The online review space is crowded with 100s of review sites, many failing to gain traction. It’s a classic chicken and egg problem. To do well, most sites or apps need many user generated reviews, however it’s hard to gain users and engage them without reviews. In recent years sites that tended to break out and gain traction have a great experience, product, and unique niche. With Judy’s Book as our reviews grew beyond 1 million, combined with another 3 million from partners, we started to see time spent on certain pages increase but users were also dropping off these same pages at higher rates. As a review site acquires more content it gains more traction, but also experiences new problems; users might abandon as it becomes harder and more time consuming for a person to filter through all the reviews and not get overwhelmed. Here are some ways we’ve tackled this issue that are working well.
Filters. Filters and searching within reviews allow a reader to get a feel for the place quickly without reading all. I tend to read a few excellent reviews, and then go to the average and poor ones to quickly get a picture of a place. We keep in mind the persona of our audience, what they would want to filter/search, and if we have enough information on the reviewers to slice our data in a useful manner.
Sorts. There are all kinds of things you can sort on; the most obvious are rating and date. Recent reviews tend to give a better picture of a place, especially since many owners read their customer reviews and use them to improve. The key here is to not go overboard with sorts and keep it relevant and simple.
Best Worst showcased. Many users want to know the best and worst about a place quickly. This is a one way to combat against fake reviews that get through review site filters since some owners will hire reputation management firms to combat bad reviews by generating good ones that are more recent, with best worst the critical review is present till a more recent critical review shows up.
Scores. At a glance ratings and scores tend to be popular with users because once the user understands the basic criteria for each item scored it can be all encompassing of their likes to quickly make a decision. They are also a good way to build a brand and syndicate your score to other places without cluttering a 3rdparties UI. KidScore is an example of utilizing scores. The challenge here is to determine if your users would benefit (what problems are you solving and questions are you answering?), identify the criteria, and then generate an algorithm to accurately determine a score from review sentiment, votes, other method, or all. You’ll need a strong development team to do this well and be able to scale.
Got your own ideas? We’d love to hear from you on feature suggestions and improvements.