Product Design Suggestion: The analysis suggestion to the bean sauce guessing product

Article Description: give "watercress guess" suggestion.

Spring Festival Eve, the new version of Watercress home-watercress guess on the line, instead of the original neighbor broadcast home. Its title is: "XX, you may like", recommend a package of popular books/videos/audio/topic/photo albums/events and so on.

The big moves of watercress again sparked hot debate. In the author's opinion, watercress guess there is a problem of ambiguous user orientation and product target.

Do users need hot recommendations? They have been in the watercress for many years, familiar with the line products, attention to hundreds, often hung watercress, and habitually in the broadcast page according to F5, after a long accumulation and interaction, has already taken the initiative to pay attention to the dry information producers and communicators. So the hot spot they recommend is often a belated suspicion: the more recommended, the more likely they are to have seen it. The problem is that the name "watercress" will allow users to produce highly intelligent and personalized expectations of the site. If it recommends what it already knows, the user will question the system's understanding and concern for itself, and without the information that the user has already identified, there are not so many new and appetizing things to recommend, because the content they are interested in can always be obtained from the neighbor broadcast the first time.

What people need watercress to guess most?

Newcomers Novice users, do not know what the watercress, how to play. Countless novice faced the dilemma (Novice's call to see this), they heard that others here to play the joy, their own to register but still not their door. I can only reluctantly to the empty home broadcast doubt: those people say good things are hidden in?

To the current neighbor broadcast not very satisfied, when they pay attention to enough information source users this situation can often be greatly alleviated. This includes some of the novelty seekers, but most of them can still be seen as growing novice users.

Single product users: by friends or search engine brought in, only familiar with the use of one or two of watercress products, other products for him are unknown mainland (for example, can only play watercress community but do not know watercress reading or vice versa). This is also a novice user of a long or short transition state, originally with their attention to the type and number of users to increase, slowly also may see the other continent's side corners.

By the information of the ocean drowned veteran: Pay attention to hundreds of thousands of neighbors, to be a radio to see too spend time to see, random extraction afraid of losing the watermelon penny wise, do not look afraid to miss what. So want to know your taste to help you select a few, save energy and avoid missing the key. Such a relatively low number of users, and their dependence on watercress is also heavier, even if you do not have to please this method will not leave. -What they need most is effective active information management, such as the classification function of neighboring broadcast: classify the neighbors, classify the broadcast, and add tags to the collection (now "like"), classify, sort, search and so on.

Among these categories of users, the most urgent concern is the first category of completely novice users. One is that if they do not come to the site will be difficult to add fresh blood, expand the scale; second, when the needs of complete novice users, the growth of novice users (including single product users) and the need to find users can also be met. The hot recommendation of home page is a good opportunity for novice users to test water, to expand their horizons to a single product user.

How do you hear the voices of novice users?

Do not use the depth user experience as a test basis, nor rely on site search or Word-of-mouth to get feedback. Because often the depth of users have the ability and influence of their own voice spread out, the most attention to the novice user's voice is often easy to drown. In the face of a complete stranger, when we do not know the site, we often do not say that they will not use and so spit slot, but doubt their ability problems, but also embarrassed to ask for help, after a wall of habitual Zang result is the first time to vote away with their feet. May not even understand the name of this thing is called watercress guess or call home-new version of the home page? What's the old version? --How do you express your opinion?

Please do user tests for novice users who have never been exposed to watercress. It's not a contrast experiment with the old two version, because it's obvious that a new home page with a variety of recommended content looks more approachable than an empty neighborhood broadcast from scratch. But the new version of how the algorithm is better experiment. However, in the process of adjusting the algorithm and the user's interaction, the situation may change every day, it is not realistic to find someone to test at any time.

The only practical method at this stage is to analyze the data directly from the browsing. In view of the watercress and neighboring broadcast home war, you can use the broadcast quantity to define a novice user. For example: Take the initiative to focus on fewer than 20 neighbors, less than 150 broadcasts a week, a shorter registration time of one months, and a total login of less than 10 times (these numbers are just examples, and the specific numbers are carefully analyzed according to the user's data characteristics). Given the low preference information left by novice users, it is also very difficult to make accurate personalized recommendations for them. Want to verify the success of the product (please the novice user), you can see whether the following indicators are significantly improved: the conversion rate of active registration of anonymous users, the new registered users within one months of the average return visits, the new registered users within one months of the average number of items, PV, active time and so on For a single product users, you can see them within one months of other products add PV, the number of items such as collection of indicators to measure the expansion effect.

