Author transcript
Sean Nolan: Hi Vivien, thank you so much for joining me today on CX Today. To start things off, could you tell us a bit about the development of Upfluence and how AI has become a larger part of the services you provide?
Vivien Garnès: Yeah, absolutely. So listen, the premise on which Upfluence was built almost fifteen years ago now was to remove the guesswork from influencer marketing, even though the term was not coined at the time. That's very much what we created the company for and what we set out to do. For the following fifteen years we've been working very hard at having the best possible data infrastructure in the industry, and that's why AI has been such a logical next step for us in our development, because we have had all this great data to train our models on and to deliver really great outcomes for the clients as a result.
From then on, I think AI has gone from a small feature, a small functionality within the app, to an essential part of your workflow, basically enabling our clients to achieve great results and great outcomes. So I think as of right now, the greatest shift is from AI adding value sporadically to having something that's fully agentic, that does the work for you, much like an agency would do for you with humans. An agent is doing it now.
Another very cool development that we've seen is an increasing number of marketers seeing marketing tools such as Upfluence as a fully headless experience. They want to connect to the MCP, they want to operate it from Claude or whatever LLM of choice they use. And if they can get away with never logging into the platform ever again, they will happily do that. That's what we've seen, and it's been a fun ride.
Sean Nolan: So interesting. You mentioned so many cool things there, including how the agents can actually play the role of agencies and do the work. It leads nicely to my next question. Where do you now see the balance between human marketers and AI marketers, and what are you seeing as the respective strengths of both of these sides?
Vivien Garnès: Yeah, of course. Interestingly enough, I think my own take has changed in recent months. For the longest time, most people, me included, certainly, viewed AI as a way to speed things up. We all had our own egos and pride, thinking, sure, AI does it faster, so I'm going to delegate that task to an AI, but surely I would do it better, because I'm a seasoned professional, I've done this for a long time. Now what the numbers tell us, having released these features for over a year now, is that there's definitely the factor of speed, but we also see a superior output and superior outcomes.
One example is that campaigns created by one of our agents, which is called Jace, within Upfluence, tend to outperform campaigns made by humans. It's not a critique of our clients, who are very competent marketers themselves. It's just that we're comparing ourselves to an AI that's been trained on the best possible practices across a very large number of campaigns, and knows its stuff.
So for me the biggest shift on this balance was that I used to see this as a way to save time and a way to increase productivity. I think the real unlock is a way to increase expertise and to increase quality as well. In combination, that's a very powerful combination. That being said, I think humans still bring their own taste. If you're the founder of a small brand, or the head of marketing of a brand you've been at for a very long time, you have such a deep understanding of the inner workings of a brand that applying that level of taste, and staying in the loop to make those decisions when it comes to strategy, when it comes to content selection, to influencer selection, even to a degree, it's irreplaceable right now.
But what I find the most exciting is all this menial work being fully automated and taken off the plates of marketers. That brings so much more time for them to double down on what makes us human. For instance, you have more time to actually connect with the creators that you are working with. Maybe invite them on premises if you have something cool to show them, or make events, or just hop on a Zoom call and speak to them and create that relationship so they understand your brand better, as opposed to just shooting them a brief because you're so busy negotiating left and right and approving content and this and that. I think that's going to move the needle meaningfully.
And I don't see AI as a way to reduce the workforce, to let go of people and to make more money. I think using the teams that you have, the talents that you have internally, AI is going to be able to increase productivity dramatically, increase the quality of outcomes, but also free up that bandwidth so you can actually focus on what matters and what makes us human.
Sean Nolan: Definitely. As you say, there's not a lot of point in getting a human marketer to do work that an AI can do, like menial tasks. They're better spent using that time on the stuff that AI can't do, which is those in-person connections, those conversations. I think that's a really strong strategy moving forward, to bring out the humanness of the human marketers and let them excel in that area.
I want to get your perspective, because I know Upfluence works across many different fields, for example influencer marketing, affiliate programs, user-generated content. Putting myself in the shoes of a CMO, what is your advice for marketing leaders and CMOs if they're unsure which strategy, which path to take, in terms of their marketing campaigns?
Vivien Garnès: Good question. I would start with the objectives. What is it you're trying to solve for? Are you trying to generate more sales because you already have great brand awareness? Are you trying to bring more eyeballs into your brand because you have a deficit, maybe relative to a competitor? Depending on what you're trying to solve for, that will dictate what you should experiment with first. Based on that, I would encourage even large brands to start with smaller scale trials, experiments, and look at this very scientifically.
