How to expand programmatic advertising up the funnel, with TripAdvisor’s Matteo Balzani

Summary

The Digiday Podcast: How to Expand Programmatic Advertising Up the Funnel, with TripAdvisor’s Matteo Balzani

Release Date: December 10, 2024

---

Episode Overview

In this insightful episode of The Digiday Podcast, host Tim Peterson delves deep into the evolving landscape of programmatic advertising with Matteo Balzani, Senior Director of Acquisition and Retention at TripAdvisor. Balzani shares TripAdvisor's strategic pivot from a predominantly lower-funnel, retargeting-focused approach to a comprehensive full-funnel strategy aimed at capturing and creating demand earlier in the customer journey. This shift is crucial for maintaining competitive advantage in an increasingly crowded digital marketplace.

---

News Highlights

1. Omnicom and IPG Merger

The episode kicks off with breaking news about the imminent merger between Omnicom and Interpublic Group (IPG), as reported by The Wall Street Journal. This merger signifies the unification of the third and fourth largest agency holding companies by revenue, potentially reshaping the advertising agency landscape.

• Tim Peterson explains, “Omnicom and IPG are the number three and number four largest agency holding companies by revenue. This merger would position them as the number one agency holding company.” (04:47)

The merger aims to enhance scale and leverage in negotiations with major media players like Disney, Netflix, Meta, and Google. However, challenges loom regarding cultural clashes and potential conflicts of interest among clients. The merger's approval is pending, influenced by regulatory environments and potential opposition from influential figures like Elon Musk.

• Tim Peterson notes concerns about the merger possibly being opposed by key industry stakeholders, stating, “Elon Musk seemed to very much signal he doesn't like a consolidation of power on the advertiser side of things because that could work against his company X.” (12:35)

2. Ad Spending Dominated by Big Tech

Drawing from GroupM's latest ad spending forecast, Peterson highlights a significant consolidation where over half of the anticipated $1 trillion global ad spend is funneled through five tech giants: Alibaba, Amazon, ByteDance, Google, and Meta.

• Tim Peterson remarks, “We talk so much about the duopoly of Google and Meta, or the triopoly of Google, Meta and Amazon. Now you have a quintopoly with Alibaba and ByteDance added into the mix.” (15:26)

This concentration raises concerns about market dominance and the potential marginalization of other players in the ad ecosystem.

3. TikTok’s Potential Ban in the U.S.

A pivotal discussion centers around the U.S. Appeals Court's decision to uphold the ban on ByteDance's TikTok, mandating its sale or removal from the U.S. market by January 19.

• Tim Peterson explains the stakes, “TikTok is pretty huge in the U.S., and so that's still billions of dollars that would effectively be up for grab if TikTok ends up being banned in the US.” (16:24)

The ban not only impacts advertisers but also jeopardizes the livelihoods of countless TikTok creators who have built careers on the platform. The potential void left by TikTok's absence could reshape how cultural conversations and travel recommendations are disseminated, directly tying into TripAdvisor's strategic adjustments.

---

Main Discussion: Programmatic Advertising Up the Funnel with Matteo Balzani

Shifting from Lower to Full-Funnel Approach

Matteo Balzani outlines TripAdvisor's transition from a lower-funnel, retargeting-centric programmatic strategy to a full-funnel approach aimed at capturing demand at earlier stages.

• Balzani states, “We need to find a way not only to capture the existing demand but to create some demand.” (02:50)

This shift involves engaging customers earlier in their travel planning journey, utilizing signals from engagement tools like TripAdvisor's AI-powered travel planning tool and the mobile app to identify and nurture potential travelers before they reach the booking phase.

Challenges in Measurement and Attribution

Implementing a full-funnel strategy necessitates robust measurement frameworks to attribute value correctly across different funnel stages. Balzani discusses the limitations of traditional attribution models in capturing the incremental value of upper-funnel activities.

• Balzani explains, “When you go into the upper funnel, you cannot measure in the same way. So we are actually exploring MMMs and this kind of tools.” (41:43)

TripAdvisor is exploring Media Mix Models (MMMs) and incrementality tests to better understand the impact of their broader advertising efforts. This involves collaborating with data science teams to ensure measurement accuracy and reliability across different campaign channels.

Success with Paid Social Advertising

Balzani highlights TripAdvisor's success in utilizing paid social advertising on platforms like Facebook and Instagram. Initially focused on retargeting, these efforts were expanded to test the incremental value of social ads in driving engagement and bookings.

• Balzani shares their findings, “We were surprised not to see an incremental value because we knew it was there, but the size and the scale of that incremental value…” (36:48)

These tests revealed significant upside, leading to increased investment in social channels despite initial attribution challenges. The ability to demonstrate incremental value provided the confidence to scale these efforts further.

Creative's Role in Upper-Funnel Campaigns

As TripAdvisor aims to engage audiences earlier, the importance of creative content becomes paramount. Balzani acknowledges that upper-funnel campaigns require more tailored and engaging creatives to effectively capture and retain audience interest.

• Balzani emphasizes, “The further you go up the funnel, the more creative becomes important and can make or break the success of a campaign.” (39:17)

This focus on unique, TripAdvisor-specific messaging is intended to differentiate their campaigns from generic, templated ads and better resonate with potential travelers.

Exploring New Channels: The Case of TikTok

Recognizing the cultural significance of platforms like TikTok, TripAdvisor is actively seeking strategies to integrate TikTok into their programmatic advertising mix, despite measurement challenges.