When watercress is positioned as a product for novice users, allow users to choose their own home page is set into a watercress guess or neighbor broadcast, can effectively reconcile the novice user and the depth of user demand contradictions (for watercress guess on the line also accompanied by neighboring broadcast of the revision, the current "Back to the old version" is not equivalent to the choice of home page function).

2. 1 The difference between the watercress and the classified products

Watercress guess carried out the watercress has been found and shared ideas, and in the traditional book audio and video based on the new community and activities of the product recommendations. "Home watercress guess is a variety of types of mix, if you want to see a special category of things, reading, film, nine points and other services in their respective watercress guess will be more in-depth." ”

Can be careful users will find: Classified products (such as reading) in the "watercress guess" and home "watercress guess" the content of the recommendation is completely different. Why should there be two different recommendations for results? If one of them is more accurate, why not just replace the other one?

Because they have different goals and different positions. Home should be simple and quick to make people understand what the site is, what, and the classification of products suitable for more in-depth, specific, personalized more strong recommendations. With the complex of watercress products, information architecture is becoming more and more difficult to understand, blocking a lot of people want to come in to play the new and some not keep up with the changes of the old users, may also artificially create more single product users. At this time, a home page integrated watercress core products of hot recommendations, to help novice users to explore, quickly understand the whole site of watercress guess came into being. This positioning is different, is the home page watercress guess with other watercress guess the biggest difference.

2. 2 How to attract new users

Since the positioning of the home page is to attract and cultivate new users, then how to play the role of watercress to achieve the goal?

Although the watercress community has a high standard of users and original support, but the face of BBS, SNS, light blog Fierce competition. Although the status of the book audio-visual commentary is leading, but in the online viewing and free download flooded network environment, coupled with "want to see, is watching, see" the words, novice users are easy to misunderstand it is also like starting point, potatoes, songtaste can immediately consume products. So when the watercress like now in the home recommended the latest hottest book audio and video products, new users tend to be filled with hope that a little open can read online, listen to music online or directly to the movie, but when the point is very confused how to find online to watch. Disappointed, probably will not come back again.

How to keep novice users and create repeat customers? In this fast-food era, the user's patience is particularly rare, a short period of time can not see the target will be quickly transferred, even if the use of purchase books and other products online to buy books also need to wait a day or two, let alone a lot of resources are very difficult to buy, even if the internet search This frustration to find the process is very easy to kill the enthusiasm of people, this can not eat the bitterness and sorrow may be lost on the novice users more frustrated, lose the enthusiasm for the exploration of watercress. At this time to give novice users a little sweetness, it is easy to leave people, watercress can fully use the existing resources, put some can immediately consume content. Let's say the book is a probation chapter book, New products in watercress Reader in the ready-made article, there is a ready-made "online watch" link to the film TV series, through the watercress radio to listen to the album, as well as the rapid consumption of watercress house made: Watercress Community topics, albums, diaries, small stations, activities and so on. At the same time, the recommended frequency of updates can also be lowered. On the one hand, this kind of content will take a long time to consume, the total amount of resources is also relatively limited, on the other hand, novice users daily login and refresh rate is not too high, so as long as the update one to two times a day is enough.

Worry about the number of recommended home page too few users see it? Is the recommendation for immediate consumer goods not comprehensive, trendy? Don't forget our goal is not to allow new users to linger on the home page, but to attract users from the peach Blossom Place to start, explore the whole station of watercress. So don't put too much content on the home page to feed the user directly, but leave a little desire to explore, so that "more" links to lead them into a broader world, such as the Product home page (for not landing users) or the product inside the watercress guess (logged in user).

As a result, it can give an unregistered user a taste of the benefits, become a repeat customer, help novice users get started quickly, but also convenient for a single product users to see other products, novelty users can also use this for a change of taste;

2. 3 developing new users into the community

After all, things rather than people-centric networks are slightly indifferent, and then lack of viscosity. If the new user each time only through the homepage recommended direct access to fast consumer goods and satisfied with the reward, a long time, not only can not help users into the community, and even the formation of hidden obstacles. To watercress this to a high standard of users for real resources, user interaction for the vitality of the site, long-term look, and watercress said together with the disappearance of "watercress guess you are interested in" and watercress community's important support: "Watercress Group", is still recommended a useful supplement home.

