Before we dive into crowd curated search, let's take a look at search and why it is important for content heavy platforms. Whenever a content rich platform is built, it will soon become evident that an accompanying search engine must also be built to retrieve the stored information. However, retrieving relevant information from millions or potentially billions of records is not a trivial task. The problem itself is so complex that there is a discipline dedicated to solving it, called Information Science.
The top six internet sites (as ranked by Alexa) all feature a prominent search bar placed on the navigation panel (Figure 1). Facebook, YouTube and Amazon have chosen to place the search bar right next to their logo, highlighting how important search has become for these platforms. Google, the world's number one website, was initially built around this single problem - search! At Flipp, where we focus on weekly shopping content, search is equally important in delighting our customers with relevant weekly content.
Figure 1: Top 6 websites showing the importance of search
Search is Navigation. This is the primary way people discover content on a platform. No one really has the time to learn your app or web hierarchy. There are many popular apps and sites competing for users' time and they are constantly evolving. When was the last time you used the menu bar? or even used search filters? Unless you are forced to, you will naturally stay away from them. Platforms should be designed to be frictionless in discovering content, otherwise users will lose interest very quickly.
Why is Curation important in search results?
When the results are not relevant to the user, they will not take the next action and click on the results. Curation is important because it helps to increase the search results' relevancy. Figure 2 (below) demonstrates the difference between non-curated and manually curated results when "cake" is searched in Flipp. What exactly is manual curation? It's the process in which a human actually checks each and every search term and manually arranges the sort order of the products. We can see that the "Manually Curated" version shows a much cleaner, more relevant set of search results to the user. However, while some of the manual arrangement can be automated, it is still a resource intensive task.
Figure 2: Manual curation of the search term "cake"
Why is Crowd Curation important?
While manual curation is a great way to get started, it is not a scalable solution, and a better approach is needed. This is where crowd curation comes into play; using the 'wisdom of the crowd' to sort the order of the search results. A simple approach is to see what items users are clicking on the most and bumping them up to the top of the search results. Figure 3 shows an example of the before and after for search term "bbq".
Figure 3: Crowd curation of the search term "bbq"
Simply measuring the click count on an item and sorting based on that yields better search results. However, because items change on a daily basis in the Flipp app, the search results require a periodic tune-up. Ensuring that we do not penalize the new flyer items against the older items (that have received more impressions and therefore collected more clicks) creates an interesting challenge for our software team.
Search is also slightly different on mobile platforms. Because the screen size is smaller, we have to also consider what is actually displayed in the view port (above the fold) vs what is below the fold. There is a greater chance that a user will click on an item that is shown at the top rather than items farther down the list that they have to find by scrolling down. If the user does take the effort to scroll down to find an item, then that has to also be taken into account when improving the sort order of the search results.
Search is a very broad topic. We have only scratched the surface of it. There are many interesting improvements we can make to revolutionize the product. We envision search to become more personalized and suggestive so that users can quickly find relevant content that they are really looking for.
The most important measure of a search engine is the quality of its search results. Figure 4 shows that the gap between the number of searches and the click on their results is widening. How do we close the gap and measure success?
Figure 4: Searches vs Clicks
Click Through Rate (CTR) is the ratio of users who click on a specific item to the number of total users who view the search results. CTR can be used to measure the effectiveness of the search engine. There are several techniques to measure quality. As suggested by our engineers at Flipp, we are using Discounted Cumulative Gain (DCG) to measure the quality of our ranking algorithms.
At Flipp, we want the user experience to be magical. We are a data driven company that constantly looks at data to find new ways to improve the lives of millions of users. Search is just one area where we are applying this principle.
Want to be part of enriching the weekly shopping experience using the latest technologies? Check out what the Engineering department at Flipp has to offer on our careers page.