• Balzani remarks, “Our strategy is really to try and be where our customers are, where travelers are. And if travelers are on TikTok, we better figure out how to reach those customers.” (40:19)

This entails developing new measurement techniques and creative strategies to harness TikTok's vast, engaged user base effectively.

Looking Ahead to 2025

Balzani outlines TripAdvisor's roadmap for 2025, focusing on:

1. Expanding Programmatic Channels: Moving beyond Google and search affiliates to incorporate diverse channels like CTV and podcasts.
2. Enhancing Measurement Frameworks: Implementing robust MMMs and incrementality tests to accurately attribute upper-funnel activities.
3. Leveraging AI and Data: Utilizing AI-powered tools to gain deeper insights into customer behavior and optimize campaign performance.
4. Creative Innovation: Developing unique, engaging creatives tailored to different funnel stages and platforms.

• Balzani concludes, “This is not a matter of if, I would say it's a matter of when and how much, because there is conviction at every level in the company that that is the right direction.” (46:13)

---

Key Takeaways

• Scale Matters: The impending Omnicom and IPG merger underscores the importance of scale in negotiating with major media platforms and managing large advertising budgets.

• Ad Spend Concentration: A significant portion of global ad spend is controlled by a handful of tech giants, highlighting the need for advertisers to diversify their media strategies.

• Platform Stability: The potential ban of TikTok in the U.S. could disrupt advertising strategies and creator ecosystems, necessitating adaptability from brands and platforms alike.

• Full-Funnel Strategy: TripAdvisor's shift to a full-funnel programmatic approach demonstrates the evolving nature of digital advertising, emphasizing the need to engage consumers earlier in their decision-making process.

• Measurement Innovation: Accurate attribution and measurement are critical for validating the effectiveness of upper-funnel campaigns, driving the adoption of advanced methodologies like MMMs.

• Creative Excellence: Successful upper-funnel campaigns require innovative and tailored creative content to effectively capture and retain audience attention across diverse platforms.

---

Notable Quotes

• Amy Fox, Chief Product Officer at Bliss: “We want to make sure that the average income of our control group is similar to the average income of our exposed group... they’re neutral and they’re distinct.” (00:00)

• Matteo Balzani, TripAdvisor: “We need to find a way not only to capture the existing demand but to create some demand.” (02:50)

• Matteo Balzani: “The further you go up the funnel, the more creative becomes important and can make or break the success of a campaign.” (39:17)

• Tim Peterson: “If travelers are on TikTok, we better figure out how to reach those customers.” (40:19)

• Matteo Balzani: “This is not a matter of if, I would say it's a matter of when and how much, because there is conviction at every level in the company that that is the right direction.” (46:13)

---

Conclusion

This episode of The Digiday Podcast offers a comprehensive exploration of the strategic shifts in programmatic advertising, emphasizing the necessity of moving up the funnel to engage consumers earlier. Through Matteo Balzani's insights, listeners gain a clear understanding of the challenges and opportunities in expanding programmatic strategies, the critical role of measurement, and the imperative of creative innovation. Coupled with the latest industry news, the episode provides valuable perspectives for brands, agencies, and publishers navigating the complexities of the digital advertising landscape.

Transcript

[Amy Fox]

We want to make sure that the income, or the average income of our control group is similar to the average income of our exposed group, where we're going to do our measurement to make sure that when we do our measurement for our Mercedes car, we're not predisposing A to perform better than B. They're neutral and they're distinct. They're similar in terms of their audience makeup.

[Christina Ko]

That's Amy Fox, chief product officer at Bliss, our sponsor on this episode of the Digiday podcast. Later in the show, custom talks with Amy about how smart holdout groups present advertisers with a new omnichannel measurement method that addresses current challenges in the space.

[Tim Peterson]

Hello and welcome to the Digiday Podcast. My name is Tim Peterson. I'm the executive editor of video and audio at Digiday. And today joining me is not Kimiko McCoy, my normal CO host. Kimiko is on some well deserved time off at the moment, but we have Sarah Patterson, who's our producer, coming onto the show. Hi, Sarah.

[Sarah Patterson]

Hi. Thanks for having me.

[Tim Peterson]

In a way, I guess we always have you. Now we get to have you for everyone else.

[Sarah Patterson]

Thanks for letting me talk this time.

[Tim Peterson]

There we go. Yeah. So Sarah's super instrumental in helping Kimiko and I kind of come up with the news of the week that we'd like to discuss before we get into the interviews each week. So it's great to have Sarah on the show this time around. Sarah, how was your weekend?

[Sarah Patterson]

It was pretty good. I. I went to a live podcast taping because, you know, the job doesn't stop on the weekends. It's my passion, so. So I did that.

[Tim Peterson]

So, live podcast. It's Avian. Is that just like a live talk show or did it feel meaningfully different?

[Sarah Patterson]

So this one is actually different. I. I went to a. I guess it's not taping. They don't actually release the episodes on their feed, but I went to a live showing of who Weekly, which is a pop culture podcast. And so what they do is they basically do these sort of PowerPoint presentations about different topics that are fascinating them right now. It's basically like a big pop culture lecture for like an hour and a half.

[Tim Peterson]

Oh, wow. And does it go out as a podcast too, or it's just in person?

[Sarah Patterson]

It's just a live showing. You have to. You have to buy tickets to see it.

