Crowdsourced Curation

Let’s Curate Together

Let’s Curate Together

YesterYear It Was About Aggregating Ratings and Reviews

When we write a review about a product or a service on Yelp, Google Places, OpenRice or other review app, we’re typically writing from our own perspective. How do I like the taste? How do I feel about the service? We do not have a specific audience in mind – other than ourselves. We tend to do this because review apps like Yelp do not differentiate audiences; all reviews are applicable to all audiences. In other words, these review apps do not curate for specific audiences. 

Having more and more reviews without curation leads to diminishing value. Try looking for a good Korean restaurant using a review app. You’ll get back hundreds of places – most with an average of 4-5 stars from thousands of reviews. How can you use this information to make decisions? Well, you can spend a lot of time and effort to sift through all the places and read the reviews and hopefully you’d find something suitable. The problem is we are inundated with too much information for us to process efficiently. 

Curating Gives Reviews A Target Audience

What we need are recommendations that are targeted for a specific audience. For example, “Best Korean restaurants in Hong Kong for diners seeking best value for money”, “Best Korean restaurants for those with finer tastes”, “Healthy diners’ guide to Korean restaurants”. 

While publications like Tatler, TimeOut, SassyHK, and others provide curated recommendations targeted to their demographics, they are providing the recommendations from a single person’s perspective and they have an inherent conflict of interest – they are often paid by the businesses they recommend. 

Today It’s about Crowd-Curation 

We want curation, but from sources without conflicts of interest, and from people with some depth of knowledge in the field we’re seeking recommendations. Many people have begun seeking recommendations from online forums or community groups feeling that those venues provide more authentic suggestions with sufficient depth of knowledge. This is today’s trend: to get crowd-curation from like-minded people. 

While online forums and community groups provide us with a channel to ask fellow group members, they are not designed for finding relevant recommendations. Location-based or product-based searches are generally not features in these channels. 

A platform designed specifically for crowd-curation of recommendations is what we need. On this platform, users join groups with people with similar interests, values and lifestyle. Together, they share, validate and curate recommendations. Like most social platforms, posts would be inherently trustworthy as authors hold reputational risk, and they would also be contemporary as content will be sorted on recency and popularity. 

What differentiates the crowd-curation platform from other community-based social platforms is its focus on recommendations, and its flexibility to capture, organize, curate and validate content specific to the fields of interest: 
  • Recommendations for restaurants, for example, would need to capture data on: cuisine type, location, quality of food, quality of service, decor, pictures of food, pictures of interior, etc. 
  • Curation of restaurants would have a couple of dimensions: the targeted audience based on the group the recommendation is posted in, and what it is being recommended for (e.g. “Best dim sum for your money”). 
  • Validation of recommendations would be based on members voting and commenting on the recommendations. 

The Bumping App – For Crowd-Curation

Install the Bumping App and try it out for yourself!


Playground for Co-curation

A Playground for Co-Curation


We use Instagram to discover interesting places and products. People you follow often share pictures of delicious dishes at restaurants they visit everyday, and each time you see one of those posts you really want to go to the restaurant. However, if we want to make a simple query like “Show me recommended Chinese restaurants in Hong Kong”, this cannot easily be done in Instagram. Instead, users have to resort to bookmarking posts so they could later browse through the bookmarks to remember posts they found interesting.

Perhaps it’s more pertinent to ask if people want to use social platforms for getting recommendations at all. After all, there are apps made just for reviews like Yelp, Google Places, and Openrice. Well, if you ever tried to find a restaurant to go to on one of these apps, you’d know it’s not easy. They do not curate, and the reviews are not often updated.

And that’s why we often resort to branded articles such as Timeout’s “Top 10 Italian Restaurants in Hong Kong”. We like them as they’re generally well-written, and they’re often updated within the last few months – they’re relevant. We also tend to trust these curations given that they are branded recommendations. However, we also know that there’s an inherent conflict of interest in most publications; publishers are often paid by the businesses they recommend.



We know people want up-to-date recommendations from trusted sources curated to their tastes, but what solutions have worked that help people discover?

Online forums and community-based social apps provide part of the solution. They enable users to solicit recommendations from like-minded people in real-time thereby getting up-to-date recommendations, but they don’t organize recommendations for browsing and discovery later.

Xiao Hong Shu, a Chinese social platform that caters to people interested in beauty and fashion, enables users to share their experiences with products and services in the beauty and fashion industries and categorizes the shares. Users can then easily discover and search for reviews they are looking for. Because content is created by users that do not have conflicts of interest (at least not that we know of), people trust them. And because it’s used as a social platform, users post updated experiences everyday. Today, users of Xiao Hong Shu will perform a search and look at reviews on the app before buying any fashion, beauty products. Can an app like Xiao Hong Shu be adapted to recommend products, places and services other areas of interest?



A conceivable solution is a community-based social-platform that enables users to co-curate posts. Instead of having a single blogger or publisher create and organize content, having many passionate individuals with a common interest perform the categorization would be much more effective and valuable to other consumers.

Wikipedia, for example, has adopted this model. Users of Wikipedia can become moderators and help organize, curate and govern the appropriateness of content created by other users. The trustworthiness and general value of Wikipedia is very much directly attributable to the effort provided by the moderators and the contributors.

Central to the co-curation platform is the ability for users to curate posts for any topic. To this end, the platform needs to support the ability for users to associate a post to two key dimensions. The first is an extensible categorization structure needs that users can use to categorize posts hierarchically. For example, a post about “Dim Sum” should be categorized in a hierarchy such as: Restaurants->Chinese Cuisine->Dim Sum. The second dimension that needs to be integrated throughout the platform would be the locational dimension. Posts need to be associated to a specific region so that users can find relevant recommendations.

Here is a diagram of a potential dimensional structure for a flexible curation platform:



The Bumping App is designed to enable social curation. The Bumping App is a community-based social-platform where users can co-curate content so that they and other community members can support businesses and topics that they deem important.s