Connect with us


Alex Song Joins Startup Savants to Discuss His Data Tech Startup Proxima

Adriaan Brits



This episode of Startup Savants features Alex Song, the founder of Proxima, a data intelligence technology that provides customized data sets that will help businesses acquire customers more efficiently and effectively on social media platforms. 

Talking with podcast hosts Annaka and Ethan, Alex discusses the changes in iOS 14.5 that are making life more difficult for advertisers, along with lessons learned in founding his company and his long-term goals for the business.

Tell us about Proxima, the problem it’s solving, and then how it is solving that problem.

“Proxima is a data intelligence technology where we have proprietary access to over 60 million unique personas such that we’re able to build customized data sets to then improve the targeting that digital businesses can have on platforms such as Facebook, Instagram, TikTok, et cetera. We’ve been able to help combat some of the challenges caused by iOS 14.5, arguably the biggest problem in technology over the past 12 months. And we’re very proud to be helping over 50 clients navigate these troubled waters to find more efficient acquisition in a very challenging time.”

Proxima claims a return on advertising spend that’s, on average, 30% to 40% more efficient than standard campaigns — how is this achieved?

“What really changed when iOS 14.5 was deployed was that instead of automatically being opted into an application, receiving your behavior actions, et cetera, when you download an app, you now on an iOS device have to actually opt in physically by clicking a button. And 80% of people are actually opting out of being tracked, which means that social platforms like Facebook and TikTok no longer know what actions you take after you click an ad. That means previously, when you would click an ad, and you end up purchasing, that would be fed into the targeting algorithm. And it would say, ‘Hey, people like Alex like buying this. Show this ad to more people like Alex.’ Now that’s not available. And what the platforms have to do is go much more broad.

And you’ll hear a lot of people say, ‘Okay, we’re doing broad audience targeting.’ And really, what that is akin to is like putting up a digital billboard in Times Square and then shuffling the ads very quickly and hoping that people at some point will see it at the right moment and then end up purchasing your product. What we do is we actually have a database of over 60 million personas where we have seen what they purchase, how much they spend in various categories. And we could then use that to run targeting off of our audiences. And it’s something we call lookalike audiences within the platform, such that people that look like those buyers that are spending the most money on these services and products, we could then target more people that look like them.”

So iOS 14 has reduced the specificity of targeting that marketers can do — is that correct?

“That’s right. It also hurt the attribution in that it’s less clear how much dollars you’re spending profitably because the signal of conversion and attribution has been blocked. It’s really those two components, attribution and then targeting.”

You’ve mentioned what’s changed on the marketing side. What’s actually the change that’s happening on the user side?

“I think the users definitely experience less efficient and relevant advertising. There is … this feeling like Instagram was listening to you, and it’s like, ‘Whoa, how does it know that I want these things?’ And really, that’s a testament of just the targeting algorithm being amazing and looking at all this behavior and seeing what we’re doing within the app. It quickly figures out what kind of advertising we would be most interested in. Now, I haven’t heard anyone say that for a while. And a big part of that is it’s just less efficient, right? People are getting served advertising that is just less relevant to them. 

And you also hear this movement towards engagement, meaning the social ad platforms today are much more focused on entertaining you, keeping you on the platform as long as possible. TikTok really led the charge there. You think about all the video content versus photo content. That’s really why you hear in the news how Instagram was trying to build reels to catch up to TikTok. It’s all about minutes and seconds and hours on the app and how long you can keep them engaged.

In order to keep them engaged, you need to keep putting out relevant creative content. And by mixing in advertising that is more relevant, everyone will have a better experience. And the truth is, right now, from an advertising perspective, it is less effective, less relevant for the average consumer like us.”

So the marketers have less ability to target the users that they think would buy their products, the users have less specific marketing advertised towards them, and it sounds to me like maybe the platforms would then have less ads clicked over time. I’m not seeing a positive here.

“It is the biggest problem of the last 12 months across technology in the US. And it is a meaningful problem in that, especially going into a downturn. Everyone’s budgets around acquiring customers [are] very important to achieve the growth expectations of all these startups out there and also more mature businesses. Every investor is looking for growth. It becomes much harder to realize those goals when your return on investment is just no longer where it used to be.

