The Leadership Letter

Real correspondence from the people running real companies — and what it reveals about leadership.

When You Lock Your Own Team Out of the Market

If your internal product can only win by being kept on a leash, you haven't built a strong product — you've just built a wall.

Brad - Did you know that they're putting long-tail AFC onto AdX now? Weren't you involved in some discussions (strategic in nature) where the group decided to put long-tail AFC off of AdX for a good reason?

From an earlier message in the chain: "if GDN has access to the same AdX inventory as any other DSP out there, that weakens the value proposition of GDN as a buyside platform. If this is the case I would like to have GDN buy on all exchanges, not just AdX, since all our competitors are doing so. We want truly equal inventory access."

The response captured in the email: "GDN actually has exclusive access to long-tail AFC pubs in addition to the other stuff that is available thru AdX so don't worry. GDN is still differentiated on inventory access. No, you cannot buy on other exchanges; that will weaken the value proposition of AdX as a sellside platform."

Kim then reflects: "So at this point I (GDN as a buyer) am beginning to wonder why we only restrict GDN to AdX."

Context from Drew Bradstock: "The original thinking was that it was low-CPM, low-CTR long tail inventory but it turned out to behave closely to the other tiers. Matthew Young-Lai eventually felt it was fine to add and he is a tough judge of getting any weak inventory into AdX."

1. Core Message

This email chain captures a single, uncomfortable question asked by Woojin Kim: why is GDN (Google's buy-side ad network) restricted to buying only on AdX (Google's ad exchange), when that restriction is justified solely by the need to protect AdX's value as a sell-side platform?

The document reveals a deliberate policy — GDN cannot buy on competing exchanges — and a parallel one: long-tail publisher inventory is kept exclusive to GDN buyers on AdX, not available to rival buyers. Both policies are designed to make each Google platform look stronger by locking the other one in.

2. What the Executive Is Really Thinking

Kim is not raising a moral question. He is raising a business question: if GDN's differentiation depends on exclusive inventory access, and if that exclusivity is being eroded by adding more inventory tiers to AdX where competitors can also bid, then GDN's competitive advantage is weakening.

The response from within the chain is explicit: "GDN is still differentiated on inventory access. No, you cannot buy on other exchanges; that will weaken the value proposition of AdX as a sell-side platform."

The logic is a two-sided lock: GDN is propped up by exclusive inventory, and AdX is propped up by GDN's captive demand. Each platform's value is partly borrowed from the other. Kim's note — "I am beginning to wonder why we only restrict GDN to AdX" — is a signal that someone inside the company noticed the circularity.

3. Key Management Lessons

Exclusive Access Is a Competitive Moat — Until It Isn't

What it means

GDN's sell to advertisers was partly "you get inventory nobody else can buy." That edge depends entirely on keeping that inventory off competitors' platforms. The moment the exclusive inventory expands or leaks, the moat shrinks.

Why it matters

Moats built on access agreements or internal policy decisions are fragile. They can be changed by a single internal email, a regulatory order, or a partner defection. Moats built on technology, brand, or network effects are stickier.

MBA Perspective

This is a textbook Competitive Moats situation — specifically a moat that is maintained by restriction rather than by genuine superiority. The distinction matters. A restriction-based moat requires constant enforcement and creates internal tension, as this email shows.

Real-world application

If you run a marketplace or platform, ask yourself: is your product winning because it is better, or because you have blocked alternatives? If a competitor got access to the same inputs tomorrow, would you still win? Build toward the former.

Two-Sided Platform Strategy Creates Internal Conflicts

What it means

Google runs both the buy side (GDN) and the sell side (AdX) of the same market. The policies that make AdX strong (exclusive demand from GDN) are the same policies that constrain GDN (no buying on rival exchanges). These goals pull in opposite directions.

Why it matters

When you operate two sides of a market, decisions that benefit one side often cost the other. Without a clear priority framework, teams will make contradictory choices and ask questions like this one.

MBA Perspective

Platform Strategy teaches that two-sided platforms gain power by controlling both supply and demand. But the document shows a less-discussed cost: the internal buyer begins to question whether they are being disadvantaged for the benefit of the internal seller. That tension is structural, not personal.

Real-world application

If you are building a platform that serves both buyers and sellers, write down explicitly which side you will prioritize when their interests conflict. If you cannot answer that question, your team will answer it inconsistently — and someone will eventually send an email like this one.

Internal Exclusivity Requires Ongoing Justification

What it means

The original decision to keep long-tail AFC inventory off AdX had a rationale: Bradstock notes it was thought to be "low-CPM, low-CTR long tail inventory." When the data showed it "behaved closely to the other tiers," the justification for keeping it exclusive disappeared — so it was added to AdX.

