1. Core Message

Jonathan Bellack does not want Google to build a product around Header Bidding (HB). He wants Google to invest in its own server-side solution, called Jedi++, and run tests to show HB hurts users and advertisers. His deeper worry: if Facebook Audience Network (FAN) or Amazon back HB, publishers may sell to them first and skip Google entirely. His comfort: as long as Google's DFP is the 'system of record for guarantees,' publishers still have to call Google.

2. What the Executive Is Really Thinking

Bellack is reading a competitive threat to Google's position in the ad stack. HB lets publishers ask many demand sources for bids at once, before Google's exchange gets to act. That breaks Google's 'first look' advantage and could put 'GDN and DBM (and the AdX buyers) at risk.' His response is two-pronged: (1) build a better server-side alternative so no business model only works in HB, and (2) gather data on HB's harms to users and advertisers to shape the narrative. He also identifies the structural moat that still protects Google — guaranteed reservation deals run through DFP, so DFP remains a 'must-call.' The whole memo is about preserving control over the auction sequence.

3. Key Management Lessons

Identify the exact mechanism that protects your business

What it means

Bellack names the specific reason Google stays essential: it is the 'system of record for guarantees,' and guarantees are the biggest chunk of publisher revenue. That, not brand or scale, makes DFP a must-call.

Why it matters

Most leaders describe their moat in vague terms ('we have the best tech'). Bellack points to a single workflow dependency. That precision tells the team what to defend and what is negotiable.

MBA Perspective

This is a clean Switching Costs argument. Publishers cannot route guaranteed deals around DFP without rebuilding their booking and reporting systems. The moat is operational lock-in, not feature superiority.

Real-world application

For any SaaS founder: write down the one workflow your customer cannot easily replace. If you cannot name it in one sentence, you do not have a moat — you have a feature.

Do not productize what you want to kill

What it means

Bellack explicitly does not want Google to ship an HB product. Building it would legitimize the model and undercut the server-side path he wants to win.

Why it matters

Companies often hedge by shipping a version of the competing approach. That hedge can cannibalize the strategy you actually believe in.

MBA Perspective

A commitment problem. Embracing two architectures sends a mixed signal to publishers, advertisers, and your own engineers. Picking one — 'double down on Jedi++' — concentrates resources and forces the market to a decision.

Real-world application

If you believe the future is your platform, do not ship a polished version of the rival workflow inside your product just to keep customers calm. You train them to want the wrong thing.

Use measurement as a strategic weapon

What it means

Bellack wants to test how HB affects 'latency, data usage, and battery' for users and 'viewability, CTR, brand lift' for advertisers.

Why it matters

If the data shows HB harms real metrics, Google can argue against it credibly to publishers, advertisers, and regulators. Data reframes a competitive fight as a quality fight.

MBA Perspective

Resource-Based View: proprietary measurement infrastructure is itself a competitive asset. Whoever defines the metrics often defines the winner.

Real-world application

Before your next competitive review, list the three metrics on which your approach genuinely beats the alternative — and instrument them. Without numbers, your story is opinion.

Anticipate the next move, not just the current one

What it means

Bellack already sees past HB. Publishers cannot make unlimited header calls without slowing pages, so the market will shift to server-to-server. The risk is that 'one HB company builds a more open Jedi than ours.'

Why it matters

Winning today's battle on HB does not matter if you lose the next round on the server-to-server architecture that replaces it.

MBA Perspective

Disruptive Innovation logic: the first wave (HB in the browser) creates the demand, but the durable winner is the second wave (server-to-server). Skating to the next puck matters more than blocking the current shot.

Real-world application

When a new model attacks you, ask: 'What is the constraint that will force this model to evolve?' Build for that next form, not the one in front of you.

4. Strategic Analysis (MBA Style)

Competitive Strategy

Defensive on the workflow, offensive on the architecture. Protect DFP's role as the guarantees system of record. Push Jedi++ as the open, server-side answer to HB so 'there's no business model that works in HB but not in our server-side solution.'

Risk Analysis

Three risks named or implied: (1) FAN or Amazon funding HB to flip publisher loyalty with offers like 'if we can pay you $10+, don't even bother checking with Google'; (2) HB becoming the new yield manager that sits above Google; (3) a competitor shipping a more open server-side product — Chris's 'Sith,' described as 'open season with no rules' — that publishers prefer over Google's.

Build vs Buy Analysis

The memo is a clear Build call. Bellack wants internal investment in Jedi++, not acquisition of an HB vendor. Buying an HB player would, by his logic, legitimize the model he wants to suppress. The document does not discuss acquisition options.

Market Dynamics

The ad-tech stack is shaped by call order. Whoever gets the 'first look' at impressions captures the best inventory. HB is a publisher-side rebellion against Google's first-look position. Two structural facts shape the fight: page performance limits header calls, and guaranteed deals concentrate revenue through DFP. Both favor Google — for now.

Long-Term Strategic Implications

If Jedi++ wins, Google keeps control of the auction sequence and the rents that come with being the must-call layer. If a rival builds a more open server-side product, Google's role shrinks to one bidder among many, and the guarantees moat is the last thing standing. The memo reads like a leader who sees both outcomes clearly.

5. Hidden Insights

• The real fear is sequence, not price. Bellack's worry is publishers being told 'don't even bother checking with Google.' Losing the call is worse than losing the bid.
• Openness is a competitive weapon Google may be unwilling to match. Bellack notes a rival could build a 'more open Jedi.' He sees Google's rules as a constraint a challenger could exploit.
• Guarantees are the last redoubt. The memo's calm closing line — DFP remains a must-call as long as guarantees flow through it — implicitly admits that everything else is contestable.
• Internal candor about uncertainty. Bellack hedges his own prediction: 'I still think the limit will be zero, but for sake of argument assume it's more like 3-5.' He is willing to plan for the case where he is wrong.