[Tim Peterson]

Oh, wow. Okay. Well, I guess that ties in nicely with our episode this week because we have a live podcast taping that we were running into this week's interview so last week we had the Digiti Programmatic Marketing Summit in Nashville, Tennessee, and I was there. And so one of the sessions that I did was with Matteo Balzani, who's the senior director of acquisition and retention at TripAdvisor. And we talked a lot about Programmatic, but we Talked about how TripAdvisor specifically is going from its programmatic strategy. Really just focusing on, like, retargeting people and getting them. People who are already interested in booking travel to be booking travel through TripAdvisor and instead is starting to look at how do we get people to even be thinking of TripAdvisor in the first place? Basically, the idea of he makes the point. If you sharpen the pencil too much, you run out of pencil. And so they're trying to figure out how do we make sure that we're getting more lead in the pencil? I don't know, maybe it's a mechanical pencil, but then it doesn't need sharpening. But I thought it was a really interesting conversation. The folks at the summit seemed to be really into it because people kept parroting that line about sharpening the pencil through our town halls and other sessions as well. So I'm excited for people who weren't at the summit to be able to hear this one.

[Sarah Patterson]

Yeah.

[Tim Peterson]

And then so before we get to that, we have to get to the big news. And there's bigger than normal news this week. This broke over the weekend. The Wall Street Journal was first to report it. Omnicom and Interpublic Group IPG are merging. Though technically Omnicom is acquiring ipg, but this is basically a coming together of two massive agency holding companies. So just some light, casual news right ahead of the holidays.

[Sarah Patterson]

Yeah. When you say it's a coming together of two massive agencies, I am not as much of an expert in this world as you are. So how explain it in kind of layman's terms? Is this sort of like if Coke and Pepsi were to merge or if CBS and Fox were to become one network?

[Tim Peterson]

Yeah. So Omnicom and IPG are the number three and number four largest agency holding companies by revenue. So it'd be sort of like If Disney and NBCUniversal are WPP and Publicis kind of the 1 and the 2. This is like Warner Brothers Discovery and Paramount kind of coming together. Paramount which owns cbs, Warner Bros. Discovery which owns HBO and a lot of other. All the Discovery networks. So the number three and the number four coming together, which would make them the number one. But there's still a lot that's going to need to be sorted out before this actually closes. Like Omnicom had tried something actually on an even bigger level about a decade ago when it was supposed to merge with Publicist Group, which is in that top two agency holding companies that fell through about a year later in 2014. They couldn't. There was some culture clashes. There was also. They billed it as a merger of equals, but it was not actually a merger of equals. There is a fight for control between the companies. This one, I imagine there's still going to be some of that though, because it's Omnicom acquiring ipg. It seems more clear that Omnicom is the one in control here.

[Sarah Patterson]

Why do they even want to come together?

[Tim Peterson]

Yeah, yeah, that's great question. So I mean, when it's the advertising business, size very much matters. Scale matters when it comes to being on the sales side of things. Scale matters when it's the agency side. So part of this is if you're a bigger agency holding company, you have a bit more leverage when you're doing deals with the Disney's and Netflixes and Metas and Googles of the world, just because you have more money from the client perspective that you can control. And there's been something of a trend towards, especially in the upfronts, agency level deals where instead of a TV network or streaming service doing a deal directly with a Coke or a Pepsi and just having the budget managed at that level, doing it instead with Omnicom or with ipg, who can then say, look, we can commit this money upfront because we'll be able to figure out which clients can spend it. We don't have to have individual clients feel on the hook because a lot of clients these days aren't so comfortable being on the hook for spending tens of millions, hundreds of millions of dollars a year ahead of time committing to that. So having the two agency holding companies come together means that's even more money that they can kind of throw around, which will be important from a negotiation standpoint. In addition to the money, it's also just data that they'll have across their clients and across the companies, which is important from just a targeting measurement standpoint, but then increasingly from an AI standpoint as well. Because what we're seeing with AI, both on the generative front, but also just standard AI, is data, is the oil is kind of like that crude oil that powers everything. So you want to have that data. There's also an AI consideration with this deal of a lot of especially smaller and mid Sized advertisers are starting to see tools from Meta and from Google. These like AI based ad buying tools and AI based ad creative tools as mitigating some of the need for them to have agencies. Now these are smaller advertisers. So to what extent Omnicom and IPG are losing out on a bunch of clients budgets remains to be seen. But that being kind of the trend, could lead to a day in which you have bigger brands saying, I don't really need an agency handling so much of my business as I did in the past. I can do a lot of this stuff in house using this technology from these other companies. And so coming together, so the thinking goes, should help Omnicom and IPG to weather some of that. It's kind of like bundling up during a storm.

[Sarah Patterson]

My other question when it comes to their clients is, and I don't know if these lists are public or if it is sort of common knowledge who, you know, who's working with what agency, but is there any concern as they merge of sort of conflicts of interests or exclusivity issues around, you know, brands that may be working with, with either Omnicom or ipg?

[Tim Peterson]

Yeah, historically the answer would be like, yes, absolutely. Especially the bigger brands were very much, you know, Coke and a Pepsi did not want to be part of the same agency holding company. And I imagine that'll still be the case for some clients. There will be some conflict. There will be some brands that see this as a reason to change agencies or as a reason to negotiate for better terms potentially. Because as much as brands don't like being part of the same agency as their competition, we've also seen a lot of brands kind of split up their work among different agencies. Like I think Amazon did this earlier this year where it used to be with ipg, and then it decided, okay, actually we're going to go with wpp, but we're also going to go with Omnicom. So there may be more of that happening as well, of people taking maybe some business away from IPG and Omicom. But you may also see some brands say, oh, this is going to be the number one agency in terms of revenue once this closes. That could be advantageous to me. I'm actually going to take more of my business here. At least if you're Omnicom and ipg, that would be your hope.

