Edmunds Builds Industry's Fastest Automotive Ad Platform

Edmunds is an American online resource for researching new and used cars, with over 16 million page views a month.

As they grew, Edmunds realized that car manufacturers and dealerships were clamoring for an opportunity to get more visibility on Edmunds's site - leading to the decision to incorporate sponsored car listings into their site/app. These ad units would involve car manufacturers/dealers paying to promote their car listings in relevant car searches.

Such an ad product would be a win-win-win: dealers would love the opportunity to drive more visibility; users would get a native, unobtrusive ad experience; and Edmunds would create a new, high-margin revenue stream.

So Edmunds decided to build their own sponsored listings platform for displaying sponsored car listings in search results.

Below details that process, why/how they did it, and what the results were.

What Edmunds wanted

In designing their platform, Edmunds had a few key goals:

1. To have car listings look and behave like organic listings - including information like the right car make and model, MSRP, and display image
2. To have fast ad response times. Using standard JavaScript tags they were seeing 500ms-2s+ responses when they wanted < 50ms
3. To drive more revenue through high-margin, highly-targeted native ad units
4. To release the product in weeks, not years

Edmunds uses Google Ads Manager as their standard ad server - but it offered little in terms of customization around the ad's look/feel, response times, and search term targeting. They tried but were unable to create the exact sponsored ad unit they wanted using Google's JavaScript tools.

They could have alternatively tried to build this custom ad platform from scratch - but that would have involved hiring new engineers or borrowing some from other projects. Because of this, they partnered with Kevel, whose ad serving APIs made it easy to build the platform they wanted in weeks.

What Edmunds built

The sponsored car listings platform Edmunds built involved only backend changes. To the user, the layout of the page and app stayed the same, and the car listings looked and behaved like they did before.

With this product, when someone searches for a specific car, then a car manufacturer or dealer can pay to have their car listed in the "Deals Near You" section.

For instance, Toyota can pay to have a listing for their 2019 Tacoma appear whenever someone searches for "Toyota Tacoma".

With this ad platform, Edmunds now had:

1. A new revenue-driving ad product that barely changed the user experience
2. High-value ad units that loaded just as quickly as organic content did
3. An ad platform that was compliant with all privacy laws since it relied on intent and search terms, not PII

The results

Edmunds saw positive results immediately upon launching:

Revenue growth	These new ad units have led to 30% ad revenue growth
Fast response times	Edmunds got response times as low as 35ms - ensuring that these paid listings didn't slow down page/app load speeds
Launch date	Using Kevel they released this product in just weeks - letting them launch 9x faster than if they had tried to do it entirely by themselves

How they did it

Edmunds built their platform by migrating from JavaScript ad tags to server-side ads, enabled through Kevel's ad serving APIs.

This enabled to Edmunds to:

1. Request ads on the backend - By pinging Kevel upon ad request, Edmunds gets a response in JSON containing details about the promotion to display, which they then insert directly into their site/app.
   Below is an example of the type of information returned in the JSON response, which contains data they can place into their apps - such as the image URL, MSRP, and other metadata.
2. Have fast response times - Since ads are requested server-side - not through page-slowing JavaScript tags - Edmunds can place this content with little delay - as fast as 35ms (an order of magnitude faster than most tags).
3. Have a malware-free and privacy-law-compliant platform - By not using client-side (on-page/in-app) code, Edmunds ensures no third-party can insert malware or drop cookies that could leak user PII. Plus, since search targeting doesn't involve PII, they don't have to worry about asking for consent.
4. Optimize revenue via auction bidding - Using revenue optimization tools, if there are multiple listings that are eligible for a given search, then the algorithm will select the promotion that'll drive the most revenue for Edmunds.

Takeaways from Edmunds's sponsored listings ads

Edmunds's results highlight the importance of treating ads like an extension of your core product, rather than something to outsource to a third-party.

While building innovative, custom ad platforms takes vision and engineering time, it comes with many benefits, such as having full control over your ad revenue. If Edmunds had stuck purely with Google Ads Manager, they would not have seen the 30% revenue growth they saw.

Have you built your own sponsored listings platform? We'd love to hear from you!