Why it matters

Policies outlive their original logic all the time. The dangerous version is when a policy that once had a real business reason continues to be enforced purely by habit or internal politics.

MBA Perspective

This is a Resource-Based View issue. The value of a resource (exclusive inventory) depends on whether it is still rare, inimitable, and non-substitutable. When the data showed the inventory was ordinary, its strategic value as an exclusive asset declined. The team correctly updated.

Real-world application

Audit your internal policies annually. Ask: "What was the original reason for this rule?" If the condition that created the rule no longer exists, the rule deserves a fresh look.

Competitive Differentiation Cannot Be Purely Defensive

What it means

Kim's question — why is GDN restricted to AdX only — exposes a risk: GDN's value proposition is partly constructed by preventing GDN from accessing better alternatives, not by GDN being genuinely superior.

Why it matters

A product that looks good only because its users are blocked from leaving is not a strong product. It is a captive. Captives eventually escape or attract regulatory attention.

MBA Perspective

Switching Costs is the relevant lens here, but in reverse. Normally, you want customers to face high switching costs when leaving your product. Here, the cost is being imposed on an internal team. That is a signal that the product may not be able to compete on its own merits in an open market.

Real-world application

Before you add a contractual lock-in or a policy restriction, ask: if we removed this restriction tomorrow, would our product still win? If the honest answer is no, the restriction is masking a product problem you should fix instead.

4. Strategic Analysis (MBA Style)

Competitive Strategy

The strategy visible in this document is vertical integration used to reinforce platform dominance. GDN (demand) and AdX (supply) are linked by policy so that each strengthens the other's market position. Competitors who only operate on one side — a pure DSP or a pure exchange — face a bundled opponent. The competitive logic is: if rivals cannot access GDN's demand, AdX is more attractive to publishers; if GDN buyers cannot go to rival exchanges, AdX gets all of GDN's spend.

Risk Analysis

Kim's question names the risk clearly. If the differentiation of GDN rests on exclusive inventory, and that exclusivity can be diluted by policy changes inside the same company, then the product's competitive position is fragile. The secondary risk: an internal buyer who cannot access competing exchanges may underperform versus external buyers who can shop across platforms — which would eventually show up in campaign results and client retention.

Build vs Buy Analysis

This document does not involve an acquisition decision. The "build vs buy" question is not relevant here, and applying it would be forced. The relevant structural question is whether internal integration creates enough combined value to offset the constraints it places on each component platform. The document does not resolve that question.

Market Dynamics

This email is a window into a concentrated ad tech market where one company controls significant infrastructure on both the buy side and the sell side simultaneously. The policies described — exclusive inventory tiers, restrictions on where an internal buyer can spend — are only possible when a single firm has substantial reach on both sides of the market. A firm with a smaller position could not enforce these restrictions because buyers or sellers would simply route around them.

Long-Term Strategic Implications

If the strategy succeeds, Google maintains a differentiated position for both GDN and AdX that independent competitors cannot easily replicate because they do not control both sides. If the strategy fails — through regulatory intervention, defection of large publishers to rival exchanges, or advertiser demand for open buying — the whole structure unwinds simultaneously: GDN loses its exclusive inventory advantage, and AdX loses its captive demand from GDN.

5. Hidden Insights

The internal buyer is the first one to notice the problem. Kim is not an outside analyst or a regulator. He is a GDN insider who sees the policy and asks whether it makes sense. When your own internal customers start questioning why they are restricted, that is an early signal that the restriction may not hold under scrutiny.

Differentiation by exclusion is not the same as differentiation by value. The document draws a direct line between GDN's differentiated "value proposition" and the fact that competitors cannot access the same inventory. The differentiation is real, but its source is policy, not product quality. That is a meaningful distinction for anyone building a platform.

Inventory quality gatekeeping has its own strategic value. Bradstock notes that Matthew Young-Lai was "a tough judge of getting any weak inventory into AdX." This detail is easy to overlook, but it points to a separate, more durable competitive advantage: a reputation for inventory quality that functions as a signal to buyers. This kind of advantage is harder to replicate than a policy restriction.

The question Kim did not ask is as important as the one he did. Kim asks why GDN is restricted to AdX. He does not ask whether GDN is a better product because of that restriction or despite it. That is the deeper question the document leaves open.

Court Exhibit
United States v. Google LLC (Ad Tech)
1:23-cv-00108 (VAED), Trial Ex. PTX0103 — DOJ public archive
July 26, 2012
Public domain
View the primary source →