WEBVTT 00:00:00.640 --> 00:00:04.480 Sonia, thanks for joining us at the Future of Advertising panel. 00:00:04.480 --> 00:00:07.840 You are the Chief Business Officer of Moloco. Tell us a 00:00:07.840 --> 00:00:09.320 little bit about Moloco. 00:00:10.000 --> 00:00:14.240 Thanks, Ernesto. Great to see you. Moloco actually stands for 00:00:14.240 --> 00:00:18.440 Machine learning company and what we do is we help companies 00:00:18.440 --> 00:00:22.440 bring ads that are engineered for outcomes. How do you do 00:00:22.440 --> 00:00:26.800 that? We do it in two ways. One is around personalization where 00:00:26.800 --> 00:00:33.280 we figure out which ad to put in front of which user that is that 00:00:33.280 --> 00:00:36.880 will drive the conversion. And the second thing that we do is 00:00:36.880 --> 00:00:40.160 because we use first party data off the advertiser, we know 00:00:40.160 --> 00:00:41.480 exactly what the value of the 00:00:41.480 --> 00:00:45.040 predict, the predicted value if the user is going to be and we 00:00:45.040 --> 00:00:47.640 use it to price what the ad should be. What price should we 00:00:47.640 --> 00:00:51.360 pay for the ad in a way that drives better outcomes for the 00:00:51.360 --> 00:00:51.920 advertiser. 00:00:52.680 --> 00:00:56.280 Interesting. What's the biggest challenge that you face in 00:00:56.280 --> 00:00:59.400 working with the existing customers or finding new 00:00:59.400 --> 00:01:00.040 customers? 00:01:01.160 --> 00:01:02.360 You know, it's a, 00:01:03.320 --> 00:01:06.280 as they say, a lot of the proof is in the pudding. When we go 00:01:06.280 --> 00:01:10.360 say performance, I'll give you an analogy, right? You look 00:01:10.360 --> 00:01:13.960 at two cars, they look exactly the same, but they may have very 00:01:13.960 --> 00:01:18.320 different engines. One could be V8 or one could be V6. The only 00:01:18.320 --> 00:01:21.600 way you know, if it works is if you drive it. So the biggest 00:01:21.600 --> 00:01:26.280 challenge with new customers is like, hey, let us prove to you, 00:01:26.280 --> 00:01:29.560 give us some data. Let's prove to you is that's one of our 00:01:29.560 --> 00:01:32.040 biggest challenges. How do you articulate? Because nobody wants 00:01:32.040 --> 00:01:33.440 to know the design of the engine, 00:01:33.640 --> 00:01:36.040 they just want to know how fast it goes. So, so that's one 00:01:36.040 --> 00:01:39.280 challenge with our new customers, with our existing 00:01:39.280 --> 00:01:43.880 customers. One of the hardest things about performance 00:01:43.880 --> 00:01:46.120 marketing or marketing is sustainability of your 00:01:46.120 --> 00:01:49.480 performance, right? And this is where we work really hard 00:01:49.480 --> 00:01:52.640 because our machine learning models are based on deep 00:01:52.640 --> 00:01:55.840 learning, where we're constantly learning from every single ad 00:01:55.840 --> 00:02:00.120 campaign that's running. And the key thing about machine learning 00:02:00.120 --> 00:02:03.720 is because consumer habits are also changing, how do you keep 00:02:03.720 --> 00:02:04.360 your models 00:02:04.440 --> 00:02:07.720 updated, right? So that's the second challenge, which is how 00:02:07.720 --> 00:02:10.240 do you keep sustainable performance happening when the 00:02:10.240 --> 00:02:13.561 world around you continues to change? And like, for example, 00:02:13.561 --> 00:02:14.161 even 00:02:15.761 --> 00:02:19.681 we just went through COVID where a lot of people were sitting at 00:02:19.681 --> 00:02:24.361 home doing things online. The LTV models were built on that 00:02:24.361 --> 00:02:24.881 phase. 