[Sarah Patterson]

So do we think this deal will actually go through? You mentioned that 10 or so years ago there was another attempt at a big merger. I know that the Trump administration may be more friendly to these big mergers, maybe not. But is there any concern that, you know, we're talking about this and nothing will actually come of it?

[Tim Peterson]

Yeah, I keep ping ponging back and forth on that one because like you mentioned, the conventional wisdom seems to be with the next Trump administration, deregulation is going to be very much a part of the firmament. And therefore it's going to be much easier for mergers of this size to happen when you don't have a aggressive and very antitrust mindful Federal Trade Commission in place. So that could make things easier especially. And then also, it seems like some of that concern over control and culture class issues that sank the Publicis Omnicom deal don't seem to be in place here. It seems very much. Again, Omnicom is acquiring ipg. The flip side of that, though, is Elon Musk, which, of course, things always come back to Elon Musk. So over the summer, Musk went against Garm, the Global alliance for Responsible Media, which was kind of this trade group that represented advertisers and was kind of like a centralizing force in a way, although I think force is probably overselling how much influence Garm actually had with advertisers. But I was part of the World Federation of Advertisers. And so Musk seemed to very much signal he doesn't like a consolidation of power on the advertiser side of things because that could work against his company X if you have one organization or one outsized organization being able to dictate spending. And so I could see Musk whispering into Trump's ear to not allow this to go through because this would consolidate power from an advertiser perspective. And that could put someone like Musk or his company X at a disadvantage. That's kind of, I think one of the, it feels like this is going to go through, but that's one of the bigger flies in the champagne from how I'm looking at things. Again, this all just broke less than 24 hours ago. So having not given it a ton of thought, my mind you don't have.

[Sarah Patterson]

A fully formed prediction of exactly what's going to happen. Well, I think this isn't, this also isn't supposed to fully come through until the second half of 2025, I believe, or the deal won't be finalized by then. So I guess we have, we've got some time to figure it out.

[Tim Peterson]

Yeah, time to figure out a lot of things on that front. And so but I mean, there's a lot of money up for grabs as one of Omnicom's and IPG's rivals, Group M, which is part of WPP, the largest agency holding company, if I'm not mistaken. They just put out their ad spending forecast and they had originally thought ad spending was going to top $1 trillion next year. Turns out, no, we got there a year early. It's this year they're expecting more than $1 trillion to be spent on ads around the world across media channels. So it's a lot of money. One thing that I think has really stood out about this report is more than half of that money is going to go to five companies. Alibaba, Amazon, ByteDance, Google and Meta, which is a wild consolidation of spend. That's more than half a trillion dollars that five companies will get. We talk so much about the duopoly of Google and Meta, or the triopoly of Google, Meta and Amazon. Now you have. What would that be? A quintopoly.

[Sarah Patterson]

Quintopoly, I think. Yeah.

[Tim Peterson]

With Alibaba and ByteDance added into the mix, I guess. You know, one thing that will remain to be seen is, is ByteDance still getting that share a year from now if it's no longer available in the US which seems like a very real possibility now? By that, I mean, on Friday, before everything else kind of hit the fan over the weekend, on Friday, U.S. appeals Court in Washington, D.C. upheld this decree that ByteDance, which owns TikTok, will have to either sell off TikTok or no longer let it be available. It'll be banned in the US this was something that's been in the works since, you know, spring of this year, but with the election of Donald Trump, or in a way, reelection of Donald Trump.

[Sarah Patterson]

Yeah. What are we calling it? Are we calling it a reelection? Is it just like. Does it reset after it feels like the Biden one?

[Matteo Balzani]

Yeah, I think it's like, not technically.

[Tim Peterson]

A reelection, but I do find myself mistakenly saying re election. So. But anyways, with Trump coming back into power, the idea was, oh, is he gonna be. Is he gonna be the saving grace for TikTok in the U.S. is he gonna say no, like, TikTok can stay because he had started using TikTok during his campaign. As it stands right now, it's kind of not going to be up to him. Or at least the deadline for ByteDance to sell TikTok or TikTok to be banned is January 19, which is before Trump will take office. Now, TikTok or ByteDance has appealed the ruling and asked for a delay. Unclear if it'll get it. But if TikTok. We've been talking enough about this for, well, really the past few years, ever since Trump had originally looked to ban it.

[Sarah Patterson]

Wasn't he the one who originally was gonna ban it, but now he's also the savior.

[Tim Peterson]

Ban it or have it be sold to like Larry Ellison and Oracle were kind of the frontrunners at that time and nothing came to pass there. But that's a big vacuum both in terms of just people's time and how they spend their time. I will have a lot of time free it up in my day if TikTok goes away. But then as we were just talking about when it comes to ad spending, because, okay, yes, ByteDance's advertising business goes way beyond the U.S. but TikTok is pretty huge in the U.S. and so that's still billions of dollars that would effectively be up for grab if TikTok ends up being banned in the US as looks like a very real possibility now, more so than at any other point.

[Sarah Patterson]

And not just the billions of dollars up for grabs from the advertising directly to TikTok, but then also the money that TikTok creators have started to earn, both, I guess, from the TikTok creator front, but then also from brand deals. Like there are now people who've built entire careers around this app who have left their other jobs in order to do this app full time. And now in the blink of an eye, it could all go away.