We have a lot of channels where the unit economics tend to deteriorate as you reach scale, but I've never seen the opposite happen. So I think the challenge for CMOs is to leave everybody's egos on the side door here and really focus on having a replicable unit economics for that specific channel. If you start with influencer marketing over UGC and over affiliate because you want to solve for awareness, find that metric that works for you, monitor it, have the infrastructure in place to monitor it properly. So you remove, again, the guesswork from the equation and really focus on having a small program work, maybe 10 creators. Maybe you don't need a tool like Upfluence. Maybe you need a spreadsheet to make it work at first. And once you've really experienced it yourself and felt the pain, once you're ready to graduate to the big leagues, then you can invest in the infrastructure and so on and so forth. But a lot of the time brands throw money at a problem as opposed to trying to figure it out and really understand the inner workings of a specific channel, and I think that's a problem.
So this is how I would approach it. One caveat is that you mentioned influencer marketing, UGC, affiliate. I think a great program that operates at scale has all three in a way that's really intertwined, and that's something you figure out as you go. But there's a really cool play that you can have with something that's very ongoing, always on, when you have these creators producing content: you put marketing dollars into this content to amplify it. That's really when you have a positive flywheel that keeps on going, that generates sales, that generates awareness. And that's when I think you really hit the sweet spot.
Sean Nolan: Definitely. It's interesting. Something we're seeing as well is that when CMOs launch a big campaign, as you say, it can be a very expensive campaign sometimes. You don't have the ability to launch this big campaign and then realize afterwards that actually our messaging wasn't quite right. It's such a huge operation, and then to realize afterwards that we need to do a bit of a U-turn, it's way too late by then. So I think your advice to start small and to test messaging, see what is actually making a big impact with your customers, can be really impactful when you then go on to your larger campaigns and diversify more from there.
Vivien Garnès: And if I can piggyback on that, I think something that too few people realize is that your creators can be your outsourced messaging lab. Creators do this great thing where they appropriate the brand message and put it forward to their audience in a way that they know will work. It might not be your exact perfect marketing message the way you intended it, but it's going to work for the audience. So there's a right balance to strike between how much trust you give to the creator to carry out that message in a way that will resonate with their audience, versus how much control you want to have on this message specifically.
If you work at scale with creators, everybody will put their little spin on your marketing message, and that's a fantastic sounding board to say, okay, this actually works, we can double down on that, we can put more marketing dollars on this. Or maybe it's a new ICP, an ideal creator profile, that we didn't know about and that we need to invest in. It's like your Black Ops team for your marketing message. But you can test at scale for perfectly tight audiences.
Sean Nolan: Absolutely. Leading on from that, about how to get the most value from your marketing campaigns, I wanted to ask about proving ROI, because that's one of the most consistent challenges for marketing leaders. How do we prove the ROI of our marketing spend, whether it's on a new campaign or whether we're adopting a new martech software and we need to prove the value of that? What's your advice on measuring ROI in the current state of marketing that we're in now?
Vivien Garnès: It's funny you should ask, because back in the day I did my doctoral thesis on influencer marketing ROI. There are a few things to consider. Number one, and I'm going to repeat myself from a previous question, I would define what success looks like. There's a very strict method to calculate ROI, which is: what are my sales, what are my costs. But many times there's a lot more that goes into it as well. So I would start by defining success, having access to the right metrics, the right pipeline infrastructure to measure my GMV or my revenue, my client acquisition cost, versus baseline, and I would manage my own expectations, or the expectations of my management, as a result as well. Because maybe they've seen a case study of Gymshark doing a 20x ROI on their influencer marketing campaign, but not everybody's Gymshark.
So I would start with that. Then I would really double down on the setup. There's one way that is probably the most pessimistic way to look at creator ROI, which is to use coupon codes for your attributable sales tracking. If I create a Sean20 coupon code for you specifically, I know that if that code is being redeemed, the conversions are coming from there specifically. I also know from experience that a lot of the time the coupon code isn't being redeemed, because either the tracking broke between the app and going to the commercial website, or there was a redirect at some point and the attributes in the URL got stripped, or there was no URL at all and someone mistyped the coupon and couldn't be bothered to find it again on Instagram or wherever it was. So what I like to say to clients is that whatever ROI you calculate strictly on a coupon basis is the most pessimistic ROI you can look at.
