User research: An online survey of critical details

Introduction: When the online questionnaire survey, the site's traffic is a major indicator, but how to select the location of the questionnaire survey, users, delivery time related to the efficiency of the questionnaire survey

Online questionnaires are often considered the simplest and easiest method of user research, so some of the most important details are often overlooked. Share my personal experience below:

Can you get a sufficient number of questionnaires?

When it comes to doing online questionnaires, many people think directly about the problem design, the problem is design is a learning, but often no one to analyze the feasibility of online questionnaires. Speaking directly, can you get enough questionnaires when resources allow? A lot of sites are very important pages of the position is very hot, want to put a questionnaire entrance, it is best to have their own heart. For example, the questionnaire entrance occupies the place, the size, occupies the time. If the resource side tells you that I can give you an entrance to a research questionnaire at the bottom of the page, only give you 1 months, can you estimate how many questionnaires can be recycled, and estimate the number of questionnaires enough for you to analyze the data you want? At this point you need to have an important indicator: the page per Yi Yuan PV You can recycle how many questionnaires. Of course, if you have done similar statistics before, I believe you must have a spectrum in mind. I'm sure someone will ask, if I have not done online questionnaires on this page before, how can I foresee the recovery of the questionnaire? Here, I have listed a number of surveys to respond to the questionnaire, can be used for reference.

As far as my experience is concerned, the recovery of questionnaires will normally be between 1-5/million PV. Of course, the different types of Web sites, the different functions of the page, the style and location of the research entrance and other factors, will affect the recovery of the questionnaire. It is best to do a survey after each of the different pages of different research on the recycling situation, so that after doing similar research can easily predict the recovery of the questionnaire:

Number of questionnaires collected = Daily page pvx delivery days x Estimate questionnaire Recovery (how many questionnaires per Yi Yuan PV can be recycled)

Two, can you have enough time to put in?

Remember that we used to do an online questionnaire: Ask the user whether a feature meets their needs. The result was an ironic result. In the first survey, 70% of users thought the feature met their needs, and only 30% of the second survey users thought it met their needs, as the research issue and the portal were virtually unchanged. Two questionnaires were filled in tens of thousands of times, each time in the release of two weeks. These two surveys are very close to each other, and users of the site as a whole have not changed. Therefore, the proposal for important pages and functions of the questionnaire to have a certain time accumulation, preferably in more than one months. The different users of different sites have a certain period of time to visit the site, in the launch of the questionnaire is best to consider to ensure that your questionnaire in the target audience has sufficient exposure.

Do you subdivide the users who answered the questionnaire?

Different users will have different answers to the same question. For example, when you ask whether certain features meet the needs of your users, you should not forget to ask the users who filled out the questionnaires if they have used this feature. Whether it is a website or other product or service. There are always certain features that are used by only a subset of the users who would consider the feature useful. However, some users will comment on features that have not been used, and most of them think that function is useless. If your online questionnaires are just filled out by these users, you do not distinguish between users who fill out an online questionnaire and use this feature. Then, you will probably make the wrong decision, remove or weaken the function after analyzing the wrong data.

Don't forget we still have cookies.

With cookies, we can not only prevent "some active users" from repeating questionnaires. We can also use cookies to automatically record the user's relevant information to help you analyze the data. For example, in the research site search function, record the member ID, visited the site URL, search keywords and so on. If you find that a user has left a very low rating for your site's search function, you can easily know by the member ID what type of user is the person who filled out the questionnaire. The URLs and search keywords that he has visited can help you restore the maddening scenario that users encounter when they use the search function: Perhaps the user sees a search result that is not relevant to the keyword, or that there are too few search results or even search results.

Five, don't be lazy, daily tracking data

Doing an online survey does not mean putting the questionnaire online. It is best to add the access statistics code to the questionnaire. Then, see how many users open the questionnaire every day, how many users submitted a complete questionnaire, how many people did not fill out the questionnaire to fill out where people come from. Don't think it's superfluous. Because of this data analysis you can find a lot of problems ahead of time. I've met this kind of thing. I remember once a survey of online questionnaires published, I saw the next day data found that open the questionnaire few people, and the survey is mainly aimed at foreign users, the results found that the user IP access questionnaires are from the domestic. As a result, I immediately communicated with the relevant technicians and found that two U.S. servers were not deployed when the online questionnaire was deployed. Sweat, or take care of your questionnaires like a vegetable garden every day.