[Tim Peterson]

Yeah, yeah. Although I think for a lot of those people, if they haven't already been diversifying to other platforms, they need to get new agents and new managers because one diversification of platforms has been really important and a lot of creators have been mindful of it for a decade at this point. Like I remember 2014 at VidCon, which is like the annual creator convention, talking with a lot of YouTube stars at the time who were looking at Facebook, had just started opening up the native video. And so there was a lot of conversation about like diversifying beyond YouTube, not wanting to be too dependent on YouTube, even though YouTube had a strong revenue sharing program and there are brand deals, there are opportunities there. But just like still an algorithm driven platform, you don't want to be beholden to an algorithm for your livelihood. And over the years we've seen that expand to Instagram, including the ill fated IGTV and Snapchat and you know, Twitter now X as well as TikTok too. So it has. There have been all these other platforms and again, when there was the first threat of a TikTok ban. A lot of the conversation among creators and creator representatives was yeah, need to make sure TikTok isn't the only outlet to audience and to revenue. And so yeah, I think if you're a creator and at this point you're like, oh shoot, it's actually going to happen, I'm screwed probably because you shouldn't have waited this long.

[Sarah Patterson]

Well, yeah, I mean I think about vine, which is very similar. It was another short form video scrolling app that also suddenly disappeared one day. And that is who is responsible for Jake and Logan Paul. And they I think are two examples of people who very successfully diversified their, their content and were able to spin it out to other places.

[Tim Peterson]

Yeah, and notably in their case and a lot of other, you know, cases they pivoted especially to YouTube. Like Logan Paul used to do daily vlogs on YouTube. So I would expect that's already been happening with TikTok stars. Like I remember on this show we had Alicia Marie and Remy Cruz on this would have been spring of 2022, I think like April 2022. And they talked about at that time hearing from a lot of up and coming. So Lisa Marie and Remy cruz are like OG YouTube creators. These are big YouTube creators that have been around for a long time and have been very successful, albeit are still like pretty young. And so a lot of TikTok creators who I would assume grew up watching them and as they became creators themselves may have modeled themselves off Remy and Alicia. Remy and Alicia were saying on the show again, this was 2022, that they had TikTok creators asking for them for advice on diversifying to YouTube. So what'll be interesting here is to what extent does YouTube pick up a lot of these creators who are losing out on their TikTok presences? Assuming this all goes through like with the way December's going, who knows what's gonna happen by the end of the year.

[Sarah Patterson]

Honestly, I'm not Super against a TikTok ban. I maybe need it to be forcibly removed from my phone and maybe I'll like pick up a book again or something. So I don't feel very strongly about it not happening. I would maybe welcome it a little bit.

[Tim Peterson]

Okay, I'll miss it. I think other than YouTube, it's been the social platform that's probably I've enjoyed using the most and hasn't been. Definitely I waste too much time on it, but I don't feel like all of that time is a waste when I'm on it.

[Sarah Patterson]

Yeah, I am. I I am curious to sort of see how the culture at large would then change without it because I do feel like so much of what drives our cultural conversations come from TikTok. So much of what drives how even like news organizations present their information is driven by TikTok. So it's interesting to me to think about how culture at large changes. If this one website that has only been around for five or six years suddenly goes away, it also would fundamentally change the way I think a lot of people would get recommendations for things, including travel. I know I use TikTok a lot to look at new places to go to, new hotels to stay at, et cetera. Which I think brings us back to our guest for this episode.

[Tim Peterson]

Yep, matteo Balzani from TripAdvisor. We don't talk about travel recommendations, but we do talk about how TripAdvisor is using programmatic channels to get people to book travel through its platform and how it's starting to reevaluate its use of programmatic to not just get kind of the low hanging fruit of customers, but to get people into the flow much earlier. So go to that conversation now. Thanks for joining me, Sarah.

[Sarah Patterson]

Thank you.

[Tim Peterson]

We are going to take a quick break to hear from our sponsor and we'll be right back.

[Christina Ko]

I'm Christina Ko, senior editor at Custom Digiday Media's In House agency. In this podcast Interstitial Story sponsored by Bliss, we speak with Amy Fox, Bliss's Chief Product officer, about how smart holdout groups present advertisers with a new omnichannel measurement method that addresses current challenges in the space.

[Amy Fox]

I mean the key foundation to measurement is having some kind of a control group that you can use to then baseline any kind of lift or any kind of defending of a decline against that's like the heart how all measurement has been done. Now in the past there have been two kind of core ways of splitting an audience into control and test and control. And the most common one in the digital space has been to take a bag of IDs or a bag of cookies and just take 10% of them and put them in a holdout group. Of course the downsides of that is it requires IDs to be present, of which there are very, very few.

[Christina Ko]

Nowadays as many advertisers shift away from relying on cookies and various IDs due to several downsides related to them, they need alternative measurement methods. With this, many are looking to geography as a guide.

