FBX discussion

[ … ] Not shown: Greeting; the group met after Tuesday’s discussion.

Fundamentally, we think the heart of the gamble revolves around how much Facebook believes it can shape the marketplace within which it operates. We have two options: Go For Broke, or Take The Money.¹ (Editorial note 1)Editorial · 1Known at the timeThe email turns a running argument about FBX, Facebook’s exchange for outside ad buyers, into a bet on one question: how far Facebook can shape the ad market rather than plug into it. Two named paths make the choice easy to discuss; they also hide options in between, such as opening some features and withholding others.

[ … ] Not shown: Inventory and identity data listed as the first two advantages.

3/ Platform advantage. Our developer ecosystem could be leveraged to fortify both inventory (with an ad network) and data (with Neko or equivalent install focused product).² (Fact note 2)Fact · 2With hindsightWhere this copy comes from: the email is Exhibit 40 to a declaration of 16 May 2018 by Six4Three’s counsel in Six4Three v. Facebook (San Mateo County Superior Court, CIV 533328), which quotes this sentence in paragraph 41. The court files were leaked anonymously to the journalist Duncan Campbell, and Computer Weekly and NBC News published them on 6 November 2019; NBC reported that they “remain under protective order” and that Facebook had not questioned their authenticity. We did not find the email in the Six4Three documents the UK Parliament’s DCMS Committee published in December 2018.SourcesNBC NewsSealed exhibits from Six4Three v. Facebook (658 MB PDF)6 Nov 2019NBC NewsLeaked documents show Facebook leveraged user data to fight rivals and help friends6 Nov 2019DuncanCampbell.orgFacebook Leaks6 Nov 2019UK Parliament, DCMS Committee (Internet Archive copy)Note by Chair and selected documents ordered from Six4ThreeDec 2018

[ … ] Not shown: Consolidation and disintermediation listed as the first two threats.

3/ Friction. If people get very comfortable working with DSPs it will be hard to move budgets later even if we have a superior product.³ (Editorial note 3)Editorial · 3Known at the timeThe sharpest threat on the list, because it is about habit, not technology: once advertisers run budgets through outside buying tools, a better Facebook product may not win them back. It is also the argument for moving first, which the rest of the email weighs against the cost of waiting.

Our decision to deprioritize FPAN/Starling/Neko was essentially a decision not to press our advantages⁴ (Editorial note 4)Editorial · 4Known at the timeA candid line. It names an earlier staffing decision as a strategic choice: not hiring for the products was, in effect, a bet that Facebook’s advantages would last. That reframing is what lets the email ask for engineers rather than for agreement.

[ … ] Not shown: Hope the advantages outlast the delay; debate; a sketched chart.

The black path is the Go For Broke path where we wait and withhold premium features from FBX. Until we actually launch our products (interfaces, fpan, starling, etc...) we all agree we will make less money than if we had released news feed and custom audiences to FBX.⁵ (Fact note 5)Fact · 5With hindsightOn 26 March 2013 Facebook began a desktop-only alpha test of FBX ads in News Feed, with TellApart, MediaMath and Nanigans as the buying partners. Our sources do not say how that relates to this thread.SourceTechCrunchFacebook Expands FBX Retargeted Ads, Based On Your Online Browsing, From Sidebar To News Feed26 Mar 2013

[ … ] Not shown: The two outcomes of the Go For Broke path.

The blue path is the Take The Money path where we follow the lead of Antonio and his team and integrate with FBX.⁶ (Fact note 6)Fact · 6With hindsightAntonio Garcia-Martinez is on the To line. TechCrunch later described him as “a former Facebook product manager who helped build FBX”.SourceTechCrunchFacebook will shut down FBX, its desktop ad exchange25 May 2016 We will certainly make more money in the short term

[ … ] Not shown: The choice depends on sensitivity to the variables and risk tolerance.

As for the opinion of our group, we have been imagining a future where we have a premium channel in News Feed to capture brand and high end DR dollars at high prices and then another channel with FPAN replacing RHC for everything else.⁷ (Fact note 7)Fact · 7With hindsightThe email does not define FPAN, RHC or Starling. On 30 April 2014 Facebook launched Audience Network, its mobile ad network, so that app developers could make money without selling their own ads. No source we found says it is the product the email calls FPAN.SourceTechCrunchFacebook Audience Network Mobile Ad Network Launches At f830 Apr 2014

[ … ] Not shown: That path needs a short time to launch; staff are short.

Consequently, we propose to reframe this choice in the following way: Are we willing to commit to fully staff Starling, FPAN, and Neko in the 3-6 month time frame?⁸ (Editorial note 8)Editorial · 8Known at the timeThe best move in the email. An open question about market power becomes a yes-or-no question about hiring, with a deadline. The trade-off is that strategy now depends on recruiting, which the email itself says is short of engineers and designers.

[ … ] Not shown: If so, Go For Broke, as a substantial investment.

If not, then the risk is just too great and we should just Take The Money we can get, when we can get it, by opening up more features to FBX.

The one thing we should most likely not do is take neither path,⁹ (Editorial note 9)Editorial · 9Known at the timeNaming the drift option as the worst one is the email’s practical point. Withholding features from FBX while not staffing the alternatives would give up the short-term money and the long-term position together. which means making a decision relatively quickly is important.

[ … ] Not shown: Sign-off before a flight to Antarctica.