Court Exhibit·2 OCT 2018
Why Scale Is The Real Moat In Ad Auctions
In ad markets, scale isn't just an advantage — it's the price of admission to compete at all.
Source document — Microsoft 'Baja' Financial Model Presentation to Apple — October 2018 · United States v. Google LLC (Search) · 1:20-cv-03010 (DCD), Trial Ex. UPX0116 — DOJ public archive
Excerpt · In Kate Plikus's own words
Scale evidence: Impact from original Yahoo supply
Key Takeaways: We have empirical proof from our Yahoo! Search deal that incremental volume drives up RPMs. With the influx of Yahoo! Traffic into our ecosystem in 2012 we were able to realize a one-time boost to our RPMs driven by injection of scale. In addition, having a bigger marketplace enabled us to increase our innovation which had driven up our RPMs even further over the next couple years.
We believe that injection of Apple volume would enable us to realize RPM lifts from scale and innovation similar to the Yahoo! deal. We would expect that these lifts would help us get to [redacted] of Google on iPad and Mac in the US. In international markets we would be able to get to [redacted] of Google on iPhone.
Notes: RPM increases with scale as advertisers increase bids to be on top in the auction, and algorithms improve with more data. The curve is initially steep as small increases in scale meaningfully improve advertiser participation, making it possible to offset fixed cost of investing in another platform. Product innovation shifts performance to a higher RPM curve, not along the same curve. Most of subsequent performance improvements come from innovation and demand expansion.
1. Core Message
Microsoft is pitching Apple on a search deal. The core argument: more search volume mechanically lifts revenue per thousand searches (RPM). They claim empirical proof from the 2012 Yahoo! deal — that a sudden injection of traffic gave them a one-time RPM boost, and then enabled further innovation that lifted RPMs again over the following years. They expect Apple's volume would do the same, helping them close the gap to Google by specific (redacted) percentages on iPad/Mac in the US and iPhone internationally.
2. What the Executive Is Really Thinking
Microsoft is trying to convince Apple that Bing can be a credible alternative to Google — but only if Apple gives it the volume to become one. The pitch quietly admits Microsoft's core problem: without more queries, Bing cannot match Google's RPMs, and without matching RPMs, Microsoft cannot pay Apple enough to win the deal. So Microsoft is asking Apple to believe a forward-looking promise grounded in a past data point (Yahoo 2012). The Yahoo example is doing heavy lifting here — it is the proof that Microsoft's claims about scale are not hypothetical.
Underneath the pitch is a clear view of how search economics actually work: advertisers bid more when more users are watching, and machine learning systems get sharper with more data. Both effects compound. Microsoft wants Apple to understand that paying Bing more over time is not a hope — it is a mechanical consequence of giving Bing scale.
3. Key Management Lessons
Use past deals as proof, not promises
What it means
Microsoft does not just claim scale lifts RPMs. They point to the 2012 Yahoo! integration as empirical evidence it happened before.
Why it matters
Prospective buyers discount promises. They believe receipts. A prior comparable outcome is the most persuasive thing a vendor can bring to a negotiation.
MBA Perspective
This is Resource-Based View in practice — Microsoft is framing the Yahoo experience itself as a proprietary capability: they have done this integration before and know what happens when traffic floods in.
Real-world application
When pitching a large client, lead with a structurally similar past deal and the measurable outcome, not with projections.
Separate one-time gains from compounding gains
What it means
Microsoft carefully distinguishes between two effects: a one-time RPM boost from the scale injection itself, and ongoing RPM lifts from innovation that scale made possible.
Why it matters
Most executives lump all benefits into one number. Separating step-change effects from compounding effects makes the model more credible and tells the buyer where the durable value sits.
MBA Perspective
This is Economies of Scale done honestly — recognizing that scale produces both static cost/revenue advantages and dynamic learning advantages, and that the dynamic ones matter more long term.
Real-world application
When forecasting deal economics, split the model into immediate impact, learning curve impact, and product-driven impact. Each has a different probability and timeline.
Understand the shape of your scale curve
What it means
Microsoft notes the RPM-vs-scale curve is initially steep — small volume gains produce big RPM gains — because they unlock advertiser participation and offset fixed platform costs.
