Voyantis

Upside shifted from two tCPA campaigns chasing conversion proxies to one tROAS campaign optimizing for actual customer value, bringing in 15% more of their highest-value users as a result. Here's a look at how we tackled it together: 👉 Designed a model specifically for auction impact, amplifying the spread between high-value and low-value users. 👉 Engineered those predictions into user-level signals that influence Google’s algorithm, delivering each within an hour of sign-up in the format, timing, and frequency the algorithm could digest. 👉 Built a system that keeps the signals sharp as Upside's business and the auction evolve. The shift consolidated two campaigns into one that delivers a single, accurate signal of customer value. Powered by the right signals, Google's algorithm knows exactly what to do - and that learning compounds over time. These are the most rewarding partnerships to be part of. We know what value-based bidding looks like when it's done right, and love seeing clients get there. Proud to partner with the Upside team on this. Read more about their journey: https://lnkd.in/dnfadJz7

Improving your ad signals in a constant loop and making it a critical system. That was the centerpiece of our webinar last week about signal engineering, which Barbara Galiza and I did together last Thursday. Over 300 people signed up, 70 attended the live stream, and over 150 have already watched the recording. This shows me that the topic is on more agendas than I thought. Before working on these kinds of projects, signal engineering felt more like a niche problem. But with shifting ad campaign types and more closed ad platform optimization systems, this remains one of the few ways you can steer your ad targeting. Thanks again to Voyantis for sponsoring this session and making it available to so many people. If you like, Voyantis is offering a free Signal Engineering Gap Analysis (link in the comments). If you like to watch the recording, we also made it available on YouTube (link also in the comments). If you have any questions about Signal Engineering setups and how to approach them, feel free to drop them in the comments here, or reach out to Barbara and me via direct message.

One week after MAU and we're still buzzing from the time spent at some of the best spots in Vegas, with some of the brightest minds in the industry ⚡ We kicked off Monday with an exclusive dinner and tour inside Allegiant Stadium, one of the most iconic venues in the world. That was followed by happy hour at Morimoto, and a Roaring 20s cabaret night at the Mayfair Supper Club with front-row views of the Bellagio Fountains. We ended the week with our own Alexandra Ivacheff taking the stage for a session on why acquisition and retention have to work together — and why the teams that get that right are the ones that scale efficiently. Big thank you to our partners at RZR, Singular, Remerge, and M&C Saatchi World Services for making it all happen. 🙌 #MAUVegas #GrowthMarketing #Voyantis

In real estate, for every serious homebuyer there are dozens window shopping. At the moment Google bids on them, they look exactly the same. Lennar builds over 80,000 homes a year. They know what serious intent looks like – someone who books an appointment with a New Home Consultant – but that appointment gets booked weeks into the journey. Google is bidding on day one. The Lennar team partnered with us to bring that insight upstream – predicting each lead's likelihood to book an appointment while Google was still deciding who to chase. The impact showed up in two ways: 👉 A better user mix, with lead-to-appointment rate climbing 40% 👉 More efficiency with cost-per-appointment dropping 15% Higher-intent buyers and a lower cost to acquire them. The case study breaks down exactly how we tackled it together: https://lnkd.in/d86cZ2Sx

Building a predictive model is one thing. Operating signal engineering as an ongoing discipline within your growth org is another. This Thursday, growth experts Barbara Galiza and Timo Dechau are walking teams through what it actually takes in a live virtual workshop. Three things you'll walk away with: 👉A read on whether your team is ready for signal engineering and what to put in place first 👉The design decisions that make it work and the failure modes that break it 👉The operating model that compounds as your platforms, business, and customers evolve Signal engineering can feel abstract in theory. This session is built to ground it in practice, so attendees can see how it applies to their own growth context. If this resonates, come learn alongside other growth teams working through the same shift: https://luma.com/bkvlcsgb

In ride-hailing, a user's first trip tells you almost nothing about whether they'll become a regular rider. It's the second trip, the third, the weekly habit that builds a business. But Google can't wait for the habit to form. It optimizes on the signals it receives in the first hours, and by the time the real pattern emerges, it's already chasing the wrong riders. That's the challenge inDrive brought to us. With 400 million downloads across 48 countries, they had built a strong acquisition engine. What they wanted was to make it smarter — acquiring not just riders, but the ones who'd come back week after week, market after market. We partnered to give Google a predictive signal of long-term rider intent within hours of acquisition, not weeks later when the habit had either formed or faded. The approach pays off in two ways. First in acquisition, with more frequent riders at the right price. Then in compounding, with every ride, every referral, every market expanding the return. We broke down how we built this into inDrive's growth engine. Read the full case study: https://lnkd.in/eA-t7JKq

So you want to get ad platforms optimizing your campaigns for customer value, not just conversions… Most growth teams have built and tested predictive models, moving steadily toward optimizing campaigns for customer value, not just short-term conversions. But activating a model is much harder than building it. That work, signal engineering, has quickly become the backbone of efficient growth. Next Thursday, Barbara Galiza and Timo Dechau are leading a live workshop and Q&A on what it takes to implement signal engineering within your growth org, including the design decisions that make it work and how to operate it as an ongoing discipline that compounds ad performance. You’ll benefit from joining if: 👉 You're considering or actively moving toward optimizing for customer value and want to understand what it takes to do it well 👉 You're already running based campaigns using static or proxy values and want to push further 👉 You've built a predictive model and watched it underperform in the auction 📅 May 21, 12pm ET  🎙 Live workshop with audience Q&A Seats are capped at 300. Register now: https://luma.com/bkvlcsgb

Monday night at MAU, you're invited to a seat at the Driver's Table. 65,000 seats inside Allegiant Stadium, and for one night, a small group of us. We're teaming up with RZR and Singular for a curated evening for app growth leaders navigating a cross-screen world. Save your seat here: https://lnkd.in/exirixSG

MAU Vegas is one of the few weeks the mobile growth community is all in one place. We've curated three nights to make the most of it – small rooms for growth leaders to connect with peers and swap notes: → A private night inside Allegiant Stadium, with RZR and Singular → Light drinks and bites at Morimoto with M+C Saatchi Performance → Cocktails over the Bellagio Fountains, then a cabaret dinner at The Mayfair Supper Club, with Remerge and Singular Get the full rundown of each and save your spot here: https://lnkd.in/exirixSG Can't make the events? Catch the Voyantis team in between sessions: Ido Wiesenberg Peleg Israeli Alexandra Ivacheff Eithan Attia Victoria Amado Brian Michael Murch Hannah Foley

Klar, the leading digital financial services platform in Mexico, was optimizing Google campaigns for credit card approvals. But their best customers were the ones who actually used the card, behavior that accumulated over months. So they set out to change what Google was optimizing for. Here's how we tackled it together: 👉 Predicted each user's long-term value from behavioral signals, not approval alone 👉 Engineered signals tailored to how Google actually learns, telling the platform what new users would be worth the moment they first engaged 👉 Built a system that keeps the signals sharp as Klar's customer base evolves, so the algorithm stays aligned with the users actually driving revenue The difference shows up ROAS and CAC but more importantly, in the users. Every new Klar customer now generates 15% more revenue on average. Read the full case study to learn how they did it: https://lnkd.in/ebqcMPJX