Excellent Cloud Lao Wang (four) dry sum: Data brings the wonderful world

650) this.width=650; "src=" http://s2.51cto.com/wyfs02/M01/89/E9/wKioL1ghjaCwYPEYAABmzRlVYtY724.jpg "style=" height : auto;vertical-align:middle;border:0px; "title=" 1111.jpg "alt=" Wkiol1ghjacwypeyaabmzrlvyty724.jpg "/>

As a product manager, we don't just need business-sensitive intuition, we need data to drive our products to innovate! I listed the main four scenarios on how to look at the data.

Scene One: Look at the page click, understand what users are concerned about, where the card

Who to look at: Product manager and Interaction Designer

See what:

Because we have all the interfaces of the Web application, the interactive elements are buried, then we can see what the user has done in the page to do the click action. The first is the horizontal comparison of the entire page, the number of operands per page itself (number of redirects and times of opening). Then is the element to which the current page is to be concerned and the respective case, that is, the operand of the element. Of course, understanding the trend is not rare, it is better to compare the number of page operand and page element operand proportional relationship.

See what it is:

1) According to the number of clicks to determine whether the information layout is reasonable, you can also click on the user more than the part of the optimization.

2) Click More not necessarily good, click less not necessarily bad, please carefully analyze. Give a chestnut: A list page 80% of the user click to view the details, then some of the content should be displayed in the list, and we are in the details?

3) If it is a form page, then you can analyze which step the user is giving up or having difficulty with.

When to see:

1) want to the product function, information structure to optimize the time attention, perhaps can give you a little inspiration;

2) Find out the problem with the number of operations on the page, see if you can find the problem.

Scenario Two: Front-end performance issues and JS error troubleshooting

To whom: test engineer, Operations engineer, front-end development

See what:

Since we have collected the performance data of the browser and the JS exception thrown by the browser, we are concerned when we find that some operations are not satisfied with the sample or the number of errors occurred too. Look at the response time of the operation, the network, the server, the client's average time-consuming situation, the initial location bottleneck is where.

See what it is:

We can focus on their dissatisfied users or the wrong user exactly what is wrong, combined with the user's environment for comprehensive analysis, you can determine the approximate cause of the problem, most of the reasons are browser compatibility issues.

When to see:

1) The product has just on-line has a certain amount of time to see, the fastest detection of browser compatibility issues;

2) After the problem is resolved to observe a period of time, to see whether the user, the number of errors is declining trend.

Scenario Three: Identify deep usage users and potential users

Who to look at: Product Manager

What to see: This feature is best combined with access to real users, once linked to real users, is a living "people" instead of an anonymous visitor terminal. We can identify when the user first used the app, when, last time, and how often the last one months were used.

See what it is:

1) Can judge the user's stickiness by the frequency of access, and identify the user with deep usage;

2) Some users may be interested in a feature and can be converted into a deep user, although they are not used frequently.

When to see:

1) The product has just on-line has a certain amount of time to see, the fastest to find which users are the most worthy of attention of users;

2) Periodically look at the user's activity, identify the risk of loss.

Scenario Four: Understanding the experience of users in different geographies

Who to see: Operations Engineer, product manager

See what:

Take advantage of three maps, slow-operation geographic maps, error rate geographic maps, user geographic maps, and analysis of the trend graph of the operands (peak and trough of business)

See what it is:

1) Where the user access will be slower, operators do not give force? Should you consider a CDN?

2) Where the user is the most need to ensure the quality of service.

When to see:

OPS people can put a few maps on a large screen to monitor real-time

These are some of my experiences in Web application experience product summary. Welcome everyone to Exchange (Personal: uyunsoft).

Author Profile:

Wang Chuanlin

Superior Cloud Software Product Manager

Four-year UI designer, three-year business intelligence product, five-year it operation and maintenance products

Responsible for cloud Web application experience Monitoring products

This article is from the "excellent cloud dual-state Operations" blog, please be sure to keep this source http://uyunopss.blog.51cto.com/12240346/1870693

Excellent Cloud Lao Wang (four) dry sum: Data brings the wonderful world