Vivek

Q&A

Anya Pratskevich by Anya Pratskevich | February 27, 2019

As data and technology become table stakes in media buying, mobile marketers find themselves spending more time — and money — building their in-house technology stack.

For Vivek Girotra, Director of Performance Marketing at FoxNext, technology is the future of user acquisition. “Having an army of media buyers working on Excel sheets is outdated,” Vivek says. For him, the key to mobile growth is building a world-class technology stack to support human talent.

Vivek recently joined FoxNext, a division of 20th Century Fox and the Fox Networks Group, from Machine Zone. We caught up with him at the App Growth Summit to talk about the modern mobile marketing technology stack and the build vs. buy conundrum.

What are the key steps in building your mobile marketing technology stack?

First, you need to figure out what tools you need to track the entire consumer journey from the first impression to the final purchase event or in-app interaction. This means choosing the right attribution partner, reporting solution, data storage and BI software to analyze the data. This is especially important to get right early because switching technology providers can be a pain, especially if it involves changing an SDK in the app or retooling the BI system.

Second, you need to design your technology stack for scale. For example, right now FoxNext has one mobile game in the market; but in six months, we are going to have three. So our solutions need to scale and work reliably with a high volume of traffic.

The third step is hiring the right people. At FoxNext, we have an equal number of engineers and media buyers in the user acquisition team. We hire people with unique sets of skills using the framework called MECE, which stands for Mutually Exclusive but Collectively Exhaustive.

How do you decide between building in-house vs. working with vendors?

One way to think about this “build vs. buy” question is obviously the financial feasibility. I spoke to a vendor recently that charges ~20% fees for a fully managed Facebook Ads service. So if you spend more than a million dollars a year, you can hire a full-time person to manage your campaigns, with some change left to spare. Given our budgets, it was almost a no-brainer for us to keep the social media buying in house; outsourcing to a vendor would have increased our overheads substantially.

The other question one needs to ask is whether the service is a parity product or if it gives you a competitive edge in the market. For example, few mobile advertisers want to build their own attribution technology because it has matured to the point of becoming a commodity (MMP sales reps might disagree). On the other hand, investing resources into an in-house programmatic bidder is a different matter, because we can build out something that is customized for our product and our internal monetization methodology.

Tell us about the technology stack behind FoxNext apps?

We work with Appsflyer and Singular for attribution and reporting, and use Google BigQuery and Looker for our warehousing and analytics. We are building our own campaign management tool and custom bidders to support the specific networks that we work with.

How do you select the right mobile marketing vendor?

In my opinion, there are a few screening questions every performance marketing team must ask:

  • Does the vendor pass the “smell test?” Is the partner a well-known brand or some white-labeled affiliate network being run out of a garage? Having experienced media buyers on the team helps to evaluate vendors effectively.
  • Does the vendor offer any unique source of inventory or a differentiated technology layer? We buy directly from sources and avoid rebrokering/arbitrage players as much as possible. I have found the Pareto Principle (80/20 rule) to be true in most cases and would rather work with a smaller set of partners that provide outsized value.
  • Does the partner fill a gap in our marketing portfolio? This is a subjective question and depending on the team’s current structure, resources, spend levels, and future plans, it may or may not be the right time to onboard the partner even if the first two questions are answered in the affirmative.

How open are you to adding new networks or technology to your marketing stack?

While we are very selective in our choice of partners, we are always open to learning and improving our marketing strategy. We have meetings with new vendors quite regularly, and I respond to cold emails as well if there is something interesting on offer that might add value to our team.

When running tests, we work closely with our partners and have a high threshold for learning. We try to set partners up for success by giving them the bandwidth and data they need to test, optimize and scale as needed.

How do you approach fighting mobile ad fraud?

Fighting mobile fraud is like playing a game of whac-a-mole. Every time you think you have fixed one problem, something new shows up. That being said, prevention is better than cure and there are some simple steps we take to minimize occurrences of fraudulent traffic.

First, we log all of our data: clicks, installs, publisher IDs and anything our MMP and partners can send us. Even if it doesn’t seem useful right now, it will be someday. This enables us to build our own dashboards, and set up alerts or conduct spot checks in case of any discrepancies.

Second, we have found that judicious partner selection is a great filtering mechanism. Working with reputed partners who provide transparency and are working towards the same goals goes a long way in fighting fraud effectively.

Third, we apply strict attribution windows and have disabled fingerprinting completely.While this limits the scale we can get on certain partners (especially with users who have LAT enabled), I find that the benefits of simply not allowing for the possibility of fingerprinting fraud far outweighs the costs of having to deal with suspicious traffic.

Read more interviews with mobile marketers from the App Growth Summit: