Cambridge Analytica. google.com/amp/abcnews.go…
Gonna use this article to link some bits.
“Warp your perception of what’s real around you.” 🤔 
Like um this?!? (Link to thread: )
“Harvest their data.” And “Then, that app would crawl through their friend network and pull all of the data from their friends also.”
1) On that “harvest”. Helped acquire >5,000 data points on every voter.
The ‘dashboard’ allowed drill-down capability to individual voter level.
2) Please 📌 the “tunneling”.
1) “weaponizing the Internet”. And 2) The campaign saying it “relied on voter information gathered by the Republican National Convention.” 3) The claim the campaign didn’t use data from other sources.
1) “weaponized” term shows up elsewhere. Like here from a so-called grassroots amplifier/ally.
🤔 Selaty - Dir. for Citizens for Trump “grassroots” group also managed 42 groups/pages on FB.
Here are the public ones from Selaty’s FB profile 1/2.
Here are the public ones from Selaty’s FB profile 2/2.
2) “Relied on info from RNC”.
GOP data repository is one ginormous loop of data in/out across the entire apparatus.
In 2016, Cruz and Carson hired Cambridge Analytica early on. All of the data, models, etc. from those campaigns fed in/out of the GOP data repository. So, when Trump campaign accessed RNC Data for voters, the data append process had already integrated into 1.
3) That claim from Trump campaign. Pfft. Cuz, there were embedded staffers from Cambridge Analytica. Oh, and Google and Facebook too. Oh, and RNC Data Analysts who worked on the GOP data repository.
Not to me to mention those subcontractors. Like Harris.
More on Harris.
And, of course, ALL of this. newsweek.com/trump-cambridg…
Yo GOP. See ⬇️ totally ties you into the whole apparatus. See what they did there. 😳🤣
And, can’t forget all the bragging that laid it all out. 👀 Note again the tie in to the GOP Data Repository.
This thread breaks down quite a bit too.
From ⬆️ Meeting of key staffers to plan: KAC. Kushner. Bannon. Bossie. Parscale. RNC analysts. Media & Data vendors.
👀👇 “regularly updated voter data” to “determine media plans and voter contact”. Translation ➡️ data models (testing and refining).
👀👇 Data modeling/voter scoring for field ops fell to tried & true RNC vendors - TargetPoint Consulting, Causeway Solutions, Deep Root Analytics. The mega data thread goes into each of these in-depth. 🚨👀 CA handled “persuasion advertising and targeting.”
Speaking of Deep Root Analytics. Let’s not forget that ginormous data file left wide open.
Ok. Let’s walk through the info from breach. A lot of in-depth into how the repository was built is in the thread as well.
GOP data repository ➡️ 9.5 Billion data points. Scoring 198 million potential voters on likely political preferences using modeling across 48 different categories.
At time of breach discovery, last data update around Jan 2017 inauguration. 🚨a “treasure trove of political data and modeled preferences used by the Trump campaign.”
dra-dw directory ➡️ “A massive collection of data prepared specifically for complex analysis.” 1.1 terabytes. “File directories named for a number of high-powered and influential Republican political organizations.”
“An additional 24 terabytes was stored in the warehouse but had been configured to prevent public access.” Including some inaccessible files that indicate other allies/data users.
More about Deep Root Analytics. I did a deep dive in a thread as well. Note: the RNC moved in quickly to ‘coordinate their data team’s efforts with the Trump campaign’
Data Trust ➡️ GOP data repository. The data_trust folder contained “comprehensive and detailed repository of potential 2016 voter information.” NOTE: last section - RNC ID ➡️ unique link to merge different sources of data.
To put RNC ID into context. From my database experience, there was always a unique identifier that allowed for data to be properly matched/appended to each data record. So, a way for the computer to know who is who to match data uploads. Hope tat makes sense. 🤣
All sorts of data included: First name. Last name. DOB. Home address. Mailing address. Phone #. Registered party. Self-reported racial demo. Voter registration status. Etc. Also included modeled data. Every field will populate if data available based on RNC ID.
Vendors were paid by RNC for their contributions. Last section: more data = better targeting. Modeling/targeting helps identify who your are talking to so that you can adapt what messages each audience receives. What buttons to push. What issues are triggers. Etc.
Lastly, the breach showed that each voter is scored across 46 columns. Indicating the modeled likelihood of supporting the policy, candidate or belief. 🤔 So each voter scored for 46 categories.
Will close this out with one last nugget of 🤔
From ⬆️ Darren Bolding left role as Chief Technology Officer at RNC to join Cambridge Analytica as CTO. 🤔🤔🤔
Praise be! Thank ya @foxycuda