Why it matters
If you know where you sit on the curve, you know how much a marginal customer is actually worth. A subscale platform should pay more for the next unit of demand than a saturated one.
MBA Perspective
Classic Network Effects with a fixed-cost overlay. More users attract more advertisers, more advertisers improve auction density, denser auctions raise prices.
Real-world application
In any two-sided market (marketplaces, ad networks, payment networks), map your unit economics as a function of volume. Use that curve to price early deals aggressively and later deals conservatively.
Innovation shifts the curve, scale moves you along it
What it means
The document draws a sharp distinction: scale moves you along the existing RPM curve; product innovation shifts you onto a higher curve entirely.
Why it matters
Leaders often confuse the two. Adding users is not the same as building a better product. Both are needed, and they compound differently.
MBA Perspective
A cleaner version of Disruptive Innovation thinking — improvements that change the performance ceiling versus improvements that just push you closer to it.
Real-world application
When reviewing growth initiatives, ask: does this move us along the curve (more of the same) or shift the curve (new capability)? Fund both, but track them separately.
4. Strategic Analysis (MBA Style)
Competitive Strategy
Microsoft is competing on credibility, not just price. They cannot beat Google on current RPMs, so they reframe the contest around future RPMs that Apple's own traffic would create. The strategy is to make Apple a co-investor in Bing's improvement.
Risk Analysis
Microsoft is trying to defuse Apple's biggest fear: that switching from Google means leaving money on the table. By arguing scale closes the gap, they reduce Apple's perceived downside.
Build vs Buy Analysis
For Apple, this is a buy-vs-build decision on search. Apple could build its own search engine, but doing so means starting at the bottom of the same scale curve Microsoft describes. Microsoft's pitch implies: do not build, because scale takes years and you would be paying the steep part of the curve yourself.
Market Dynamics
The document reveals that search advertising is a scale-dominated industry. The auction format means whoever has the most queries gets the most advertisers, who then bid the most, which funds the most innovation. This is a self-reinforcing loop and explains why search has consolidated to so few players.
Long-Term Strategic Implications
If Apple agrees, Microsoft gets the volume needed to make Bing structurally competitive — possibly for the first time. If Apple refuses, Microsoft's argument confirms its own ceiling: Bing's RPM gap to Google may be unclosable without an external traffic injection of this size.
5. Hidden Insights
- Microsoft is implicitly admitting that without a partner like Apple, Bing cannot close the gap with Google on its own. The pitch is partly a confession.
- The redacted percentages are the entire commercial heart of the deal — how close to Google Bing thinks it can get. Whatever those numbers are, they are below 100%, meaning even with Apple's traffic, Microsoft does not expect to match Google.
- The document quietly explains why Google's position is so durable: each query Google handles makes its auctions denser and its algorithms smarter, which raises advertiser bids, which funds more innovation. Breaking that loop requires a one-time external shock — exactly what Microsoft is asking Apple to provide.
- The phrase "offset fixed cost of investing in another platform" hints that Apple itself faces this math if it ever considered building search in-house.
How this surfaced
- Source type
- Court Exhibit
- Case / record
- United States v. Google LLC (Search)
- Citation
- 1:20-cv-03010 (DCD), Trial Ex. UPX0116 — DOJ public archive
- Date authored
- October 2, 2018
- License
- Public domain
- Original
- View the primary source →
Netflix Management ·Netflix ·21 APR 2014
Pick Your Fight: Beat Each Rival On A Different Edge
When you face two different types of competitors, you need two different weapons — not one slogan that pretends to work against both.
Reed Hastings & Netflix Management ·Netflix ·16 JUL 2020
Tell investors bad news before they ask for it
The best leaders don't hide a slowdown behind a great quarter — they name it, explain it, and move on to what comes next.
Ted Sarandos & Greg Peters & Spence Neumann & Spencer Wang ·Netflix ·18 OCT 2023
How Netflix Turned Content Spending Into a Profit Machine
When your cost structure and your competitive moat are the same thing, spending more is actually the safest move you can make.