And that’s where we feel very fortunate that we had the foresight over five years ago to begin creating a network to allow us to get access to this underlying data. And the truth is, I didn’t realize how impactful and how challenging iOS 14.5 would’ve been if not for the fact that I was also a D2C founder of multiple eCommerce businesses. So I was literally feeling the pain myself, trying to solve it for my own businesses, my own brands that we’re struggling to acquire those customers. And then when we did a deep dive and started to bring in really smart data scientists and data engineers to then take on this problem, we started to realize, wow, we are actually in a position to solve this, not only for ourselves but for many of our clients out there.”

How do you see this trend evolving in the near term? And if anybody needs to be worried, who is it?

“I want to be really clear. We are very pro data privacy, and I actually believe that our solution being a third party in what previously was a walled garden in terms of the large platforms owning everything is actually very healthy for the long-term of this industry and this ecosystem. Ultimately more players like us taking a part in that digital and data landscape means that we’re decentralizing power. And when you split up the power into more people and more stakeholders, you then allow yourself to have more market dynamics that are competitive. And therefore, the end user, the consumer, will be better protected versus just a few very powerful, large technology companies owning all the data. 

There will continue to be more trends to protect customers and consumers from their data being used inappropriately. And I think what’s happening when you shake things up like this is that you force the larger companies to start thinking more long-term about how you balance privacy with their own business goals … And I think the movements that we’re seeing for data privacy will only get better for the consumer. Our role within that is to make sure that we secure our position as someone that keeps everything aggregated [and] anonymized and adheres to all these data privacy requirements and is always going to be standing by the consumer and ultimately helping the consumer have a better experience on these platforms without giving up their data.”

Can you give us an overview of the type of data you’re collecting and where you’re collecting that data from?

“So we launched a partnership network in 2016 called DojoMojo. DojoMojo is … a really great place for young digital businesses, brands, media companies to create partnerships where they will promote another brand in [exchange] for them promoting their products and services as well. We created this network where now there’s over 15,000 businesses on this platform, and they’re proactively working with each other messaging, connecting almost like a LinkedIn for marketers that want to work together.

Now, the number one question that these businesses ask is, well, which partners are going to be best for us and which ones won’t be as good for us? And the way that we answer this question is we said, ‘Hey, look, give us the information you have about your customers. We’ll keep it aggregated, anonymized, and encrypted in our database. We’ll never share this data to anyone else, but we’ll just score everybody and tell you … which brands will be good for you and which brands won’t be.'” 

The information they’re providing us is engagement stats around, let’s say, an email service provider, who’s opening this email, how engaged are they, how often are they clicking as an example. An SMS message, for example, ‘Okay, well, how often are they actually engaging when they receive a text?’

And then also we’ve also received transactional data like how much have they purchased and how often they are buying. We never share that information with anyone. But what we can do is, on an anonymized basis, score these individual personas and then share hashed emails, very specifically hashed emails, into these platforms. We can then leverage those anonymized and hashed data sets to then provide the improved targeting that you highlighted is consistently 30% better than what a brand is doing on these platforms on their own today.

So the information that is shared with Proxima has been shared in a way that the user is okay with, and there’s no personally identifiable information being shared at all. It sounds like what it’s meant to be in the first place.

“Every brand that we work [with must] adhere to [California Consumer Privacy Act] compliance. And that is for us in the US, the current standard for data privacy. And that allows … any consumer any time opt out of their data being accessed by that business as well as us. So ultimately, I think the balance here is making sure that it is consumer privacy first and then making sure that the business objectives are also met in a way that ultimately makes everyone benefit as a whole.”

Why is social media the channel you focus on?

“Social media has individual profiles for every user, and that is the channel that most well aligns with the fact that we have individual stats and metrics and trends on a user-by-user level. We’ve started to test on YouTube as well in that they have some similar audience-type functions, but something like Google AdWords, for example, doesn’t have the same profile based targeting available. And therefore, we can’t add as much value. So ultimately, that’s where there is the biggest opportunity to help improvement.

I also think right now, that’s where digital marketing is hurting the most. And that’s also why firsthand, as D2C operators and founders first, we knew that was where we had to start … And then as we continue to improve that and make that a lot better for our clients, we’d be happy to then transition to these other arenas.”

Is this also where the next generation of consumers is, coming up on social media rather than traditional channels?

“I think you’re spot on there … I think the answer is that in different media formats, we’ll continue to see the social platforms be where people want to spend the most time because it’s the easiest, most cost-efficient way to create a bunch of interesting content. And at the end of the day, that’s how we engage with media today. We want to escape. We want something to relax with. So we want something to be fun and kill some time and have some laughs.”

As far as changes across these social media platforms, either algorithmic or privacy related, how do those impact what you do at Proxima and how do you adapt to those changes?

