*This article originally appeared on Alex Chalmer's LinkedIn profile*
From my experience at Publift, sustainable ad revenue growth doesn’t only come from plugging in more programmatic demand. It comes from tailoring auction decisions for each publisher.
In 2025, 61.3% of experiments delivered persistent revenue improvements, with an average uplift of 5.56% through deliberately designed experimentation.
What’s been most interesting isn’t just what worked, but why it worked. We see that outcomes improve when experimentation is personalised to each publisher’s setup, rather than scatter-shot testing driven by so-called “best practices”. Some examples to call out are:
- Dynamic ad refresh strategies, tailored to ad-unit behaviour, reduced low-quality ad requests, which have driven revenue growth for over 95% of publishers, while over 55% experience improved viewability.
- Timeout strategies tuned to auction conditions increased bid density without destabilising auctions, with over 90% of publishers seeing an incremental revenue impact.
- Responsive Prebid auctions, configured around site layout and structure, decreased ad-serve time and, in turn, demonstrated a sustained increase in viewability for over 40% of publishers.
The takeaway for publishers
Optimisation isn’t just about running more experiments. It’s about designing experiments that make sense for each publisher. We work with an array of publishers across different verticals, so it is important that we recognise their differences and identify opportunities to build durable revenue streams.
This is the kind of work I’m proud to be doing at Publift. This year, my team is focused on unpacking and understanding different behaviours in the Publift ecosystem and constructing more relevant, personalised optimisation strategies inside Fuse, so each publisher’s setup is tuned to how their site actually performs.
Credits to Amy Jansen-Flynn for the slide from our 2025 wrap that sparked this conversation.