00:02:25.561 --> 00:02:28.881 Now we're two years out of COVID. The LTE models are very 00:02:28.881 --> 00:02:32.121 different, right? So you need to keep up with the models with the 00:02:32.121 --> 00:02:34.641 real time data to make sure you're driving the optimal 00:02:34.641 --> 00:02:38.881 outcomes. And tell us a little bit about how easy or difficult 00:02:38.881 --> 00:02:43.201 it is to get the data from your customers also get third party 00:02:43.201 --> 00:02:47.561 data privacy concerns. How do you deal with that? Yeah, it's a 00:02:47.561 --> 00:02:50.881 great question. The good news is, for example, in the app 00:02:50.881 --> 00:02:55.401 ecosystem, we have a pretty mature ecosystem of attribution 00:02:55.401 --> 00:02:56.321 providers 00:02:56.441 --> 00:03:00.841 which are privacy centric, which are safe. So the advertisers 00:03:00.841 --> 00:03:04.641 give them the data and we work with those third party partners 00:03:04.641 --> 00:03:08.281 to ingest that data and drive specific outcomes, right? So we 00:03:08.281 --> 00:03:11.161 don't create custom integrations with every customer. We actually 00:03:11.161 --> 00:03:14.121 work with the ecosystem that's privacy friendly and drive those 00:03:14.121 --> 00:03:14.921 specific outcomes. 00:03:15.801 --> 00:03:16.201 Great. 00:03:17.121 --> 00:03:18.241 What brings you to Cannes? 00:03:19.681 --> 00:03:22.521 You know, I've been in advertising on and off for about 00:03:22.521 --> 00:03:26.481 10 years. I've always found there's a religious debate 00:03:26.481 --> 00:03:30.201 between brand marketing and performance marketing. I think 00:03:30.201 --> 00:03:35.361 we've reached an age where there is a way that both can coexist, 00:03:35.361 --> 00:03:40.881 but with better measurement. And I just wanted to see, you know, 00:03:40.881 --> 00:03:45.481 how is marketing kind of evolving over time? Are people 00:03:45.481 --> 00:03:49.201 still using sample based techniques for measurement 00:03:49.201 --> 00:03:49.641 versus 00:03:49.721 --> 00:03:53.481 sensors based? And just wanted to see what where the wind is 00:03:53.481 --> 00:03:57.001 blowing because typically the wind is like, ohh, brand is 00:03:57.001 --> 00:04:00.281 moving to performance or people are sick and tired of last click 00:04:00.281 --> 00:04:03.521 attribution and they're moving upper funnels. So I just came in 00:04:03.521 --> 00:04:07.801 to say where is the wind blowing this time? But I feel like there 00:04:07.801 --> 00:04:12.601 is a real question on fundamentals and how we can make 00:04:12.601 --> 00:04:15.081 the brand marketing more effective as well as performance 00:04:15.081 --> 00:04:19.001 marketing more effective. Yeah, it is a religious debate for 00:04:19.001 --> 00:04:19.681 sure. 00:04:20.121 --> 00:04:26.361 You just attended our panel and we had Disney, Coca-Cola, Known. 00:04:27.242 --> 00:04:30.922 What did you take away from the panel? Sure. First of all, I 00:04:30.922 --> 00:04:34.002 thought it was a great panel, A great topic and really good 00:04:34.002 --> 00:04:38.802 frameworks explained by BCG too. Two things that struck me, one 00:04:38.802 --> 00:04:43.282 was when the likes of Disney and Coca-Cola say that they are as 00:04:43.282 --> 00:04:46.602 much of a technology company as they are a consumer products 00:04:46.602 --> 00:04:49.442 company. Was refreshing to see and I think more people will 00:04:49.442 --> 00:04:53.042 start adopting that mantra also being a technology company and 00:04:53.042 --> 00:04:56.442 measuring everything. So one which is really good. The second 00:04:56.442 --> 00:04:57.762 thing is when 00:04:57.762 --> 00:05:01.402 Vijay from Coca-Cola talked about there's a difference between 00:05:01.402 --> 00:05:05.082 size and scale, it was great to see. And to me, the parallels 00:05:05.082 --> 00:05:08.642 that came to me was in brand marketing, you know, there's a 00:05:08.642 --> 00:05:12.202 difference between reach and frequency versus what outcomes 00:05:12.202 --> 00:05:16.122 is it driving. And I think a lot of people confuse the two. So it 00:05:16.122 --> 00:05:19.802 was really refreshing to get that insight from Vijay as well. 00:05:20.562 --> 00:05:24.642 Great. Sunil, thank you for stopping by the BCG Cabana here at Cannes. 00:05:24.642 --> 00:05:25.762 Thank you.