And now we're starting to actually measure properly the halo effect, which is not exactly new. For people who do campaigns on TikTok Shop, for example, they get a ton of eyeballs on their product and they look at the sales and see a big drop between the top of the funnel, when they had all these impressions, all these eyeballs, and the actual conversions. Then they look at the rest of their conversions that were untracked and they see that they're up three hundred percent. A lot of our clients now do this work proactively, and some of them leverage AI to do that. They look at the logs and try to coincide with the dates of publication of some very viral creator posts and say, okay, probably a lot of these untracked conversions could be attributable to that creator. I don't have exact causation here, but it's something to consider when looking critically at the ROI that the campaign has generated.
And then there are all the other factors that a lot of brands and marketers completely ignore. Generating that 2x ROI, how many tens or hundreds or thousands of hours of my team, that I pay fifty bucks an hour, a hundred bucks an hour, how much of their time was spent on that? A lot of the overheads are completely ignored. If you want to be very strict and intellectually honest about the way you look at ROI, that's something you factor in as well, at least to a degree. And obviously the more you factor in technology, the more that number of hours goes down. It's the same with AI. I'm going back to the first couple of questions that you had. It's one of the factors that really alleviates the amount of time that you need to spend on your overhead costs as well.
Sean Nolan: Thank you so much. Definitely, as you say, because there are so many channels now, so many ways for brands to share their content, that traditional funnel idea, things get very complex and things get lost in that funnel. It's not as simple as people might think. So that conversation on ROI is really important.
Taking a step back now and getting your perspective on the marketing industry as a whole right now, what's one common mistake that you're seeing marketing leaders make at the minute, and what's your advice on how to avoid that?
Vivien Garnès: Only one? I'll give you a few, if you can indulge me. I think something we've seen a long time in the making for creator marketing, but I don't think it has caught up to all the marketing channels, is thinking in terms of campaigns. Campaign meaning a finite amount of time but seasonally relevant. I'm going to have my Black Friday campaign, then my Christmas campaign, then my Valentine's Day campaign, whatever makes sense for my industry. And the clients that I find are most successful with creator marketing have always-on programs. They have evergreen, quote unquote, campaigns. They look at creator marketing as a channel as opposed to a series of one-off campaigns.
I think this is very smart for many reasons. Number one, you amortize all your efforts a lot more, because the creators you recruited once, you double down, you triple down on them. And as long as they're ROI positive, all the work you've put up front to recruit them, to create that relationship with them, you can capitalize on that very easily. But there's also the idea of looking at this community of creators that you created for yourself in a very ROI-esque way, meaning all the top performing creators, I'm going to double down on them. All the ones who are sub one ROI, I'm going to remove from the program, and really admit new creators to that program on an ongoing basis. I find this is how the best performing brands operate. So that's the first idea.
The second idea I would have, and it's very intuitive in terms of ROI, is do not be stingy with creators when it comes to the incentives that you put in front of them. I think incentives can be seen as a spectrum. On one end, if you, Sean, are the creator and I am the brand, and I offer you a million bucks for an Instagram post, you're very likely to say yes. This is a very generous and very incentivizing proposal. That being said, I will also shoot myself in the foot in terms of ROI. This is probably going to be a horrendous ROI from the campaign perspective. I don't doubt you have reach on Instagram, but it might be a tall order.
On the other end of the spectrum, if I reach out to you saying, "Hey Sean, I'm Vivien, please talk about my brand. Okay, thanks, bye," you're very unlikely to do it. But if you do it, the ROI is virtually infinite because it costs me nearly nothing. And so obviously the truth is somewhere in between. I think the art here, not exactly the science, is to find a proposal that's enticing for the creator to work with you, but also has a higher perceived value than monetary value. Meaning if I offer you this mic, whose monetary value is three hundred bucks, maybe my production cost is fifty bucks plus shipping, so it costs me 60 bucks. The perceived value for you is 300 bucks. That's how you get leverage.
But there's also the consideration of whether I pay you a flat fee, or do I pay you an incentive on an affiliate basis, so that every time you sell one of these mics, you make money. So you have uncapped earning potential. If you do great, you're going to make bank. Or is it a combination of both? There's really something that needs to be properly tested, because what you want is to maximize the equation between having an incentive that will work great to bring creators over to me, but at the same time keeps a very interesting ROI. And I think a lot of mistakes are being made there.
I touched on it earlier, and I'm not going to go too long, but there's a natural bias to want to control the narrative, especially for older brands or brands in the luxury industry. They are really attached to how they are phrased, how their brand is being presented, and they want to hold the pen for creators and approve everything. I think that's a mistake. It's completely understandable, but I think it's a mistake, because, like we touched on prior, these creators have grown their audience for years, sometimes decades. They know exactly what resonates and what doesn't. So giving them the opportunity to rephrase your messaging in a way that will work, not doing that is a huge missed opportunity here.