The user will invoke the experience under the guidance of the copy and the interface, build up the mental model, and speculate the logic behind it according to the display and feedback of the system, revise and perfect the expectation. The closer the product's performance model is to the user's mental model, the easier it is for users to understand and use it. A good performance model can help us predict the effects of action or the use of rules, and vice versa can mislead users into setting up incorrect mental models and thus frustrate expectations. Customer satisfaction is achieved by comparing the actual effect of the product with the expectation of the user. Therefore, the success or failure of a product is determined not only by the product itself, but also by the user expectations it evokes.

According to our users and product positioning, first from the name of the product and copy to start, adjust it to people's psychological expectations.

As mentioned earlier, this product is mainly used in the face of novice users, with a fast-consuming resources to attract them to register, get started, and let the depth of the user occasionally use it to pass the time. Because the novice user personalized recommendation can not be very accurate to the depth of the user unknown personalized hot spot, always have the means of luck, so you can be named "Casual look." Like shopping malls, do not know what this shop or have no specific goals, in the face of enthusiastic service staff, we will answer a "casual look." Like Google's "good luck", this relaxed tone and positioning, will not allow users to build unrealistic expectations, but also in line with the actual product positioning.

Watercress guess this combination of recommendations presents a considerable challenge to the UI. Because of different types of product resources, how to harmonize them harmoniously in the same interface? At present, watercress guess this kind of improved waterfall flow design is a very good innovation.

However, considering the applicable premises of waterfall--a large number of fragmented and disorderly content, no reliance on complex indexes; pictures are very important, text is not important, users browse for no clear purpose, and need a wide range of scanning to do screening; users only do very light interactive consumption of content It is not very suitable for the current watercress guess: first, the information itself is not homogeneous disorder, the quantity is not big enough; the second is that apart from photo albums are smaller or no pictures, users must read the text in order to know what it is; third, this random, fast, low interactive browsing may be suitable for users in depth, but hinders the need to quickly understand the content of the site, To enhance the interaction and exploration of novice users.

Figure I: Current watercress guess screenshot (logged in user)

Under the new product name and positioning, the UI should be adjusted accordingly, trying to integrate all kinds of resources at the same time, to create a bookstore-like easy to explore the feeling. We will redesign this interface from the user requirements analyzed above.

4. 1 overall structure

"People" and "things" separate, with "things" mainly. Watercress Station recommended content can generally be classified as "people" and "things" these two categories, the former such as watercress users, watercress Group, watercress station, belonging to the "Information producers", the latter such as books, videos, sound, online/city activities, Group topics/diaries, albums, belong to "information products." These two categories can be separated, the direct consumption of "information products" on the left main interface, the right column to focus on long-term benefits of "information producers."

Categorize the recommended items (not mixed). With the title clear content, the specific items in the form of examples arranged below, so that the whole station content at a glance. Give the "more" button, guide novice users to do a deeper exploration of the whole station, but also convenient not satisfied with the limited recommended entry depth users get more entries.

4. 2 element composition

The elements placed on the recommended product suggest that this is the reason for recommending the item, such as the Watercress Star, the rating, the number of people like, just like the item's neighbors. However, this is not the case, which leads to unnecessary misleading. In fact, since it's a "casual look" position, at the same time also subject to "can be consumed immediately" constraint, there is no need to give users a prior explanation of the reasons for the recommendation, users can point to understand the details and probation, try to listen to or watch, and the purpose of this product is to arouse the user's curiosity, to attract them to point into the

At the same time, not every book audio and video items have a brief introduction, title, author name is also different, in a unified interface style is difficult to show all the content, not as much as the first page of the larger picture, and the introduction of these contents into the mouse tips. The analogy with the bookstore scene is as follows:

Books such as audio-visual publications, often have a very professional and tailored cover design. First by the cover to attract, and then browse the content, but also more in line with people in real life cognitive habits. But for the community and activities outside the book, the text is more important in the message communication, and it is more convenient to scan and get information quickly. Therefore, the author designed different display methods according to their product characteristics and the user's attention points respectively.

This article, as the author of the Watercress Guess products to provide the analysis of suggestions, welcome colleagues to discuss correction.

Author: Ice Peach