Traditional product research usually needs to select user samples first, and then spend a lot of manpower and material resources and adopt different research methods to conduct user research. If you apply big data to user research and have a large number of historical data samples, you can pre-analyze the research problems with the help of big data, and then intervene manually, which can not only improve the efficiency of user research, but also respond to user needs as quickly as possible, greatly reducing the cost of user research. Based on this, this paper attempts to interpret the new changes in user research in the era of big data with big data thinking.
Note: What this article provides is only the idea of user research in the era of big data. If you have specific user research needs, please feel free to ask me, and I will discuss specific cases in the next tweet.
As a means of production, big data is affecting human society more and more deeply. Now in the field of e-commerce, big data recommends products that are more in line with the personal preferences of similar consumers. This recommendation method not only saves the time for consumers to find products, but also improves the revenue of e-commerce platform.
Similarly, in music, TV series, movies, advertising, user research and other fields. Big data can be used more and more widely. So, what changes have the era of big data brought to users' research methods?
Before big data is widely used, traditional user survey methods usually need to go through five steps: defining survey questions, making survey plans, integrating survey methods, designing questionnaires and summarizing survey results.
However, after big data is widely used, there are a large number of historical data samples. We can use a variety of public big data tools for early analysis and processing, and then conduct manual selective intervention to compare the two and conduct multiple rounds of tests to help product personnel find the truth of the problem.
Set up excellent research questions first, and the research will be half successful.
Setting the research question is in the first link of the whole research, and its importance is self-evident. For example, some product managers may ask "Why don't users accept video payment" or "Are there enough users willing to pay 15 yuan/month to watch genuine HD videos, if it is lower or higher?" The former's research questions are too broad, while the latter's research questions are too single.
If the research question is defined as:
What kind of users are most likely to use the paid services of video websites? How many users will be willing to pay the price of different stalls on the video website? Of all the video websites, how many users will choose this video website because of this service? What is the value of this method compared with video payment, such as advertiser sponsorship? Of course, not all research contents can be so specific and clear:
Some of them belong to exploratory research, the purpose of which is to find out the truth of the problem and put forward possible answers or new ideas;
Some of them belong to descriptive research, focusing on describing some quantitative characteristics of the project content;
There are also some causal studies, the purpose of which is to detect whether there is a causal relationship between phenomena.
Second, according to the research question, the charm of pre-analysis and processing of big data lies in collecting all data instead of sample data. For example, Didi's takeaway service and Meituan's taxi service are all launched. Thanks to the development of modern social networks, Didi and Meituan can almost make statistical analysis on the comments of newly launched services on social media such as Weibo and WeChat, so as to provide better services.
For example, you can learn about the search behavior of netizens on this service through Baidu Index, and at the same time conduct tracking analysis:
Of course, not all netizens will use Baidu search, and they may also use 360 search. At this time, we must use the 360 index:
Or users use social media such as Weibo and WeChat to express their emotions and thoughts, and may use Sina micro-public opinion to conduct word-of-mouth analysis and text mining by using Weibo Index, third-party public opinion monitoring and word-of-mouth analysis tools:
Note: The above big data tools only list three commonly used ones. In practice, the choice of big data tools needs to be determined according to the specific research problems of users.
Third, manual intervention, targeted treatment of research problems.
According to the results of big data analysis, we can manually intervene in research problems and conduct targeted research and processing. At this time, traditional research methods can be adopted. But different from the past, when adopting these research methods, you don't need to spend a lot of money on various studies. The purpose of choosing manual intervention is to feel the investigation process more truly and participate in the handling of investigation problems.
Traditional research methods usually have the following four ways:
1 observation method
This method is to observe the use of products by consumers in an unobtrusive way and collect the latest data. Some strategic consulting companies believe in observation when doing research.
The following is a fragment of Hua Hehua, a well-known domestic marketing consulting company, applying this method in Super Symbol is Super Creativity. Find out:
"For example, if you observe the consumption of toothpaste in the supermarket and observe the people walking to the toothpaste shelf, you will see a process: a customer walks past with a shopping cart and browses the toothpaste on the shelf while walking; Stop and stare at a box of toothpaste for a while, then move on; Stop, pick up a box of toothpaste and put it down at a glance; Pick up another box, turn it over, read the package carefully, and put the back copy back on the bookshelf; Take two steps forward, turn back to the box of toothpaste you first noticed, take a closer look, and put the one behind the package back on the shelf; Hurry back, the box of toothpaste seen in the fourth step is still in the shopping cart, and choose to end. "
"No, it's not over yet. He may come back later, put the toothpaste he just put in the shopping cart back on the shelf, and replace it with the box he noticed in the second step, or he may want both boxes. In this way, you can observe the whole process of buying toothpaste, and there are actually seven actions. "
2. Focus group interview method
This is a research method based on demographic characteristics, psychostatistical characteristics and other factors. Six to ten people are carefully recruited, and then they are gathered together within a specified time to discuss with these participants. Participants usually get some compensation.
Researchers usually sit in the room next to the forum and observe the discussion process in front of a mirror. It must be noted that in real-time focus group interviews, participants must feel as relaxed as possible and try to get them to tell the truth.
3. Behavior data analysis method
Users' behavior when using products can be used to observe users' psychology, and researchers can learn a lot about users by analyzing these data.
The user's browsing time and browsing content can reflect the user's actual preferences, which is more reliable than some statements provided by users orally to researchers.
4. Experimental method
Excluding all factors that may affect the observation results, we can get the real causal relationship between phenomena.
For example, video websites provide users with high-definition video services. They only accept 25 yuan for one month in the first quarter and 15 yuan for one month in the second quarter. If there is no difference between users who use the service after charging at two different prices, then the video website can't draw the conclusion that higher service fees will significantly affect users' willingness to watch paid videos.
Fourthly, after the research method is determined, the design of the questionnaire can be started.
The purpose of setting up a questionnaire is to collect first-hand information. However, it is necessary to test and adjust the questions in the questionnaire because the format, order and order of the questions in the questionnaire all affect the answering effect.
Matters needing attention in questionnaire design:
Verb (abbreviation for verb) summarizes the survey results.
Compare the results of statistical pre-analysis of big data with those of product researchers, so as to turn data and information into findings and suggestions.
Finally, you finished. According to the results of market research, you can make specific marketing decisions.
Note: In this process, the traditional investigation method does not need to consume a lot of manpower and material resources. For suspicious results, multiple rounds of tests can be conducted by controlling variables to help product personnel really discover the truth of the investigation problem.