The quiet auction model shift that changed everything
For years, performance marketers operated under the assumption that you shouldn’t have multiple retargeting providers targeting the same audience. The logic was sound. In the second-price auction world, having two partners bidding for the same user meant bidding against yourself, and driving up costs without adding value.
This wisdom became so ingrained that many marketers still follow it today. But the underlying mechanics have changed.
Starting in 2019, the digital advertising ecosystem transitioned to first-price auctions. Google led the way, and the rest of the industry followed. In a first-price auction, the highest bidder wins and pays exactly what they bid. No more second-price calculations. No more worrying about inadvertently inflating your own costs.
As Vice President of Global Product Commercialization, Jaysen Gillespie noted in AdExchanger, "In the first-price world, that fear disappears. Having other bidders in the auction doesn't change what the winner pays." This shift means that having multiple retargeting partners working on the same pool of data no longer creates the inefficiencies it once did.
Yet many busy marketers may not have noticed the change. They're leaving money on the table by not updating old practices. The data show a different reality. When partners use distinct approaches, they complement rather than clash.
Why bid collision is no longer a concern
Understanding why bid collision has faded, we need to take a look at auction mechanics. In the old second-price model, the winning bidder paid slightly more than the second-highest bid. If you had two partners bidding $5 and $4 for the same impression, you'd win at $4.01. But if both were working for you, you essentially bid against yourself.
First-price auctions work differently. The highest bid wins and pays that exact amount. Whether there's one bidder at $5 or ten bidders below that price, the winner still pays $5. Your other partners in the auction don't affect what you pay.
This change has implications for performance marketing strategy. Instead of limiting yourself to one partner to avoid bid collision, you can now use multiple retargeting partners to reach less obvious segments of your audience and capture disparate audiences a single platform might otherwise miss.
The main takeaway is that partners using different technologies and approaches aren't really competing for the same impressions. They're identifying different opportunities based on different signals. When one partner wins an auction, it's because their approach identified value that others missed.
What makes partners fundamentally different
No two retargeting vendors are the same. Having two vendors that identify and make bidding decisions based on separate intent signals means you come out ahead.
Conventional Machine Learning and advanced Deep Learning process information, identify patterns, and make decisions in distinct ways.
Machine Learning relies on structured data and predefined features. It excels at recognizing patterns in organized information but requires human input to define what matters. It's powerful but bound by parameters set by data scientists and other stakeholders.
Deep Learning, however, operates on another level. It processes large volumes of unstructured data, identifies patterns humans might never notice, and continuously adapts without manual intervention. It can work with raw, chaotic information and find meaning where traditional approaches only see noise.
Machine Learning might excel at identifying users who've shown clear purchase intent through specific behaviors. Deep Learning identifies users whose complex, non-obvious patterns suggest they're likely to convert despite never having viewed the product before. Working with multiple vendors that use complementary technologies means you cast a wider net in a pool of prospective customers.
Our internal data show that 61% of purchases driven by Deep Learning come from products a user had not viewed before. Rather than stealing conversions from Machine Learning campaigns, Deep Learning algorithms find entirely new opportunities.
Building your maximized performance stack
The most successful performance marketing stacks today apply this principle of complementary technologies. Instead of putting all their eggs in one basket, brands combine partners that excel at different aspects of the conversion journey.
A typical maximized stack includes a mix of traditional Machine Learning retargeting, walled garden retargeting on social platforms like Meta, and Deep Learning retargeting for the open web.
Each partner in this stack operates from a specific starting point. They're not fighting over the same impressions, but rather finding opportunities.
Brands who add RTB House to their retargeting stack see an average 57% increase in retargeting scale at the same ROAS.
Making the transition to multiple retargeting partners
Moving from a single-partner to a multi-partner approach takes careful planning but isn't as complex as many assume. Start by auditing your current performance stack. What technology does your primary partner use? What signals do they prioritize? What segments of your audience do they excel at reaching? Understanding your starting point helps identify gaps that complementary partners could fill.
Next, identify partners that truly differ in their approach. Ask potential partners about their core technology, not just their features. A partner touting "AI-powered optimization" might be using conventional Machine Learning, while another might employ true Deep Learning. Be clear about how your retargeting partners differ.
Try running an incrementality test. You don't need to overhaul your entire approach overnight. Start by allocating a portion of budget to a complementary advertising partner. Set clear KPIs focused on incremental lift rather than just overall performance and, of course, monitor for true additive value.
Measure your target metrics holistically. When running multiple retargeting partners, individual campaign metrics matter less than portfolio performance. One partner might show lower ROAS in isolation but drive significant incremental revenue when combined with others. Focus on total revenue and efficiency across all partners.
Address operational concerns upfront. Working with multiple retargeting partners does require more coordination, but modern platforms and attribution solutions make this manageable. Set clear rules about audience segments, budget allocation, and performance targets to avoid confusion.
Final thoughts
The performance marketing landscape has evolved, but many strategies haven't kept pace. Bid collision, once a legitimate concern, is now an outdated constraint in the first-price auction world. Marketers who recognize this shift and diversify their retargeting streams see significant incremental revenue growth.
The future of performance marketing is about building a stack of partners that complement each other through different approaches.
Ready to explore how multiple performance partners can drive incremental revenue for your business? Contact us to learn more about building a maximized performance stack that amplifies results from partners that are truly different in kind.