And maybe one last one, just to conclude, is to not optimize, to say, okay, I have a channel that works, I'm going to leave it be. I think there's always room for optimization. Culture and conversations change, and the way you need to drive your creators changes. It takes work. You need to understand how to make your incentives evolve over time as well. There's a lot of work that goes into it, more on the human side, to go back to our AI analogy. It's not the menial repetitive work I'm talking about. It's more of a strategic thing. And I think too few brands do that in a way that's really proactive.
Sean Nolan: Thank you so much, really interesting. The philosophy shift there, what you mentioned about always-on campaigns, I think that's really interesting. And I think that's why a lot of brands now, brands you might not expect, are creating TikTok accounts and fun accounts where they're just talking about the brand in general, and about workplace culture and things like that. It's because they want that ongoing engagement. It's not just about this particular season, but about that continuous engagement, keeping those creators who they know work well with the brand on board. I think that's a really great point.
Finally, I wanted to get your perspective on the future of marketing. Do you have any predictions for how you expect marketing to shift over the next five years?
Vivien Garnès: Looking at marketing through the lens and the prism of influencer marketing, which is what I know best, first I would say a lot of people were concerned about creator marketing being completely replaced by AI creators. I don't see that happening at all. I think the psychological mechanics behind influencer marketing, which is the power of trust and recommendation, are here to stay. I think also the risk of social media being completely overrun with AI slop is an existential risk for the social media platforms themselves. We've seen already YouTube taking some steps to protect the creators in that sense and to demonetize AI content accounts. So I don't see this as a threat, and I think the future will very much count on creator marketing to develop brands. So that's the first. I don't know if you can call it a prediction, because essentially I'm saying something will not change.
But I think the agentic piece is going to be huge. I think many, many more marketers will live in their Claude or LLM of choice, will coordinate their campaigns via MCPs across all their tools, all their stacks, and they'll have these stacks work together, which I find very exciting. So maybe what it will look like is they have their evergreen creator program on Upfluence, they have scheduled workflows that connect with creators, negotiate, add them to the mix automatically, review the performance, so they can really spend their time connecting with the creators, inviting them, giving them a great experience. Then work with maybe the Meta MCP to put marketing dollars on the top performing creators and make sure to have this flywheel and virtuous cycle happening.
Maybe the same thing with their UGC play. They have a social listening stream that picks up all the mentions that they get, and mentions that look very cool are added to that amplification loop, and so on and so forth. I really easily could see that happening, because a lot of our clients are already doing that. Even though I've been impressed by the speed at which they adopt the technology, I think this is going to become commonplace for the marketers of the future.
But I think the opposite side of the creator marketing economy will happen as well. I think right now AI companies have been really focused on the brand side and the agency side, Upfluence included. Guilty as charged. I think soon will come the time of creators as well, for the exact same reasons. We want creators to be able to spend more time creating. Usually a lot of these folks are not necessarily super business savvy. They don't want to spend their lives negotiating and reaching out to brands themselves. They don't want to spend their lives invoicing clients. All of that is no fun for them. It's not their genius zone, their flow zone. And so I would very easily see a creator-side agent negotiating with a brand-side agent for brand deals day in, day out, so they can focus on the creation itself.
I think also the creator industry will concentrate more and more ad dollars. I think the combination of top performing viral creator content plus amplification on socials is a winning combination, and we've only seen the beginning of it. I think this is going to become a more and more central part of the playbook for all these brands.
And one last one for the road. I think all this free time that marketers will get thanks to productivity gains, and creators will get, will result in more genuine and deeper relationships between brands and creators. We've seen a little bit of equity deals happen in the creator economy. I don't know if you've followed that, but there's an increasing number of brands that offer a one percent equity package to creators to work with them on an ongoing basis, in a way that's very genuine and authentic. We've seen more and more brands working with creators on a co-creation basis, meaning: we have this idea for a good product, I know you've been very vocal about this very specific niche use case, would you like to create it with us and go to market together? I think this is exactly what we will see more and more of in the future, because I think those types of relationships are more accurate in value for both sides.
Sean Nolan: Thank you so much. I think what you were saying about agents representing creators is really exciting, and there's a lot to look forward to in terms of deeper relationships between brands and creators. We'll definitely be keeping a close eye on it at CX Today, and hopefully we'll talk about this with you soon and see how all these predictions have played out. Really exciting times ahead. Thank you so much for your time today, Vivien.
