Engagement existence of several defects:

• Engagement is different for the different sites, it is not a uniform measure, it is difficult to define a uniform standard;
• Engagement is more defined, from a qualitative point of view to many users to participate in behavior is not well quantified, it is difficult to measure accurately;
• Not a criteria for a site user participation in the end how much is;
• Because there is no standard definition, there is not fixed, and therefore can not be used as the reference index to measure the site's overall performance.

Therefore, we can assume that the Engagement is a non-standardized measure, but bypassing those Myth of the Avinash mentioned, perhaps Engagement also not completely useless, at least it can be used as the site's own analysis of indicators to measure the user's activity.

How to reasonably define the Engagement

If you want to use the Engagement to measure the activity of users of the Site, and use it as a long-term indicator for trend analysis, then the first must determine the definition of those to be classified as Engagement behavior is relatively fixed in the short term will not occur changed. Such as for a forum, the behavior of the "post" and "thread" is the constant interaction behavior, which can be used as standards to identify whether the user participation; an activity button in the site or promotional links can not be used as identification Engagement standards, because the general promotion period, line of Engagement will be the corresponding reduced, which would lead to the Engagement instability, will be lost (or you just for the separate analysis of the significance of the analysis promotion of user participation, it is another matter).

Here to give two examples to show examples of e-commerce site definition Engagement:

• User's behavior;
• User into the shopping cart or purchase behavior;
• The user's pre-and post-sales feedback behavior;
• Evaluation of the users of goods.

And then look at microblogging can be defined for the Engagement of behavior:

• User registration act;
• Released a new microblogging or forward microblogging;
• Post comment the microblogging;
• Concerned about the new user.

Engagement of the definition of e-commerce sites and microblogging can see very different, which is why the Engagement is a non-standardized measure of the reasons the site can not compare their Engagement pros and cons of. So Engagement analysis indicators within the site, only used to measure site users active trend.

　　 Even if is a website Engagement also have a different definition of methods, such as I also can microblogging Engagement defined by adding "favorites micro-Bo" the action, so my suggestion is for the Engagement of the definition, a web site should at the outset on the formation of a standard, if the nature of the site no major changes should not be easily modified definition, or even evaluation as the site itself, it can not well play a role.

Should be noted that the Engagement is just the user a series of actions or behavior (Actions), not the site of the output results (Outcomes,). Can only be used for frequent measure of user activities, and then analyzes the recent level of activity of the user, and can not be used to measure the effectiveness of the site.

Based on the analysis of the Engagement Index

First recommended an article of Eric T. Peterson on Web Analytics, Demystified - How do you the calculate engagement.? , which provides on Engagement is very comprehensive and exciting analysis.

Analysis of the best examples, so here is my blog, for example, if I define my blog Engagement consists of the following behavior:

1. User comments;
2. Collection or to share my article;
3. Subscribe to my blog or concerns me;
4. Residence time (Time on site) over 3 minutes.

First relates to the data obtained, in fact, through click-stream + Event tracking the behavior of the majority of users can still track to get here is no longer launched in detail.

Next is how to measure Engagement. Measure Engagement indicators is called the participation index (Engagement Index) Now that we have defined Engagement, we can calculate the Engagement the Index through a number of ways. It is recommended that called the Session to Visit (or) as a unit, the first visit whether Engaged, rather than to calculate the number of times each Visit Engage, because we can not be certain behaviors precise quantitative number of times, perhaps to Post Comments can calculate the number of but the length of stay will not be able to use the number to measure, so here is unified only judge a Visit any of a class of behavior occurs that the visit is Engaged, we can use some image of the code to explain:

  boolean isEngaged (Visit) {
 if (comment | | share | | feed | | follow | | time_on_site> = 3min)
 return true;
 else
 return false;} 

Call the above function, "We can get to what Visits Engaged Visits can be calculated out of the total number of Engaged Visits Engagement Index value. But simply an Engagement Index does not seem to contrast value we can through the user's participation rate (Engagement Rate) is calculated with the ratio of the total site Visits:

The engagement Rate = Engaged Visits / All Visits

Can also use the following diagram to compare the relationship of Engaged Visits Visits:

Can also analyze the ratio of the types of conduct in the definition of Engagement:

To note here is the sum of Visits in this pie chart is not calculated above Engagement Index, a Visit may both commented on, and also subscribe to the blog, then this Visit will duplicate records to the ratio of the two types of behavior. Eric T. Peterson's Engagement breakdown of sources of analysis recommended above article has been described in great detail, interested friends can see.

If you feel that is also intended to still better actually, that user participation is not as important as, and perhaps those who subscribe to the users of the blog or in the blog comments the user should become more active, you can give different weights to these acts, Sidney related to the introduction of the blog - the most basic of web analytics metrics (8) - Engagement , if you want the weight is more objective and accurate, you can use the Analytic Hierarchy Process (AHP) before me had to calculate each behavior of the weight of the Engagement the Index is the corresponding weighted (as defined in each Engaged Visits the number of E _i, the weight of each Engagement behavior of W _i):

Analyze the Engagement of significance

Sometimes we need to be subdivided, and sometimes we need to be indicators of aggregate that to evaluate the overall performance. It is recommended that the Do not try to calculate each Visit Engagement Index is the frequency of use, number of times due to changes in the site design (page layout, interactive button design changes), and behavior to achieve diversity (if you want to spring microblogging, that will be achieved through too much) and showing the unstable state of Engagement Visits to measure more comparable, the most in total we only aggregate to measure the entire site of Engagement.

So Engagement for web sites, in the end what's the use? There is a simple example, when an e-commerce website orders in recent days there has not been a significant change, but the user activity had increased sharply, because of a product quality problem, a large number of users to the site Feedback and complaints, then if the site is monitored by analysis of Engagement with the user's active, it can quickly detect and respond to. In fact, this example also shows that for some sites Engagement The higher performance of the site are not necessarily better, there is no direct link between the two.

The WEB2.0 has entered the era, between users and Web sites to interact more frequently, the site requires more active users to increase its own vitality, and user involvement to enhance information exchange between the sites and users for the site with dynamic at the same time to create for the site with more information, many of which are site itself is ignored or can not provide valuable information, while the injection of these "wisdom" of the user is no doubt for the site to bring more attention.

»In this paper, the BY-NC-SA agreement, reproduced please specify source: Website data » to measure user activity "Engagement"

Related Articles:

1. Site analysis of the basic measure
2. Doubt about the definition of Bounce in Rate
3. Measure and subdivision of the site pages
4. Comparative analysis based on user segments
5. Web site of the old and new users to analyze

• Previous: optimize web site navigation design
• Next: Optimizing content recommended