Boris Mouzykantskii0:07
It's an honor to speak. I'll explain the title: we've been building custom solutions for media trading. Customization is a key theme. I'll unpack the existing boxes. Our clients include exchanges, early adopters, and now e-commerce players and brands requesting custom technology. The demand side ecosystem: DSPs create results for agencies, agencies serve advertisers, and advertisers now hire data scientists to compute their own ROI, pushing back on technology. This creates pressure for customization. I'll give a real case: a retargeting vendor shows 4% post-view conversion, but the advertiser's data shows 4% conversion without ads and 1% with ads. The data scientist explains the bias: users who leave the site are overrepresented. The solution is to avoid the attribution problem by using custom modules that identify users likely to convert without additional spend, focusing on incremental revenue. For example, an airline can exclude existing customers and oversold routes, improving ROI. This requires deep integration with advertiser systems, fast iteration, machine learning, and media spend efficiency. But there are internet-scale problems: access to the worldwide bitstream, filtering bots, and cross-device tracking. In summary, advertisers now compute their own ROI and push for customization. The only way forward is heavily customizing demand-side solutions. Thank you.