1, do the design why still need to look at the data?
Many designers never look at the data, either because there is no data to look at, or do not want to see, but also the same design to do well! Design is a sentimental side, why do we have to relate to the data? Let us take a look at the nature of the design first. Design is different from pure art, art stems from the artist's observation and reflection of reality, as well as the self-expression of such observations and reflections; Design is born to do things for others, even if the same need to observe and think, but this observation and thinking is not to represent the designer's ego, but to better serve a user group, So it becomes very important for the designer to understand the user. In particular, to understand the user's goals, behavior, attitude and other related situations, we say here the data is actually the user's goal, behavior, attitude and so on quantification, therefore, through the analysis of these data, we can better tap the needs of users, and thus provide users with a better experience.
Simply put, the design is to serve the user, understand the user to better do the design, data is a way to understand users.
2, what is the role of data in the project?
To understand this role, we first return to the designer to see the data of the main scene, summed up there are no two categories: one is because of the project needs, through the data of the argument, so that the design go more calmly, reasonable, not the designer of their own yy; the other is the daily monitoring needs of their own products, Always know how many people are using it, how it is used, and whether the user's behavior and expectations are consistent. That is to understand your design is used in the situation, otherwise how do you know the design is good, is not achieved the design goals, is not really help users solve the problem.
First, analyze the scenario for the data in the project. Almost all of the design process can be used in the data, summed up in this process can be cut into three parts:
Pre-design data to help you find the problem: all design before the beginning of the research and analysis, are to more clearly the needs of users, why do you want to do this design? From a business point of view, what is the value of this product to the company, the design to achieve what purpose; from the user's point of view, what is the value of this product to the user, This design for the user to solve what problems, in understanding the business demands and users of the process, we will inevitably use data, this stage, the role of data is to "find the problem" to see what the design can solve the problem, thus better clear design goals.
Of course, the specific work, most designers are more entangled, not only to consider the business requirements, but also consider the user demand, if the two can not match the time, we should do, is the sum of the two? or we will only consider the user demand, the business appeal to see the line. My personal understanding is that in the real work we are not in pursuit of the most perfect design, more is in the balance, if it is a user-oriented products, such as biased to provide users with a function of the platform, itself is completely from the user's point of view, by providing users with the function to help users solve the problem, Should be more close to the user's appeal; if it's a commercial product, for example, in favor of providing users with some content of the platform, then provide users with the initiative to find the entrance at the same time, you can moderate to the business development needs of tilt, do a moderate level of business guidance; Of course, this is not absolute, often the same platform, The same product, in different stages of development also have different needs, if it is a brand new product, business survival becomes particularly important, this time the design should be more consideration of business requirements, first to help the business survive, otherwise, this product will be hung, how to provide services for users?
Of course, good designers can always find a clever balance between the business and the user, find the intersection between the two, for example, if the product this stage is to do the user scale, and the user is seeking to enjoy personalized service, seemingly completely unrelated to the two demands, In fact, we can improve the user satisfaction through better personalized service, get good word-of-mouth, and then indirectly with the use of user Word-of-mouth to enhance the user size of the product, the two are not completely irrelevant, more time to see if they can find their relevance, seize the phased design goals.
How do you use data to identify problems with a specific example? The data represents the user's goal, behavior, and attitude, but looking at a number alone is no way to find the problem, the comparison of data is the simplest and most effective means. We know that trading relationship buyer's transaction is very important to 1688 websites, we want to improve the trading experience of trading relationship buyers, but we don't know where to start, so we do a lot of data analysis. Trading relationship How do buyers find old sellers? What are the conversion rates for different paths? What is the difference between different user search methods and conversion rates?
First, through the breakdown of the user group, we found that the transaction relationship buyer through the search payment order conversion rate is the search overall payment order conversion rate of twice times. Therefore, in search results to increase the old buyers label, easy to find old sellers.
In addition, we also found that the average member, 1-2 star membership level, is to enhance the trading relationship of key users.
Through the above data analysis, we have found some of the main problems, around these problems, followed by the optimization program.
Design data to help you determine the way: because the designer's personal experience is different, creative thinking is different, so different designers face the same problem, the solution may be very different, even if the same designer will think of different solutions, in the end which is more appropriate, in some cases the data can give you reference opinions, Provide you with "judgment ideas" to help you make decisions; All roads lead to Rome, but which is the most suitable one?
Through a concrete example to see how to use the data to judge ideas? There is a wholesale class of the Electric Dealer website (1688.com) channel Home (ye.1688.com), we found that the user's conversion rate is very low, went to study the data, and then combined with the typical user to do the conclusion of user interviews, Finally, the reason for the bottom of the conversion rate is very simple, the first page of this channel is mainly from the home page of the entire site, and the entire site's homepage is a whole industry category page, if the user is the women's industry buyers, she from a whole category of the home page click a link into another whole category of pages, Again difficult to find women this category, and then click into the list page to see the product, this path is very deep, then how to solve this problem? That is to avoid doing women's clothing users from the home page into the channel and then again to select Women's category, to see Women's goods!