[Amy Fox]

Now geographical A B splits, it's definitely not something new, it's something that's been done by the industry for a very, very long time. What we're doing with smart holdout groups is we're still using geography and we're using zip codes critically as the thing that we spin out our control group against. But we're trying to be as smart as we possibly can be. And we're leveraging all of the data that we know about zip codes. We're leveraging the data we know about, the people that live in different zip codes, what kind of consumption habits they have, what kind of spending habits they have. So all the data partnerships that we have in our platform, we leverage that data to try and take an audience that's been built and let's say the audience is made up of, I don't know, 30,000 zip codes across the US and when we spin out a control group, we find isolated clusters of zip codes spread throughout the full audience that are balanced mathematically in terms of their makeup based on a factor, a given factor. So a really good default factor that we tend to use is income. So we want to make sure that the income or the average income of our control group is similar to the average income of our exposed group where we're going to do our measurement to make sure that when we do our measurement for our Mercedes car, we're not predisposing A to perform better than B. They're neutral and they're distinct. In fact, they're similar in terms of their audience makeup.

[Christina Ko]

While a B testing isn't new, the smart holdout group method provides a more strategic approach for advertisers, ensuring they're conducting a sound experiment and are well set up for feedback. Future campaign measurement studies to make it.

[Amy Fox]

A bit easier, we've tried to bucket three or four things that we can we think of that work really well. So for example, brand surveys. Super simple technique that's been used for years and years and years to measure brand opinion that works really well with this framework. Instead of trying to serve a survey to someone who's, you know, an ID or a cookie that's seen your ad, you're basically serving the survey in these zip codes that have seen exposed to the advertising and then not also serving it in the in the zip codes held out as your control and you're effectively actively trying to monitor the difference in the response. If you're a car brand and you are trying to get people to sign up for a test drive or brochure, most of the time you'll be capturing details about that customer, including, for example, where they Live. Now, you can't share all that personal information for advertising, but you can look at aggregated customer sign up information information and you can start, start to see if you can see kind of pattern analysis between abs and holdouts. So for me, this is sort of a version of using that kind of client econometric data that makes it really exciting.

[Christina Ko]

You've been listening to Amy Fox, chief product Officer at Bliss, our sponsor, on this episode. And now back to the Digiday podcast.

[Tim Peterson]

Welcome, Mateo.

[Matteo Balzani]

So it's December, the end of the year. I feel like this is a good time for everyone in all facets of life to kind of be reassessing how things are, making plans for 2025 when it comes to TripAdvisor's programmatic strategy. How would you describe the strategy right now as you, I imagine, are thinking about how you want to evolve it next year?

Yeah, I think that the approach that we have taken until now in the last. I've been a TripAdvisor for three and a half years, but I would say that's the same approach that we had since 2019, since we started to recover from the pandemic. And the restrictions related to the pandemic is really to focus on capturing demand as much as possible and be as efficient as possible and really focus on growth but at a profitable rate. And of course we will not change that approach in terms of everybody wants to be profitable. But I think that right after the restrictions were lifted, of course demand for travel at some point was a zero and then very quickly, even up to 100, almost overnight. And so the focus was really on capturing all the pent up demand that was there. But at some point you continue to optimize, you continue to restrict your audience, you continue to focus on the most profitable pockets and you kind of optimize yourself to the ground because competition is increasing, CPCs are increasing. And so we need to find a way not only to capture the existing demand but to create some demand. So I think that we have been on a path in the last year with some tests, but I think that next year will really be when we try to go. I don't want to say all in, but we are being more aggressive in pursuing a full funnel strategy. I would say that in the last years really we focused on the lower funnel with some attempts to go mid funnel. We had some test at the top of the funnel, but not in a way that would impact the overall performance was more like to understand what works and what doesn't. And I think now we Have a good plan in place to focus on deeper engagement with the customer and not only to look at the lowest part of the funnel, but look at other signals that we get in terms of engagement with our trip planning tool, in terms of using the app in terms of community engagement. Of course, TripAdvisor is powered by our community and so those signals will help us expand our funnel and really try new channels and new audiences that before, because of this focus on profitability were more challenging. But I think with that comes new challenges. Right. Because of course the way you measure mid funnel and upper funnel cannot be the same as lower funnel. But it's something that you have to explain also to your CFO and to your leadership because you go from a place where you say, I'm spending X million dollars and I'm getting back hopefully more than X million dollars right away.

As opposed to like, and I got this brand list for you.

Yeah. So there is the brand lift part, which is difficult to sell. But also in terms of attribution and measurement, you need a different framework. Again, you don't want to hold the other channels to the same standard because of course it cannot be right. When you're capturing somebody who's been on your website and give you a signal that they want to book a hotel and then you retarget them. Clearly that will work to some degree. In other areas, you need to create a different mindset, which doesn't mean that you spend money and you don't want to measure it and you just ask for a blank check. Actually the opposite. But you need to, to explain how will you measure and what will you measure, what you need to actually prove it and then also I think, clarify how you will see that in the P and L. Because of course you will not see that impact the day after. You will not see it one week after. You will need to have a different time frame where you evaluate if something works or not. Versus with search, let's say you have a one week, two weeks attribution window. And if it doesn't work, you know right away, and if it works, you know right away, you can invest and with other channels, you know, again, it's, it's a path that you have to go through and kind of a journey not only with your team, but with the, with the whole company and with leadership. And I think that now we are really at that turning point where we have a good, I think we have a good vision of where we want to be and we need to put that in, in practice.

Okay, what Brought this about because I imagine it could make life very easy for you to just be like, we're going to focus on lower funnel. We can measure results. We know this profitability pocket, as you put it, and you just coast.

