>> perspectives

A/B Testing at Yieldmo

Together with the Ad Format Lab, the AB testing team at Yieldmo looks at both the qualitative and quantitative approach to mobile UX.

What does this mean? It means that while we continue to move the mobile advertising conversation forward – we are also doing it in a way where we maintain value to all three sides of the mobile experience – users, content developers and marketers.

Always. Be. Testing.

This results in our obsession with testing.  We measure every change we make to our formats and see how it impacts mobile consumers, marketers and our publishers.

Because the mobile landscape is changing so fast and in so many dimensions (devices, operating systems, content, user profiles), it would be difficult for any ad tech solution to claim to have a hold on the market with a “killer ad” (which quickly becomes copied and commoditized). It’s not a way to sustain a business with our partners.  We can’t predict which apps or even mobile behaviors will become paramount in the next 3-5 years, but we can adapt trends that we’re seeing right now.

We can do this with more precision and maneuverability because we have code on the page.  This allows us to do the work with minimal impact to our publishers. That code gives us the blank canvas that allows the Ad Format Lab to create new ad treatments. They take their creativity and user design experience and come up with unique ways to capture the attention of mobile users.

Our testing team takes those formats and throws them out into the wild to gauge their effect on those mobile users. We then take those results and peel apart the numbers to better understand how and why the ad format was effective or not. By isolating every component of the ad, we can think about how to improve it to make it more impactful. We can do this because we can capture all the data variables and measure each component down to the pixel if we want to.

To put it in perspective—advertisers and agencies can’t do this with a lot of other ad partners.  If you send the same ad through an exchange—you have no control as to where those ads are run, and any data you get isn’t first party and often spotty.

Measuring beyond the click

Mobile has given marketers an incredibly personal vehicle to send messages to the customers in the way that speaks to them the best.  How best to gauge if the message has actually been received in a positive and impactful way?

Instead of relying on third party measurements, we just look at what the readers are doing.  Are they playing the videos?  Swiping through a carousel?  Re-triggering the ad animation? We work with marketers to understand what the ad is intended to do, and because we built it—we can tell you if it actually worked. And sometimes we don’t get it right, and that’s OK – that’s the whole point of testing. It’s better to fail and know how to do it better next time, than to run campaign after campaign using a buckshot approach.

For many of our advertisers, clickthroughs and conversions might be their only measurement.  At Yieldmo, the click is just the beginning. All of our statistical analysis and iterations start after the click because the mobile device is completely different than the desktop. Applying the same metric is backwards.  And as brand-marketers know, there’s so much more than the click.

As our new formats bring new functionality that draw users at the top of the funnel – your clients/advertisers will be looking to get data about that functionality:

If you’re a brand, we offer faster and more nimble ways to get your brand impact results. We’re working to match our data and analytics with the way that brand marketers and agencies think.

Testing for Publishers

Obviously, more effective marketing impacts the publishers that work with us as well. While we employ standard AB testing techniques, one of the exceptions is that our tests are run across a network, and not just on one site. A #0099CC button background color may not be as effective on one site as for another site.

If a publisher is testing every component of their site, they should be able to do the same with the ads that run on their site. We’ve created tools to allow publisher-focused teams to do per-site, per-placement testing. Again, the goal here is to maximize revenue opportunities, without sacrificing the user experience.

When we talk about “contextual” advertising – we aren’t just talking about content.  We can help marketers find insight not only within the specific format features, but also within the context of the content. We think about how our ads need to fit within the hardware, software, content and the specific user.

Ultimately, our design decisions think about these combinations.  We then take advantage of our diversity across these parameters to test our theory out in real world.  

At the end of the day, marketers work with Yieldmo because we can deliver value with minimal waste. Avoiding waste also means that users are seeing products and services that they want to see at the right time. Avoiding waste means that marketers are willing to pay more for the same amount of traffic.

We listen to customers all the time…with data. The numbers are telling us that standard formats are not effective against the constant change in the way we consume mobile.  Publishers and advertisers work with us today because we’re creating trends, and not just following them.

Looking at all sides of the Ad Tech ecosystem (which most importantly, include users), we’re working to solve their mobile ad problems without swags and guesstimates.

Posted by: Lucas Piazza, Director of Marketing