Many Branch customers engage with their users across multiple platforms: web, mobile web, mobile app, and physical locations. More recently, the rise in viewership of connected television (CTV) has many of our customers interested in this new, growing channel. But many face the conundrum: Is CTV advertising an effective way to engage with my users? I asked Moloco‘s head of CTV, Jake Richardson, to answer some of the questions I’ve heard from Branch customers to help you decide if CTV is right for your brand.
— Adam Landis, Head of Growth at Branch
First, can you tell us a little about your company and your role?
I should start by noting that Moloco is a portmanteau for Machine Learning Company — which is at the heart of everything that we do. Moloco’s machine learning platform delivers growth and performance to app publishers by unlocking their unique, first-party data. One of our core products, Moloco Cloud DSP, has quickly become one of the fastest-growing DSPs, enabling performance marketers to accelerate user acquisition and lifetime value. We thrive largely because we power lower-funnel outcomes for our customers. This has also allowed for our unprecedented growth and, more recently, has allowed us to begin scaling a CTV business.
I joined just under a year ago to lead Moloco’s CTV business — both supply (ensuring premium, performant inventory) and demand (sourcing advertisers to buy that inventory). We have seen a lot of early success and are excited to continue building out the product and bring it to market alongside partners like Branch. Prior to joining Moloco, I was director of CTV strategy and partnerships at The Trade Desk. Before that I was a mediocre attorney.
CTV is growing rapidly, both in ad spend and viewership. Who are some of your customers, and what are they aiming to do with CTV?
Much of our early success has come from our core customer base: gaming. And many of our gaming customers have come to us for CTV because they want to see measurable outcomes and tangible results. These customers have tried either linear TV or buying CTV through large branding-focused agencies and platforms, but they do not have the stomach for spending without understanding what KPIs that spending drives.
As we continue to expand, we’ve seen success and traction on performance CTV in verticals like finance, travel, and real-money gambling. Of course, that list is endless because advertisers are increasingly looking for more value from their CTV buys.
What are some of the fundamentals you see with your most successful CTV customers? Which verticals should be most strongly considered when testing this channel?
For us, there are two types of ideal customer profiles depending on where our advertisers are coming from in the market. First, there is the advertiser who’s robustly established themself in the performance business and is expanding into CTV, which they consider a necessary channel (think: a large, global advertiser with household name recognition).
And second, there is the advertiser who hasn’t expanded into CTV yet but stands to gain significantly from integrating it into their strategy. As cost per acquisition (CPA) has risen, advertisers who were “raised” on direct response (DR) marketing have had to diversify into more brand or “brandformance” marketing channels, including CTV. But at their core, these advertisers are always looking for measurable outcomes from all the media channels they buy from (think: direct-to-consumer apps and products).
Which customers would fare poorly with this type of advertising?
The biggest place where we see misaligned advertiser expectations is when advertisers expect their CTV inventory to perform at parity with bottom-funnel mobile user acquisition. Oftentimes, we have this trouble with smaller or purely performance-driven gaming customers. Suppose a customer’s entire growth budget is allocated to achieve return on advertising spend (ROAS) at the cheapest possible KPI. In that case, CTV probably won’t be the best fit, as customer acquisition costs can be higher in CTV. Added to this, in mobile, there is a fully closed attribution loop: the customer views an in-app impression, clicks on it, downloads the new app, and all of this takes place in a tight feedback loop. In CTV, the complexities expand. So, too, must the advertiser’s tolerance for differential consumer behavior and new measurement frameworks to take into account the differences CTV requires.
We know all about these complexities; our cross-device attribution tracks users downloading an app on their mobile device from ads shown on a CTV. It is a great feature but was incredibly complicated to build. How would you categorize the difference between a beginning customer and a sophisticated customer? What capabilities or methodologies are different between them?
Because performance CTV is relatively nascent, many of our customers have similar first-time learnings — whether they’ve been running CTV budgets en masse for the last half decade or are first dipping their toes in the water.
What we notice with customers who are just starting out is a bit of a warming-up period: either transitioning from bottom-funnel DR to performance CTV, or from top-funnel branding CTV into performance and outcome-oriented CTV. One of the issues we often run into is looking for click-through attribution, or extremely short view-through attribution windows, rather than looking at CTV holistically — as a part of their entire marketing funnel.
In contrast, many of our more sophisticated customers will evaluate CTV not as a standalone channel (though it can stand alone on its performance) but as just one part of their advertising strategy. These customers will leverage our robust analytics tools, including our incrementality offering, A/B testing, and experimental design — all designed to tease out the causal impact of CTV advertising. So far, we’ve had great success with customers who want to make their CTV spend perform as well as it possibly can without comparing it directly to other “last-touch” types of channels.
