There are several types of user ratings, and each has its place depending on the context of use. However, they all share some common principles and benefits.

The ability to rate an item should entail nothing more than a minimal amount of investment by the user. The mechanism should be associated with the item and easy to find and positioned in a user flow at the optimal instance, such as after reading an article or making a purchase. The option to leave a rating or review should not be presented at a moment of bias, such as when they are about to delete an app from their phone, as is shown in Figure 10-1.

Figure 10-1. In the early lifecycle of apps on smartphones, a call to action to rate an application on the iPhone often presented itself at the moment of deletion, which is clearly not the best time to expect a positive rating.

Users who write reviews or leave comments should be held accountable. They should have an identity associated with their words and opinions—both to keep them honest and to help build the reputation of that identity. They should only be allowed to vote once lower the risk of some users gaming the system.

It should be very clear what users are rating or to what they’re lending their opinion. They should be leaving this information about a service, experience, or thing, not about a person—although with services such as a real estate broker, rating an agent’s knowledge and expertise is helpful to others. The criteria for rating should be clearly articulated. The simpler the rating mechanism, the less clear this becomes.

You should expose ratings everywhere. For example, you could reveal the rating for an item in search results, on browse screens, and details screens. You can use them to clarify searches and filter results regardless of the type of rating pattern used (see Figure 10-2 through Figure 10-5).

Figure 10-2. The Google Play store shows ratings at the top level to highlight and promote items (Google Play application on Android).

Figure 10-3. Google Play showcases the ratings for applications on a product page as well as ratings for applications people viewed that are similar to the item being viewed.

Figure 10-4. Google Play includes an item’s rating in search and browse listings, both on their mobile devices and on the Web (Google Play application on Android and http://play.google.com).

Figure 10-5. Google Play shows other ratings and reviews of the community as part of the review process (Google Play application on Android).

Because ratings and reviews help to build reputation, there should be recognition for a person’s quality contributions—they should be reflected back to themselves and the community. People are more likely to trust another user’s reviews over a critic, especially if that person is qualified, is like them, and is presented as being authentic and honest. You can present and clarify all of these attributes in the reviewers profile and through your identity system.

Lastly, don’t just implement it and walk away. You should analyze the data collected from ratings, reviews, and feedback and use that data to modify, tweak, and adjust the offerings to your users based on the things they tell you. The pact that you make when asking someone to sign up and give feedback is improvements to the content, the quality, or personalization of their experience.

The user wants to promote a particular piece of content within a community in which they participate. This promotion takes the form of an upvote for that item, and items with more votes rise in the rankings and are displayed with more prominence, as depicted in Figure 10-6.

Figure 10-6. The question and answer site Quora uses a highlighted button labeled “Upvote” to encourage users to participate if they think the answer is good. The downvote mechanism is downplayed (Quora Android application).

Provide a voting mechanism, attached to each item in the community pool. Clicking this mechanism counts as a vote in favor of that item’s promotion:

Highlight popular items:

Try to ensure that users are voting on items that they have actually consumed (read, watched, listened to):

Items with fewer votes are not punished for their lack of popularity; they merely fall into obscurity, and disappear into the deep end of the popularity-ranked pool.

You might want to consider a moderation control that lets the community decide to remove an item altogether, but don’t make this control prominent. The emphasis for this pattern is on promoting the good, not punishing the bad. You should downplay the downvote or exclude it altogether.

Be aware that a vote to promote type rating is often used as a substitute for Favorites or Bookmarks by some people. It is important to periodically analyze what the community is doing to see if both types of mechanisms need to be part of the available actions, as shown in Figure 10-7.

Figure 10-7. Digg uses both an upvote (Digg) mechanism to adjust placement and a Save control to collect articles into a private list (http://www.digg.com).

This pattern is popular with a lot of user-contributed content aggregation sites (most notably on the link-popularity sites Digg, Reddit, and Newsvine, and the Q and A sites such as Quora). Such systems for collective choice are a good way to promote community participation; they provide a low-cost means for promoting popular content. Note, however, that popular content does not necessarily equate to quality content, so no promises of content quality should be made.

A user wants to express a like/dislike (love/hate) type of opinion about an object (person, place, or thing) he is consuming, reading, or experiencing, as illustrated in Figure 10-8.