“I fully expect that there will be more changes. Our business was created from a very disruptive moment, and it would be foolish for me to think that I’m always going to be on the good side of the disruption. And as a result, we’re always thinking about where the second bounce of the ball will be. And right now, we are benefiting from that first bounce, but we are looking ahead and trying to understand by speaking to experts on the data privacy side, by talking to the creators and the advertisers, and understanding where their needs are going to be met best.

I think there are many different ways we are preparing for what’s to come, but I couldn’t tell you exactly where that’s going to go. I just want to make sure that whichever path it goes, I’m ready with a plan to make sure that our business will continue to be able to serve our clients in the ways that they need us to serve them.”

What are some marketers’ reactions you saw to iOS 14.5?

“The initial reaction I saw was just a lot of frustration and desperation. I say this with complete empathy and compassion because I was literally at the helm of a D2C company when this went and was rolled out … It literally launched right into the iOS 14 challenges. And in the initial testing, all the unit economics looked great, and suddenly everything was wonky and we didn’t know what to do.

I would say what really changed, the number one action that is prevalent for all good marketers today, [is] adhering to the changes of the algorithm itself. Everyone is super proactive and creating a lot of creative assets very consistently and trying them on a very regular basis. Because what has also happened in the algorithm [is] that the social platforms are now becoming entertainment platforms. You’ll get just as much credit almost as an ad that you watch all the way through the full 10 seconds, 15 seconds, whatever it is, as if someone clicks on it. And that’s a big statement because before, the ads that would benefit in the auction were the ones that got clicked on because it would result in a sale.”

Now it’s like, ‘Oh, your creative [ad] is still going to get shown because it’s not causing people to bounce off our app.’ It was fun, it was interesting, which then now means all the creative studios are actually the real beneficiaries of iOS 14.5, because now they’re busier than they’ve ever been spinning up amazing creative, very innovative, engaging type creative assets. And you will start to hear more and more digital businesses. They’re talking about bringing studios in-house, video production. It is a part of their marketing team now.”

Are these more entertaining ads that people actually watch for the full 10 seconds? Are they leading to as many sales?

They are much more focused on testimonials and real people-based advertising. It’s less of the glamorous lifestyle stuff. It’s more like, ‘Hey, this is real. Here’s why.’ I think that’s been a trend I’ve seen in creative assets. I’ve also definitely seen people be more cutting-edge or more edgy, in that they’re trying to get you to remember them and then tell your friends … And you’ll see that videos are actually super fun as a content medium because you can do so many layers within it. 

Ultimately that’s better for us as consumers because we now have a more interesting form of entertainment and engagement. But for the marketer at home, they need to start thinking, if they’re not already, ‘How do I balance the original direct-response, very unit-economic-focused ROI that I was looking for with this more brand awareness engaging type creative content and still make all of that perform in terms of [my] business needs and objectives?’”

So we’re going to start seeing all sorts of commercials that feel more like progressive and Geico – basically just tertiarily related to the business that they’re offering, but much more entertaining?

“Yeah, you’re going to feel less like you’re being sold to and more like you’re being spoken to or engaged with in either a very entertaining way or a very genuine and honest way. And I think those are the most feasible paths to cut through all the other noise that’s out there.”

What is Innovation Department?

“For the first 10 years in my career, I was hardcore finance. And in that financial role, I spent a lot of time with CEOs and management companies, effectively telling them what they should be doing with their businesses. And I had this inflated sense of how I could then be a really good manager and CEO of my own businesses. Not true at all. It’s much harder to do than tell. And my way of balancing out my finance experience as an investor with the desire to be that entrepreneur that I had mentioned to you before was my childhood dream of one day building my own businesses.

My approach was to build a structure which was part investing and part operating. And we took a platform approach where we would invest in businesses alongside other investors, like a very nice partner in a VC deal. Or we would go find early-stage founders and say, ‘Hey, look, come be a part of our ecosystem. We’ll give you some seed capital. We’ll give you access to our resources, engineers, project managers, marketers, designers, et cetera. And then we’ll all build it together and learn from each other until this really rising tide lifts all ships away.’ And that became the Innovation Department. 

Now it’s really evolved into being more focused on direct-to-consumer and ecommerce businesses with a very strong lean towards retail and wholesale as well because, obviously, in today’s D2C category, you have to be omnichannel and very balanced. And we stopped taking flyers and bets on things that we didn’t understand. And then we started really focusing on things that we were good at and started to gain more economies of scale by doing things at repeatable to find that success that we’re looking for.”