Like linear TV, one of the biggest challenges we hear is how to best measure the success of our output of CTV budgets. How do you suggest your customers approach measuring campaign success?
The inherent challenge in CTV is that we are broadcasting advertisements on a household-level device, which can be distinguished from just about every other channel which is an individual device (e.g., phone, computer, tablet). And, as with anything, measuring success relies most importantly on advertiser trust in the measurer. As you mentioned, Branch’s cross-device attribution will give advertisers a fair accounting of TV ads that ultimately lead to action. It allows the advertiser to run an ad on the big screen and then later see that that advertisement led to a download on the mobile device. This is a fundamentally complicated problem to solve. Doing so in a credible and trustworthy way is even more complex.
In thinking about how Moloco has approached measurement, what has allowed the company to thrive in the mobile DSP space is also what has allowed for our early success in the CTV space: our outcomes-driven machine learning algorithm. Rather than attempting to pinpoint what types of content a potential customer might engage with and then uploading audience data, our advanced machine learning algorithms look for patterns associated with installs, conversions, and other advertiser KPIs. This can only be achieved with our measurement partners like Branch to ensure (1) those goals are being met and (2) the advertiser is comfortable continuing to invest because they’re able to accurately measure CTV ROI.
What are the biggest fundamental shifts a marketing team will need to take to start advertising on CTV at scale?
There are a few, but I think the question depends on whether the marketing team has experience in TV (linear or otherwise). If the customer is new to any form of CTV advertising, then the shift will look more like what we talked about in the last question: viewing CTV as a distinct channel with different KPIs and a robust value proposition, albeit different from individual device-based outcomes.
For teams with linear TV or brand CTV exposure, I think a different shift will need to take place to start driving measurable outcomes at scale. Specifically, demanding more from CTV. The promise of connected television has been that we are going from broad targeting (national or designated market area) to hyper-specific household-level targeting. However, so much of the measurement apparatus is geared toward mimicking the linear ecosystem — optimizing toward reach and frequency is an extremely blunt tool if your marketing KPI is to sell more products. Whether the customer sees the advertisement once or 1,000 times, the point is moot if they buy your product.
With that as background, I think the best advice I can give customers is to consider all that CTV can do as a channel — the measurability, the specific targeting, things that never existed in linear TV — and apply that lens to make an impact with their marketing spend.
Say six months out, where do you see the future of CTV going? What should advertisers be planning for?
If we look at the historical lifecycle of each new channel — display, online video (OLV), mobile app, and now CTV — the general trajectory is the same. First, there is nascency and scarce supply; everyone wants to get in at any price. As the market responds to this influx of demand, supply increases dramatically, outpacing demand. With this outsize supply, differentiation on content becomes harder, so marketers increasingly differentiate on performance.
In CTV, we are just starting to see that rapid influx of supply. Starting with the pandemic-fueled surge in time spent streaming, and then buoyed by (1) the emergence of free ad-supported television (FAST) inventory, as well as (2) the shift from subscription video on demand (SVOD) to advertising video on demand (AVOD) by many of the premium streamers — we have seen supply start to expand dramatically. We can expect the next phase of this market lifecycle to force that supply to differentiate on performance and outcomes.
At Moloco, we’re already seeing this shift towards performance and outcomes with CTV advertisers. Whether it’s over the next six or 16 months, marketers will begin to demand more accountability from their CTV spend. We’ve seen unprecedented amounts of money rush into the CTV space in the past half decade, with only marginally more measurement than in the linear world. I believe that for CTV to continue commanding a share of advertiser wallets, it will need to prove its worth.
As the market matures, what are you most excited about, and what are you most worried about?
What worries me the most is that the CTV advertising ecosystem could get stuck in this limbo state where the focus is more on measurability than accountability. With all of the data at our disposal in CTV, the worst thing we can do is settle for the veneer of performance. Big tech is growing at an outsize rate relative to the rest of the open internet because these companies, with advanced machine learning technology, recognize the importance of delivering outcomes. If the rest of the internet can’t align on the importance of accountable advertising spend to drive those same outcomes, I fear that we’ll be quarreling over the pennies that fall off big techs’ dump trucks of advertiser spend.
However, with adversity comes opportunity — CTV is one of the few advertising mediums that isn’t completely dominated by one or two major players: there is fragmentation. Fragmentation of devices, fragmentation of streaming services and distributors, fragmentation of identity solutions. And while that fragmentation introduces complexity, it also ensures that no one player can alter the landscape on its own. This space will continue to be ripe for innovation as well as disruption.
Still have questions about CTV advertising? Branch is here to help.