Figure 10-8. On YouTube, users can rate videos with a thumbs-up/thumbs-down style of rating. The mechanism for rating is close to the video in both mobile and web instances (YouTube iOS application).

Do consider thumb ratings when opinions about a rated asset will tend to be strongly polarized. For example, if you can state the question simply as, “Did you like this, or did you not?” thumbs might be appropriate. If it seems more natural to state the question as, “How much did you like this?” star ratings are probably more appropriate.

Do consider using thumb ratings for developing personalized recommendations.

For example, Pandora uses declared music interests to create a personalized playlist of songs that are similar in style. It then uses the thumbs-up rating (Figure 10-10) on a song to add more songs like this to the music stream and a thumbs down to remove that song from the music stream.

Figure 10-10. Thumbs-up and down on Pandora is embedded directly into the music controls (http://www.pandora.com).

The benefit of rating is expressed to the user by instantly stopping the play of the song.

You can see a simple yes/no, like or dislike, thumbs-up or thumbs-down action for personalizing content or for giving the system instant feedback in mobile apps ranging from dating apps such as Tinder and Grindr, to the pet matchmaking app BarkBuddy, to voting apps such as Thumb (see Figure 10-11 and Figure 10-12).

In mobile experiences for which a quick like/dislike is associated with imagery, also consider using the equivalent swipe gestures—left for no and right for yes are the current interaction pattern.

Figure 10-11. BarkBuddy uses the heart (thumbs-up) and the X (thumbs-down) to let users quickly note likes and dislikes. The positive items (the heart) are collected into a list of favorites, and a Stats section tabulates you which breeds you tend to favor (BarkBuddy iOS application).

Figure 10-12. On both BarkBuddy and Tinder, gestures are the primary interaction, with a right swipe indicating a positive intent and a left swipe indicating a negative intent. The swipes are redundant to tapping the X or Heart. In the Tinder case, there is only system response if the other person reciprocates positively (Tinder iOS application).

Do not consider thumbs for rating multiple facets of an asset. For example, don’t provide multiple thumbs widgets for a product review intended to register a user’s satisfaction with that product’s price, quality, design, and features. Generally, thumbs should be associated with an asset in a one-to-one relationship: one asset gets one thumb up or down. (After all, Emperor Nero would never let a gladiator’s arm survive but put his leg to death! Think of thumbs as “all or nothing.”)

Do use thumb ratings when an easy, lightweight ratings mechanism is desired. The context for these ratings should be appropriately fun and light-hearted, as well.

Do not use thumb ratings when you want to provide qualitative data comparisons between assets. For example, in a long listing of DVDs available for rental, you might want to permit sorting of the list by rating. If you provide thumb ratings, this sort would have very little practical utility for the user. Instead, you should consider a scalar, 5-star rating style.

The concept of thumb ratings can be problematic for some locales.

First, the symbolism of an extended and raised thumb is problematic: in some locales, it is considered an insult, thus a plus/minus widget would probably be more appropriate; in other locales, representations of any body part (especially disembodied ones) are considered offensive; and, finally, some locales will just not “get it.”

Second, the very notion of a binary, black-or-white, love-it-or-hate-it ratings system might not be a natural fit for some cultures. Many locales have stated a preference for “shades of gray” in a polarized scale. (Note this is subtly different from star ratings, which imply “I like it exactly this much.”)

Third, in some areas it is considered rude to criticize things openly. In this instance, consider the thumbs-up-only variation.

The benefit to other users with thumbs-up/-down style ratings is that these ratings, when assessed in aggregate, can quickly give a sense of the community’s opinion of a rated object. They can also be helpful for drawing quick qualitative comparisons between like items (this is better than that), but this is of secondary importance with this rating type.

A user wants to quickly leave her opinion on an object, with minimal interruption to any other task flow she is involved in, as depicted in Figure 10-13.

Figure 10-13. Ratings and reviews on Petco.com.

Rating an object provides a lightweight model for user engagement. Ratings are often tied with reviews to encourage richer user contributions and activity.

Multifaceted ratings provide a mechanism for users to give nuanced feedback about a product or service, even if they don’t write a lengthy review (see Figure 10-14).

Figure 10-14. A faceted rating widget for a real-estate agent on Trulia (http://www.trulia.com).

Faceted ratings can be a powerful research tool for consumers looking to make a purchase. They are useful for skimming in search results and offer a lot of valuable information in a smaller footprint than lengthy reviews.

As seen on

Cars.com (http://www.cars.com)

Trulia (http://www.trulia.com)

The Lifespan of Content and What Should Be Rateable

What type of items are rateable?

Items that should be rateable by the community share some common traits.

Rateable items should have some intrinsic value.

We should never ask users to provide metadata (basically “add value”) to an item whose own apparent value is low. Or, more specifically, we should be careful to only ask for user participation in a way that acknowledges an item’s intrinsic value: it might be OK to ask someone to give a thumbs-up rating to someone else’s blog comment (because the “cost” to do so is low, basically a click), but it would be inappropriate to ask for a full-blown review of the comment. There would be more effort and thought involved in writing the review than there was in the initial comment!

Rateable items should persist for some length of time.

Rateable items must remain in the “community pool” long enough for all members of the community to cast their votes. There’s also little use in asking for a bunch of metadata for an item if others cannot come along afterward and enjoy the benefit of that metadata (Figure 10-16).

Figure 10-16. The content-lifespan spectrum.

A RATEABLE ITEM’S LIFESPAN

Highly ephemeral items such as News articles that disappear after 48 or 72 hours probably aren’t good candidates for rating.

Juxtapose this with TiVo (which is kind of the canonical example) where thumbs (up and down) are used to rate TV shows for recommendation purposes. The programs being rated have high value (i.e., they cost money to produce, and we get entertainment value out of them), and they’re highly persistent (they’ll recur at the same time next week, and in reruns, and in syndication, etc.; for all intents and purposes, TV episodes are immortal). And TiVo’s entire user experience (including user-education movies that ship with the unit, its printed manual, and, heck, the dang remote has ‘em hardcoded on there) is oriented around the thumbs-up and thumbs-down voting. I’d venture that thumb voting and the recommender system are a huge part of why many people buy TiVo in the first place. (OK, that plus “pause live TV.”)

Items with a great deal of persistence (on the extreme end are real-world establishments, such as restaurants or businesses) make excellent candidates for rateability. Furthermore, the types of ratings we can ask for can be more involved. Because these establishments will persist, we can be reasonably sure that others will always come along afterward and benefit from the work that the community has put into the item.

When it comes to explicitly input recommender systems, we should acknowledge the limitations of folks’ interest in “feeding the machine.” If they understand the benefit, and they think that the work they’ll put in will at some point be worth something to them, then people will play along.

BRYCE GLASS, COAUTHOR, BUILDING WEB REPUTATION SYSTEMS (http://buildingreputation.com)

A user has a comment or opinion about an item she is viewing on the site and wants to share it, as demonstrated in Figure 10-17.

Figure 10-17. Medium (http://medium.com) displays comments integrated within the article. Comments are indicated with a number bubble (more as an annotation than end comment) and as the next reader scrolls over the comment, the specific phrase, sentence or paragraph is highlighted in order to give context to the comment.

Comments are an easy way to promote user participation on your site and can be a conduit for multiuser conversation. Comments associated with an item give context to the conversations and the participation.

A user wants to share her opinion with others about an object (person, place, or thing) in greater detail than a simple rating or comment, as illustrated in Figure 10-19.

Figure 10-19. A review form on Amazon (http://www.amazon.com).

Qualitative fields such as Pros and Cons seem to be easier for users to create than a full narrative. They do not need to think in complete sentences, and they have more specific direction regarding what to write (positives and negatives). Additionally, readers find them easier to scan than a narrative.

The site owners want to be able to collect feedback about the site from its users, as depicted in Figure 10-22.

Figure 10-22. The Hunt’s (http://www.thehunt.com) web version of their service offers a question mark call to action to trigger a feedback entry form. Selecting the radio button automatically adds a screenshot to the entry allowing the folks on the other end to know where the user was when they needed help or wanted to offer feedback.

Giving your users the tools to offer feedback and suggestions imbues them a sense of ownership over the site. If the site is extremely social and houses a person’s online identity (see Chapter 4), users already have a sense of ownership for the site. They are using the site frequently and most likely in ways that you had never predicted. Tap into that passion.

Tapping into the community is a way to gather new ideas for improvements and future features, and can give insight into how your users perceive the site experience and your brand.