The programmatic ecosystem is in the midst of an evolution driven by significant advances in AI technology coupled with a shift to sell-side decisioning. This shift enables smarter, more dynamic curation, where both sell- and demand-side decisioning can seamlessly interconnect to unlock true performance gains.
To explore how sell-side decisioning is transforming programmatic strategies, I spoke with Ryan Lammela, group director of channel activation at Butler/Till, and Andrew Altersohn, co-founder and president of SWYM.ai.
Andrew Altersohn, co-founder and president of SWYM.ai, and Ryan Lammela, group director of channel activation at Butler/Till
1. As an outcome-driven agency, Butler/Till has focused on improving the effectiveness of programmatic campaigns, from cutting non-performing media to removing made-for-advertising (MFA) sites to optimise supply paths. How are you approaching this today?
Ryan Lammela: We have a multi-faceted approach that includes applying a “directness matters” ethos to supply path optimisation (SPO). We leverage highly trusted SSPs such as Index Exchange and highly trusted curators such as SWYM to ensure both efficiency and quality. We also leverage third parties such as Jounce to help with agency-wide MFA mitigation, careful DSP selection, and the trading efforts of our hands-on-keys staff.
This continuously evolves alongside the shifts in our ecosystem. Today, that means AI is increasingly an important factor, for example in how we approach curation to optimise media buying.
2. Let’s talk more about that—we’ve seen many examples recently that show just how transformative algorithms can be in driving better outcomes more efficiently, particularly when applied on the sell side. How do you view AI? How is AI shaping your curation efforts?
RL: We see AI tools as a better way to handle more data, faster by having a learning-based model that updates as it processes more signals. AI is the next big leap available in media and advertising.
Just as programmatic advertising allows marketers to reach more inventory in less time by removing insertion orders, AI provides a counterbalance to that efficiency, ensuring we’re buying only in the right places, at scale.
Andrew Altersohn: Today, most agencies and brands have a defined SPO strategy that often centers on a subset of preferred SSP partners. That’s a great start. But that approach is only a starting point to finding the most efficient, cost-effective, and transparent path to inventory.
Traditional SPO doesn’t address or correct for the most important part—the missing performance signal between a DSP and the SSPs with which it’s connected.
SSPs typically don’t know which bid requests are driving better performance, which is crucial for optimised curation. So, even with an SPO strategy, AI-driven curation and automation are critically important in bridging the gap to ensure that the bid requests being sent to your DSP have a higher probability of driving performance.
3. To help solve for that, SWYM developed an algorithmic solution for curation that learns from both buy- and sell-side signals, and created an Index Marketplace. What compelled you to anchor decisioning on the sell side?
AA: We saw that the underlying opportunity to reduce programmatic waste and inefficiency was not in how we buy programmatic media, but in how it’s supplied. Specifically, marketers can only bid on the QPS-limited supply of ad requests made available for bid through their DSP. And, the SSPs that transmit supply to DSPs are missing a critical piece of information—they don’t know which bid requests actually deliver on campaign KPIs.
Without such a performance feedback signal between the DSP and SSP, the supply of bid requests is often under-optimised.
That’s why we developed our SmartCurate solution. Our approach elevates the importance and role of supply partners and sell-side decisioning, and it’s changing the way brands and agencies think about their own buying and optimisation processes.
Our solution doesn’t only identify and curate top-performing inventory. It also increases the available supply of better inventory from the start by signaling to the SSP that the DSP is likely to bid on it. This is a key signal that accelerates the flywheel of better supply.
4. Ryan, how has this shift to sell-side decisioning influenced your campaign strategy at Butler/Till? How has it impacted campaign operations or your teams in navigating a new paradigm?
RL: By ensuring we’re only sending the right requests from the sell side, we can effectively reduce the “clutter” that our DSP has to parse through. This ultimately helps our clients see more bid requests that they really want and that will drive the outcomes they’re looking for. Meanwhile, it helps our sell-side partners better understand what types of inventory we don’t want so it can be monetised elsewhere.
We love win-win-win solutions and sell-side decisioning helps all parties involved.
5. What kind of results has Butler/Till been able to deliver for clients since unlocking decisioning further up the bid stream?
RL: We’ve implemented this algorithmic curation approach on a range of campaigns and clients together with SWYM. As just one example, a campaign we ran for one of our financial services clients resulted in a clear simplification and cleanup of the programmatic supply chain, reducing the number of SSPs and suppliers by 90% and domains by 52% (you can read the full case study here).
Not only does this reduce supply chain complexity and improve transparency, but such simplification and consolidation also improves brand safety by limiting the number and variety of sites the campaign runs across to only the most effective properties.
More importantly, this approach to algorithmic sell-side decisioning drove a clear impact on conversions, with this primary KPI up 56% while the cost per conversion decreased by 26%.
6. Already that’s a notable impact, and we’re still in the early days of capturing the potential of AI and sell-side decisioning in programmatic. What opportunities do you see ahead?
RL: We feel the best advertising happens when ads are not just delivered to the right person, but at the right time and in the right context as well. As the sell side develops more insights into what brands want, it’s exciting to think about the role that AI could have not just in curation but in content creation.
For instance—consider an application almost like dynamic creative optimisation (DCO), except that it occurs on publishers’ properties. Content could be updated in real time in a way that helps the consumer understand and recall the ad, all based on buyer propensity.
AA: We’re still in the early days of algorithmic supply shaping, and the potential to evolve it beyond basic optimisations is massive. The real unlock lies in integrating richer datasets—such as audience data—directly into supply curation, allowing for scaled, multi-dimensional bidding strategies that bridge the gap between supply and demand in real time.
By leveraging both supply-side signals (win rates, availability, and supply path efficiency) and demand-side performance data, we can move toward truly intelligent bid shaping—where every impression opportunity is evaluated dynamically, ensuring buyers are always bidding at the most efficient price.
But the evolution doesn’t stop there. The next phase will be driven by improved bidding efficiency models within DSPs, enabling more precise bid shading, adaptive pacing, and real-time optimisation against dynamic supply conditions. As AI continues to shape the landscape, we’re moving toward a world where both sell- and demand-side decisioning become seamlessly interconnected, driving greater efficiency, performance, and scale.
Explore Index Marketplaces and see how sell-side decisioning can boost performance and unlock new opportunities for success.