How many companies have come out of Innovation Department so far, and over what time period?

“We’ve had five businesses that we literally started from scratch and are out there in the world and are doing great business and have amazing customers that support us. We launched Innovation Department at the end of 2015.”

Swinging back, what is a buyer persona, and why is it important for digital advertising?

“A buyer persona allows a business to understand who this customer is, what types of behaviors they’ll take, and what their likelihood is to continue to want to purchase your services or your product. If you are a digital business today, you need to invest in building out your understanding of your buyer persona, as well as very intelligently organizing your customer relationship management database to ensure that you are getting the most out of every single customer … It’s five times cheaper to get an existing customer to spend more than a new customer. So really understanding your buyer persona is super, super important.

The value we bring is that because we’ve been providing so much insight and analytics to all of those businesses on the partnership marketing platform so long, we started to get really smart about buyer personas on an individual level. And now in this moment where we see that there is a lack of targeting within these social platforms, our deep understanding and our proprietary database of all these insights around these buyer personas is now super effective to help bridge that lost efficiency from a targeting perspective.”

Do the brands that you work with already have personas set and ready to go, or do you help them with that?

“Most of the businesses that we work with are already spending money on these social platforms, so they have a really good baseline understanding. They’ve probably gone through multiple brand creative exercises and figured out who they think their tribe is. We always are very happy to learn and understand, and we don’t pretend to know someone’s customers better than they do. But what’s very interesting about our solution is we are very agnostic to where the insights take us, meaning our data science approach actually launches multiple campaigns using our data sets, almost like this game ‘Battleship’ growing up.

It’s a game of intelligent guessing. You’re trying to sink your opponent’s battleships by guessing on a grid. So what we really do in our calibration period, [which is] we call it when we onboard a new client, is we launch a bunch of shots out there from our database where the data sets have unique attributes, different qualities. And we look at where we get hits, where the unit economics, the ROAS, the CPA, the cost of acquisition is really performing the best. And we literally let machine learning figure out where to guess next.

So sometimes it’s very interesting. Where it takes us is very different from where perhaps we initially thought we were going to start and end up at. And I think that’s the beauty of now having so much data that we can leverage to then be able to provide those insights and ultimately provide our clients with the performance they’re looking for.”

Once you know everything there is to know about this buyer persona, how do you then take all of that and say, ‘Here is an impactful ad.’

“We leverage something called lookalikes, [which] means that if I show you Alex, you’re going to show my ad to other people that look like Alex, but not Alex himself, because you already know who Alex is. And that allows us to really expand our proprietary database of 60 million personas of US base into a much larger ecosystem of people. 

Additionally … our customers typically don’t have millions of customers in their database. The bigger ones will have maybe a million to two million. A lot of them are below a million still. So to access our data by showing us, ‘Okay, well, here’s who we think we’re looking for,’ and then augmenting those insights with the data that we have gives them an edge in this very challenging moment today.”

How often do you revisit the personas?

“Another interesting advantage for our clients that work with us is the fact that we have live integrations with a lot of our clients so that we are continuously developing and refreshing. There’s so many advertisers out there that probably have had a bad experience buying a list or trying to target some static thing that was sold to them. And maybe it provides some value for a little bit of time, but eventually, those audiences decay. Our unique approach allows us to actually refresh the data on the back end and keep those audiences alive and performing effectively forever by continuously providing more insights and more data into what the model shows.”

What’s next for Proxima?

We are very focused on being the essential acceleration platform for all digital businesses. That is literally that big, hairy, audacious goal that we want to strive for. And it’s become especially important to us that we were solving our own pain points in the very beginning as digital business founders. And now, it’s taken on this desire to continue to expand anywhere our clients need help growing.

To start, our opportunity has really been around social media platforms because that’s where people are hurting the most in terms of inefficiencies. But we want to go everywhere where our clients need support to grow, whether that’s Google advertising, whether that’s display networks, whether that’s the next big thing after TikTok, whatever that might be. Maybe VR. But really just making sure that we are solidifying our data, which is at the core of the foundation that allows us to go anywhere our clients need us. And then from there make sure that we build the software and the technology that allow us to be effective and provide the differentiated value that makes us essential to our clients.”

Who are the people who should be pausing this podcast right now and calling you up?

If you are spending money on Facebook, Instagram, and TikTok today, and you’re not completely ecstatic about your performance, we can help.

Continue Reading
Advertisement Submit

TechAnnouncer On Facebook

Pin It on Pinterest

Share This