SUCCESS Using Location Targeting and Product Feeds
Wonderschool is an online marketplace to help parents find in-home child care and preschools programs in their area. They operate in about 20 markets throughout the US, with more locations continuing to roll out.
While they also look to get more programs listed on the site, we focused on driving new parent leads into the funnel. Our goal was not only to scale and drive more overall leads, but more specifically the goal was to increase the advertising coverage and generate leads across as many different programs as possible.
Because we were advertising in-person programs for children, we took a geo-targeted approach. We ran campaigns on both search and social. This allowed us to capture the high intent users actively seeking childcare, as well as get in front of our target market and build awareness on social. On each platform, we split campaigns up by geos to track performance by market and be able to control budgets for each.
On Facebook, we started out with general ads that drove people to the region-specific landing page for the area they are in.
But we knew to get more traffic directly to specific program pages, it would be best to run carousel ads using Facebook’s product feeds feature. We created feeds based on the different geos and pulled in different aspects of each program into the ad.
This way, parents could scroll through different programs before even clicking through to find one that would be more relevant to them, thus resulting in a higher likelihood of converting.
However, Facebook optimizes feeds based on which cards are performing best. For example, there may be 50 programs in the San Francisco area, but Facebook is only consistently showing the top 5 programs that perform the best. Usually Facebook’s optimizations are very beneficial, but in this case we were trying to get as many people to different programs as possible.
To combat this, we divided up the feeds even further. In each market, we used multiple different feeds that featured different programs based on different criteria to maximize the amount of programs getting traffic.
We found that using carousel ads to send users directly to school pages instead of generic landing pages performed significantly better. With the carousel product feed ads, we were able to highlight individual schools, show their best picture, advertise the starting price point, include the program type, and give the specific neighborhood the school was located in.
The product feed carousel ads performed much better than the general ads to the regional landing pages across the board.
We saw a 55% decrease in CPC, a 56% increase in CTR, and a 58% decrease in CPA.
On our Google search campaigns, we implemented a few different strategies to increase the amount of traffic to different program pages. For each campaign, the main ads were pointing to the general location’s landing page, as it wouldn’t make sense in this scenario to send people directly to a school page. Instead, we utilized sitelink extensions and price extensions to include additional links to specific schools that we rotated out each week.
We also broke the campaigns out even further by zip code. Rather than sending them to the main region’s landing page with a large number of listings, we could send them to a more relevant landing page showing only a few programs closest to them.
We launched the zip code test in Los Angeles to compare the hyper-targeted zip codes vs the broader region targeting. There were no changes to ads, just the geo targeting and the destination of the landing page.
The zip code campaigns resulted in better conversion performance with a 29% decrease in CPA over the broader LA region targeting.