What are the ways to solve this problem? Can be added to the homepage of the home page, so that users directly click on the female category to enter the home page, to show the user the women's goods, the user can enter the channel home, according to the industry preferences personalized data to recommend goods, the recommended inaccurate, users can also be customized; which is more reliable? Two ideas have advantages and disadvantages, in view of the former need to have external dependencies, to change the home page, so we are very much in the hope that the latter a thought can run through, but how to know this train of thought OK? First we need to know the industry's personalized recommendations can cover the size of the crowd, and how many people are willing to customize the industry preferences?
This may not be a clear question for a typical web site, But 1688.com is a member of the user has long been billion of the B-class electric business site, with such a large user scale, high user coverage, which means the accumulation of user behavior data, moreover, the B class of users have a significant feature is in a longer period of time, the industry's preference is relatively stable, if it is a main women's buyers, then she Preferences will generally be dominated by women, will not exceed the scope of clothing, there will be a small amount of clothing surrounding the procurement of matching.
As shown above, through the industry preference of personalized algorithm, we tracked a period of time to visit the Channel Home page (ye.1688.com) user data, we found that about 2/3 of the user is a very clear industry preferences, then this basic can be determined to do industry preferences personalized recommendation is reliable! But the remaining 1 /3 are users willing to customize industry preferences? We were unable to determine whether or not they were willing to customize preferences for the time being, but through a questionnaire survey of the entire user base, about 30% of the users said that custom industry preferences were a good service, based on these situations, We determine that personalized recommendations based on industry preferences can solve most of the user's industry preference problems and enhance the relevance of the content. When the program finally came online, about 10% of the people actually found custom portals and created custom behavior, and 70% didn't have to customize it to achieve the default, accurate recommendation.
After the design of the data to help you verify the solution: Our design program in the end to do good? The measure is to see if the design can achieve the design goal? This also requires data to quantify, usually using a GSM model to support design validation. G (Goal) design objectives, S (Signal) phenomenon signal, M (Metric) measurement index, the so-called design goal, is to determine what the design to achieve, what to solve the problem; metrics, we cannot imagine, must be based on the design objectives, the first assumption that the design objectives will be achieved, So what happens or the signal? List all the phenomena or signals, choose what we can monitor, and then quantify the phenomenon or signal product, naturally get the measure, but the amplitude of the index often depends on experience.
For example, a product design goal is through the design of guidance, so that more buyers to create a purchase, imagine, if the design objectives to achieve, what will be the phenomenon? There may be more people willing to buy, look at the Product Details page, click on the Purchase button and so on, and finally produced a purchase, then, what is the measure? Design just changed the presentation of commodity information, and can not change the quality of the product itself or behind the service, so we should focus on whether the design to strengthen the guidance, enhance the purchase will, whether to stimulate the user to understand the behavior, mainly refers to browsing behavior, The most typical is to reach the Product List page or the Product Details page, and so on, the quantitative result is to see the proportion of users to further behavior;
Take a concrete example to see how you can use the data to verify that your design is achieving your design goals. There was a functional module to find the origin, we in the design before the investigation, the user told us they need to find the origin, and more accustomed to using maps to find the origin, we are ecstatic, in this direction made a direct floor of origin, we firmly believe that the user told us is certainly right! But is such a design really able to help users find the place of origin efficiently? Look at the following data analysis.
Is the user's goal not to find a place of origin? Also tell us to use the map to find a place of origin is very consistent with their habits? Why do users not use this section when online??? I saw this data is very unexpected, the moment between the head, and then went to see the heat of the plate, suddenly dawned. Through data analysis, the map, even in line with user habits, but only so small the map to carry out such a complex operation, its efficiency is very below, so the map to find the origin of the function to retain, but not as the default way, using the popular, regional, near, searchable, map of the way integrated load, Finally achieved a good result!
3, how to use the data to do day-to-day monitoring?
作为一个设计师，你的作品上线后，有多少人用?这些用户是谁?有什么特征?用户具体是怎么在使用你的产品的?你的设计是否还有优化的空间?如何才能为用户打造更好的使用体验?怎么才能知道这些数据好不好，有没有问题呢?主要是靠比较、 Rely on experience, rely on this product long-term follow-up generated intuition, only in this product is very familiar with the premise, you will be able to give a more reliable interpretation of data changes.
The main way to find problems in daily monitoring is to compare data, but how to make a specific comparison? There are three most commonly used simplest contrast methods: A, horizontal comparison, and similar products to compare, look at the relative situation, and then speculate whether there are problems; B, vertical comparison, and their own past comparisons, To see whether some inspiration can be obtained from the historical law of development, mainly to see their own trend of change; C, user segmentation, this is the user according to different analysis needs, split to look at the data, to see the differences between the groups, there is no part of the user and other users to show different behavior, And then find out where the problem lies. Of course, in addition to these three commonly used comparisons, we can also do some matching qualitative research, and then make the problem more thorough. Some statistical tools can also play a role, such as using SPSS to do data factor analysis, clustering analysis, etc., can also have some unexpected harvest.
4, the data is not the core value, you are!
Say so much, I do not want to emphasize how the data is omnipotent, but in the Internet domain, any a certain user's product, you have to go to understand the data, some of these data is macroscopic, as a designer we can as a background knowledge, should go to understand, But designers are more concerned with the user's goals, behavior and attitude and other related data, attention to those micro, and users, and design plans closely related to data, so as to better understand our users, understand the user of our design program feedback, to help us better play their own value!