1. Core Message

Jonathan Bellack is asking his team a simple question with big consequences: how do we push more of DBM's buying through Google's own ad exchange (AdX) instead of third-party exchanges (3PE)? He notes that moving $1 from a third-party exchange to AdX captures a 20% revenue share Google otherwise doesn't get. He then lists open questions about how to actually do that — through buying strategy, features, or commercial terms — and admits the team is only just starting to work on it.

2. What the Executive Is Really Thinking

Bellack is thinking like an owner of infrastructure, not a participant in it. DBM (Google's buy-side tool) sends money to many exchanges. When it sends to a competitor's exchange, Google earns only the DBM fee. When it sends to AdX (Google's own exchange), Google earns the exchange fee too. The whole memo is about closing that gap. The honesty is notable: he doesn't claim Google already does this well. He asks whether there's a "defensible" way to change DBM's behavior, and whether publishers actually benefit from calling multiple 3PEs in the first place. He's also thinking ahead — mobile apps ("mApp") are an emerging surface where Google doesn't have to copy the messy web setup. Build it differently from day one.

3. Key Management Lessons

Follow the take rate, not just the revenue

What it means

Bellack frames the opportunity in revenue-share terms: a dollar moved to AdX earns Google an extra 20%. He's not chasing more ad spend — he's chasing more of the spend Google already touches.

Why it matters

Growth often hides inside existing flows. Routing decisions, defaults, and integrations can be worth more than new customer acquisition.

MBA Perspective

This is Vertical Integration thinking. Google owns the buyer (DBM) and an exchange (AdX). Each step of the stack you own adds a take rate. Owning two steps and routing between them is the economic prize.

Real-world application

A payments company should ask: when our checkout sends a transaction to an outside processor, what would it cost to route it internally? A SaaS company should ask: where in our workflow are we handing customers to a third-party tool we could own?

Ask whether the customer actually benefits from the status quo

What it means

Bellack asks: "How much lift do publishers actually get today by calling DBM multiple times through different 3PE, relative to running all of DBM through AdX?" He's questioning whether the existing multi-exchange behavior is real value or just habit.

Why it matters

If the current customer practice doesn't actually help them, you can change it without harming them — and capture the gains. If it does help them, you need a different plan.

MBA Perspective

This is a Resource-Based View question dressed as a customer question: is the third-party exchange layer providing distinct value, or is it a substitutable middleman?

Real-world application

Before changing defaults or integrations, measure honestly: do users get better outcomes through the outside option, or are they just following inertia? The answer determines whether your move is pro-customer or anti-customer.

Change incentives, don't just change code

What it means

Bellack lists "features or set commercial terms that give pubs & buyers more incentive to run DBM through AdX." He's open to both product and pricing levers.

Why it matters

You can shift behavior with a product feature, a fee structure, or both. Pricing is often faster and cheaper than building.

MBA Perspective

Switching Costs and Platform Strategy. If you can make your own pipe materially better or cheaper for both sides of the market, you create gravity that pulls volume in without forcing it.

Real-world application

If you run a marketplace, look at every place users could leave for a competitor. Ask: what feature, rebate, or exclusive would make staying obviously the better choice?

Don't recreate a messy market in your next market

What it means

For mobile apps ("mApp"), Bellack asks whether Google should "change the DBM buying strategy entirely for emerging businesses... instead of recreating the Web's 3PE marketplace."

Why it matters

New surfaces are a chance to fix structural problems from the start. Once habits and integrations form, they're very hard to undo.

MBA Perspective

First-Mover Advantage applied to platform architecture. The shape you set in a young market often becomes the shape of the mature market.

Real-world application

When entering a new geography, channel, or product line, don't port your old plumbing. Ask what you'd build if you were starting today.

Admit how early you are

What it means

Bellack ends with: "We're just starting to pivot toward these questions on the sell side, there's a lot more to do." He's not pretending the strategy is finished.

Why it matters

Clear-eyed status reports beat polished ones. Teams can only fix what leaders are willing to name as unfinished.

MBA Perspective

Not a framework — just operating discipline. Pretending a strategy is mature when it isn't kills internal urgency.

Real-world application

In your next strategy memo, separate "what we believe" from "what we've tested." Force yourself to list the open questions, not just the conclusions.

4. Strategic Analysis (MBA Style)

Competitive Strategy

The logic is to maximize Google's share of the wallet on flows it already participates in. Every dollar DBM spends elsewhere funds a competing exchange. Routing more of that spend internally raises Google's revenue per dollar and starves competing exchanges of volume.

Risk Analysis

Bellack flags the risk himself by asking for a "defensible" way to shift DBM behavior. The implicit risks: publisher and buyer backlash if yields drop, and external scrutiny if the routing looks like self-dealing. He's looking for a path that survives both customer and outside review.

Build vs Buy Analysis

This document is not about M&A. It's about utilization of assets Google already owns (DBM and AdX). The choice is between building features, setting commercial terms, or both — to lift internal routing share.

Market Dynamics

The document reveals an ad tech stack where the same dollar passes through multiple intermediaries, each taking a cut. Publishers call several exchanges hoping for higher prices. Google sits on both the buy side (DBM) and sell side (AdX) and is thinking about how to consolidate flow through its own exchange.

Long-Term Strategic Implications

If the strategy works, Google captures more take rate on existing volume and weakens rival exchanges' scale. If it fails — either because publishers see worse yields or because outside scrutiny intervenes — Google loses trust with the two-sided market it depends on. Bellack's word "defensible" suggests he already sees that second risk.

5. Hidden Insights

• Awareness of dual-role tension. By asking for a "defensible" approach, Bellack implicitly recognizes that owning both buyer and exchange creates conflict-of-interest risk. He wants the economic gain without the optics problem.
• Mobile is treated as a clean slate. The mApp aside is a quiet but important signal: leadership sees emerging surfaces as a chance to design a different market structure, not just port the web's.
• Sell-side underinvestment. "We're just starting to pivot toward these questions on the sell side" suggests the sell-side team has not historically optimized for routing share — meaning meaningful upside may be sitting untouched.
• Customer value is an open question, not an assumption. The fact that Bellack has to ask how much lift publishers actually get from multiple exchanges suggests Google doesn't yet know — and the answer will shape what's possible.