Yeah, I think that, you know, I don't see those things as either or. So it's not that to go up the funnel, you need to take away from the lower funnel. We will actually, hopefully what we will see is a virtual cycle where you are able to engage more people, bigger audiences, get new signals, and that will make lower funnel work even better because then you will have more data, you will have more opportunities to retarget customers, you will have more actions that you track that are valuable for you. So I think it's not so much giving up something, it's more adding to what we already have. And again, yes, it is more challenging. But I think that over time, in order to compete and continue to grow, you need to expand your funnel because again, otherwise you risk to optimize yourself to the ground and run out. If you continue to sharpen a pencil, at some point you run out of pencil and you get to the.

Were you starting to see that?

Not yet, but I think we want to get ahead of that to some degree. I think that travel is very competitive. We have partners and competitors that spend a lot more than we do. Of course they're a lot bigger. And so in order to be competitive, I think we need to also leverage our unique opportunity to engage with customers in different ways that are not just purely transactional. For example, again I mentioned our travel planning tool, which also now is powered by AI and we have seen a lot of engagement and we see that customers who engage with that tool are 15 times more valuable. Of course it's a smaller portion for now, but that's a signal that we have that maybe our competitors do not have. And then in the past we haven't really focused too much on the app. Well, now the app is becoming more central to the strategy and again we see of course app users not surprisingly, being much more valuable than non app users. So again, that's another signal that we want to use to expand our reach and really go to even more growth. Again, the profitability aspect remains. It's not that we are saying we don't care about profitability, it's more like how you define profitability probably changes because it's more in the longer term, you look at lifetime value of the customer and you look at the longer horizon for your tracking and attribution.

[Tim Peterson]

Got It. Okay.

[Matteo Balzani]

And then you mentioned having done some mid funnel tests this year in anticipation of making this bigger shift next year. Can you tell me more about what that.

Yeah, we tried multiple things. We tried podcast advertising. We tried CTV as well a little bit. We did a brand campaign in Q3.

And why would you characterize like podcasts CTV? I feel like oftentimes those would be characterized as upper funnel, not so much mid funnel.

Well, the way we look at them is really they were still. I think it depends on kind of the message and the goal that you have. So the way we looked at those was more from a performance standpoint. And so I think that we characterized them as mid funnel because they are not as low a funnel as search or Google hotel ads, which is the lowest of the funnel where a person already selected hotel and the dates and they're really ready to book. That is more in the middle. And it was not brand advertising, but was also not lower funnel. Again, I don't know if there is an official description of mid funnel versus upper versus lower, but the way we looked at that for us were, let's say, newer channels that we expected to behave in a different way compared to again, search affiliates and Google hotel ads. But then I think that the area where we saw the biggest upside was on paid social, social ads. And again, this is.

Were there any platforms in particular?

This was specifically on meta, Facebook and Instagram. Yeah, Facebook and Instagram. And I think that was very important for us because before this test, of course, we were active with social ads, but mostly on a retargeting basis. Again, we were considering that lower funnel and we knew that the attribution kind of penalizes that channel compared to, again compared to search or other channels. Our attribution model, multitouch attribution model doesn't look at Post View, which is a big component. And we knew that there was more value, but it was difficult to prove. And so what we did was really to run an incrementality test based on specific DMAs. And we were surprised not to see an incremental value because we knew it was there, but the size and the scale of that incremental value. Value which actually not only convinced us to invest more, but also led to a second test where we tried to figure out the level of the point of diminishing returns, increasing the spend pretty significantly up to the point where we saw revenue and profit going down. And so that opened the door to other similar channels and tests that again, we did some of it. But we are also planning to do more. Because before we were over indexing on Google and this type of channels and now we know that there is a value there that you do not necessarily see in your weekly report because what you see is probably the draw as goes down because you measured the incremental value and not the attributed value. But then if you are convinced that the test is set up properly and the value that you recognize is, is real, then gives you permission to invest a lot more in that channel. The challenge remains that you need to explain how that works because again, if you look at the report, you will not see that right away. And so that value is not necessarily attributed to that channel. Maybe it's direct, maybe people will go to the app, people will go to Google and search for your brand, but it's not something that you can pinpoint and say, hey, see, I spent 50% more and I'm getting, you know, 75% more.

Okay, and so you found paid social like worked pretty well when it comes to the midfield test, were there any components like whether they're channels or any other aspects to the approach and the execution in terms of things that you found that like don't work or that you would need to fine tune?

I think that the biggest aspect that we need to explore more, I would say is the creative because again, being focused on the lower funnel, most of the creative was really kind of basic templatized ads, maybe focusing on the specific hotel or experience that the customer looked at or similar. And I think that the further you go up the funnel, the more creative becomes important and can make or break the success of a campaign. And so I think that one of the areas where we want to invest more and we are kind of gearing up towards is, is being able to create unique messages that are really tripadvisor specific, like kind of leveraging our strength versus just doing templatized experiences like anybody else does. And I truly believe that that can open up doors to other channels that maybe we haven't even tried today because probably we deem them as not very efficient.

Any that you think may be the case for.

I think that, you know, Everybody talks about TikTok and you know, it's difficult to ignore. I don't think we figured out how to deal with TikTok either from an organic or from a paid standpoint. We are present there and we have organic content but you know, there are 100 billion views for travel related content on on TikTok at any given time. That's a lot. It's difficult to ignore. And so I think that again, our strategy is really to try and be where our customers are, where travelers are. And if travelers are on TikTok, we better figure out how to reach those customers, Those travelers on TikTok, in an efficient and effective way. And so I think that again, maybe a year ago, two years ago, we wouldn't have even considered that because how do we measure it? There's no click. It's not the way we are used to. But now I think that again, this experience and this journey towards uppering the funnel is allowing us to consider that and try to figure out, okay, what should we do? But I think there is conviction that somehow we need to be there.

