To Internet entrepreneurs: Please don't think of yourself as a typical user

Intermediary transaction SEO diagnosis Taobao guest Cloud host technology Hall

Today, looking at the latest mobile developer report from Vision Mobile, I was surprised by the fact that only 24% of developers were conducting market research, and more and more developers were choosing to believe in their own judgments, and "away from users" seemed to be the main theme of product decisions in this fast-moving era.

I'll tell you two stories first.

Doug Bowman, Google's chief visual designer, announced his departure from Google in March 2009, and wrote a goodbye,google that slammed the process of Google's data-led product design, which addressed a case- Google tested 41 kinds of blue, and finally chose one of them based on feedback data, rather than allowing designers to make choices based on their own judgment, Doug Bowman that this approach is very rigid. Later, I spoke to the user experience Director of Innovation Workshop, Wu Zhohao, as Doug Bowman's former colleague at Google, who said that in this case he was more in agreement with Google's approach Google products serve hundreds of millions of users, They use different screens, operating systems, and browsers in different environments, Doug Bowman the best Blue effects in the eyes of many of them, using only the test data to find the most common and appropriate color.

The second story is also about Google, some time ago, Google announced the closing of Google Reader, a person claimed that Google's spirit is missing, began to "discard users", because he is Google Reader's heavy users. But in fact, the opposite is true, because most users have abandoned Google reader to make Google's decision: After Google announced the decision, many users choose to import RSS feeds into other products, delicious love read the imported data to compare, found that there are about The 83% source has not been updated since 2010, and even about 24% of its sources have been permanently invalidated. The nostalgic, carefully organized list of RSS feeds is already a huge and desolate graveyard--which has led to Google's killer of Google Reader.

From the above two stories, we can see that user research is still very important in product design, because an Internet practitioner can almost never become a typical user of a common domain.

For example, when they say they have never used a browser to do something on the mobile side, they don't see that the proportion of users reading news through a mobile browser has exceeded the mobile app, and the ratio of using a search engine to a Web page has been above 80%; For example, they say they haven't watched TV for years, To the 2012 single year shipments have broken 50 million of the smart television market behind the head ...

Of course, this does not mean that the data is the only standard to drive the product, but it can provide a reference for your product design from a relatively objective angle, in fact, we should consider from the qualitative and quantitative two angles.

First of all, I think this approach is ideal for use in the early stages of a product, especially for teams that do not have enough user capacity to conduct data research. The easiest way to do this is to pick a small number of your target users, get them to use the product or ask questions, watch their responses, and listen to their feedback. It should be noted that any problem should be concise, detailed and can be answered.

In quantitative analysis, in addition to market data, the most commonly used in product design is a/b test and 1/100 test. A/b test is the deployment of two or more concurrent scenarios for a product to different users, depending on their feedback to determine the best one. In A/b test, there can only be one single variable between each scenario, or it cannot be judged, including time consistency--several scenarios need to be carried out simultaneously. 1/100 test is to deploy another version of the product to a small number of users (often the core of the user), by iterating and diffusing the feedback, the 1/100 test iterations are faster and more suitable for the start-up team than the A/b test, but it is more intuitive to be effective.

As you can see, both A/b and 1/100 tests require a larger user base, which may seem difficult for startups, but there are a variety of large test platforms that can help you achieve this. Of course, in addition to testing, there are now a lot of statistical platform and analysis platform for your products to provide data reference. As for how to grasp and measure the balance between qualitative and quantitative, I will invite some entrepreneurs to share a detailed case to illustrate, and also welcome you to participate in the discussion.

Finally, what I want to say is that not treating yourself like a typical user doesn't mean you need those typical users to tell you how to do it, but to decide what you're going to do based on their feedback, as the usability expert Jakob Nielsen says--don ' t listen to Users,watch Users work.

Note: Pictures from the network