Mile AI

Revenue loss in programmatic doesn’t always look like a drop in fill rate. Sometimes it looks like auctions clearing too low. When publishers run prebid auctions without effective floor pricing, there’s nothing stopping buyers from winning below their true willingness to pay. This makes it harder to: ➡️ Understand how much revenue is being left on the table ➡️ Scale yield without relying on more traffic Across publishers, introducing dynamic flooring consistently drives incremental lift Dynamic Flooring: ➡️ Prevents systematic undervaluation in open auctions ➡️ Improves pricing across everyday impressions, not just premium ones This is the problem we focus on at Mile, helping publishers protect and grow revenue by pricing each auction more accurately. If you’re thinking about flooring for the first time, happy to talk through what we’re seeing.

The biggest risk for publishers right now isn’t AI. It’s underestimating how deeply it will reshape revenue and distribution. Ezra Menaged has taken a different approach at Hometalk, Inc.. AI isn’t a side experiment. It’s built into how the business runs across product, content, distribution, sales, and outreach, even inside their community with their in-house assistant. The shift isn’t about replacing people. It’s about making strong teams more effective at every layer. The publishers who move early on this won’t just adapt. They’ll redefine how growth works. Episode link in the comments 👇

That “flying blind” moment is the whole reason Mile exists. Vijay Kumar has spent more than a decade inside programmatic’s complexity, building Mile to give publishers something the market rarely offers: control and transparency. Today, Mile has evolved into an AI-powered unified control layer that combines real time auction data, autonomous optimization agents for experiments, machine learning models for flooring and shaping, and an LLM that surfaces revenue insights, all within a self-serve UI. Read more about Mile’s journey in the comments 👇

March 2020. CPMs dropped overnight. Hometalk had around 55 to 60 employees and was staring down a very ugly year. Ezra Menaged, CEO of Hometalk, Inc., made the hardest call of his career. Let go of a third of the company and asked himself one question. What are we actually really good at? The answer was email. Millions of subscribers on Hometalk and a genuine core competency that had been sitting there the whole time. So instead of chasing the next big thing, he went all in on what they already knew how to do. Within the year, they launched Food Talk and Up Style. Here's where it gets fun. They started 2020 on track for significant losses. They ended it in profit. The year after that, the revenue growth continued. That turnaround is what eventually led to the Vertical Scope acquisition in November 2021. Sometimes the pivot that saves you is the one that takes you back to basics. Episode link in the comments.

📍 We’re at Prebid Ascent London today! Our CEO, Vijay Kumar and SVP of Revenue, Sarah Ruxin are on the ground at Prebid Ascent, connecting with publishers and partners from across the ecosystem. From evolving auction mechanics to the role of open-source in shaping the future of ad tech, there’s a lot to unpack. If you’re here and want to exchange ideas or chat about what’s ahead, let’s meet up!

Most publishers assume the big ad tech providers are doing it better than they could. Ezra Menaged, CEO of Hometalk, Inc., decided to put that assumption to the test. Two and a half years ago, Hometalk ran a two-month head-to-head test, pitting its in-house setup against an established ad tech provider. The results were clear. They were performing better, so they made the move. But winning the test was the easy part. What followed was building a serious ad tech team from the ground up. One that created real programmatic impact through patience, iteration, and the willingness to invest before the returns showed up. In-housing isn't a switch you flip. It's a decision you commit to, and then rebuild around. For publishers with the team and the appetite to own their operation, the upside is real. The iteration is yours. The growth is yours. Episode link in the comments.

Programmatic has become too dynamic for static decisioning. At Prebid Ascent, we’ll be digging into what comes next: sell-side systems that learn, adapt, and optimize in real time. Sam Edmonds from Venatus joins Vijay Kumar to unpack how ML is reshaping yield, from smarter bid selection to dynamic floor strategies across formats. Less manual tuning. More outcome-driven monetization. If you're thinking about how your stack needs to evolve, this conversation is worth your time.

A platform algorithm change can erase years of traffic overnight. Hometalk used that moment to rethink how distribution should work. Ezra Menaged, CEO of Hometalk, Inc., faced that challenge head-on. Back in the day, instead of relying solely on Facebook, Hometalk doubled down on newsletters, email, and owned audience channels. That focus didn’t just protect them when clicks and algorithms shifted. It gave them the freedom to experiment, iterate, and grow on their own terms. Ezra’s approach combines product thinking, AI-powered tools, and relentless experimentation to build a publishing business that lasts. The lesson is simple: own your audience, control your growth, and never stop asking what’s next. Episode link in the comments.

Many publishers rent their audience from platforms. Hometalk, Inc. focused on owning the relationship. Ezra Menaged breaks down the two things that kept Hometalk standing through every platform shift, every pivot, and every moment the ground moved beneath them. The ones still standing never stopped asking what's next. Episode link in the comments. 

Programmatic yield management is changing, and static rules are no longer enough. We're speaking at Prebid Ascent London, and this is exactly what we'll be getting into. CEO Vijay Kumar is joining Sam Edmonds, Head of Yield & Programmatic at Venatus Media, for a session on building the data and decisioning layer for publisher monetization. The conversation will go beyond theory - covering how ML can enhance bid selection, floors, and demand strategy across display and video, and what it actually takes to move from static rules to predictive, outcome-driven monetization. If you're evaluating your stack's readiness for ML-powered decisioning, this session is for you. Prebid Ascent | March 25 | 2:50 – 3:20 PM