You've mentioned measurement a couple of times now. Especially as you're looking to now go fuller funnel with the programmatic strategy, does that mean you're having to make any big changes to your measurement system?

Yeah, measurement is really one of the key aspects. And I think that again, in the past it was a lot easier and we have our own multitouch attribution. Maybe it's not perfect. I don't think a perfect attribution exists. It's not a thing, but was kind of reliable and it was apples to apples. I think that as you go into the upper funnel, again, you cannot measure in the same way. So we are actually exploring MMMs and this kind of tools. But also I think that incrementality tests are really what have been driving success for us. And again, the methodology is something that we can discuss for hours. How you do it, how you not do it, and how accurate it is. Probably not all the time will be 95% statistically significant. But as long as we have the signal, I think that we have the confidence to go in that direction. So definitely, as a company, we are gearing up for a new measurement framework that will be different because you will have still lower funnel. Again, we are not getting away from that. But then you have these other channels and again, the further up you go, the more challenging it becomes. When you go to linear TV and all this stuff, of course, it's something that we've done in the past many years ago, but we haven't done for a while. And we need to see how that works specifically for TripAdvisor and how we measure the impact. Again, there's no click, there's no direct attribution, there's no cookie. Well, in some ways there are cookies. But you heard, we talked about already Google and third party cookies not going anywhere.

Well, there's The IP address for ctv, but even that's un reliable.

Yeah. So again we are, I would say that measurement is probably the biggest challenge that we have because kind of to some degree new territory. And I think that finance and they are used to that level of reliability and consistency that is a lot more difficult as you go up the funnel.

And medium mixed models. I feel like a few years ago if we were talking about MMMs, we would be talking about, you'd be from a brand that's been around for 50 years and has media mix models that have mothballs on them. It feels like in the past two years everyone's getting into media mix models. I also know media mix models have gotten a lot more affordable for companies.

Yeah, absolutely. And I think that there are a lot of options out there. There are open source options in some ways. We have a very strong data science team and we consider even build that internally. But we are definitely looking for the right platform and the right partner.

What specific are you looking for?

I think that what I have observed is that in general the story and the narrative is the same for everybody. And at the end of the day it boils down to the technology and the reliability of the technology and how they use technology, the data. And so I think that as we evaluate partners, we really want to make sure that if you read, if you go through the slides, everybody will say that it's accurate. Everybody will say that you can track this and that, but we need to go two or three levels deeper. And that's why we actually do that together with our analytics and data science team. It's not a marketing initiative specifically, it's a company wide initiative because of course it affects the overall performance of a company. And so we want to make sure that our data engineers are comfortable with that, our data science team is comfortable with that, analytics is comfortable with that. And really every company is different, has different data structures, have different challenges. And so we need to find something that is right for us, that might not be the same, that is right for another company. And so it's not about a vendor or a provider being good or bad. Bad is more. Is that the right solution for, for us and how we are structured.

When do you think you'll have that answer on whether is the right.

Oh, we'll have it hopefully soon. We want to move fast. So I think it's. It's something that is. Tomorrow? Yeah, yeah, Half an hour? No, I think that.

Do you think like you'll have it in place by the end of this year? The Answer not.

I think that I would say Q1, probably end of the year. You know, we're already, what is it? December 3rd. But I think that definitely is something that we're looking for in Q1 to have figured out.

And then how long do you think it'll be until you get it in place?

Again, we have plans to move as quickly as possible. And again, it's not like we are going from 0 to 100 right away. We've been on that path. We have already some experiences that, you know, led us to. To have some convictions of what works and what doesn't. And we are kind of expanding and continuing that journey. But I think that, you know, in the first half of the year, we'll be. Will be very active, I would say.

Yeah. So, you know, Q1, you'll have the decision on the what does the rest of your. Your roadmap look like for 2025 when it comes to. To expanding Programmatic Full funnel?

I think that what we want to do is to use Q1 and Q2 to figure out what works and what doesn't, make sure that we have everything in place, and then based on the results, then we'll figure out which direction we want to go. Again, this is not a matter of if, I would say it's a matter of when and how much, because there is conviction at every level in the company that that is the right direction. And again, we have plenty of indications from our previous campaigns, from the data that we have that we have permission to go in that direction is more to figure out. Okay, we don't want to go all in right away. We want to find the right path. But, you know, it's something that is happening and, you know, as soon as we find something that works. The good thing about TripAdvisor, I would say, is that we are quick to move when we find something that works. We constantly experiment. We are, we have very much like a testing and experimentation mentality. So when we have a win, then it's easy to build on that win. And then if something doesn't work, we'll try something else. And we are used to that across the company. Even for lower funnels is the case. We try new vendors, we try new channels, and I think would be the same approach with the challenges that we described. I think the path is the right one.

[Tim Peterson]

Got it.

[Matteo Balzani]

At least you say you move quick. We covered a lot of ground in this conversation.

[Tim Peterson]

So, Matteo, thanks so much for taking time.

[Matteo Balzani]

My pleasure.

[Tim Peterson]

Thanks for listening to this episode of the Digiday. Podcast. If you enjoyed it, please leave us a rating and a review on Apple Podcasts, Spotify or wherever you're listening. Get more from Digiday with our daily newsletter sent out each weekday morning. Visit digiday.comnewsletters to sign up Up.