Ad revenue risk really breaks out by 2 major types of spending: 1. Brand Advertisers These companies don't measure direct ROI from their ads in the short run,

Ad revenue risk really breaks out by 2 major types of spending: 1. Brand Advertisers These companies don't measure direct ROI from their ads in the short run, but believe that it's important to put their brand and message out continuously to keep consumers engaged. Coke/Pepsi are the classic example. Very few people click on a Coke ad, land on a page, and put in their credit card to purchase a bottle of Coke right away. These companies care disproportionately about long term perception and therefore who their brand is associated with. They buy big event sponsorships for the same reason. 2. Direct response advertisers These companies continuously measure the ROI of ad spend vs. revenue impact, sometimes even on a weekly or monthly cadence. These companies have some limits (e.g. no ads on porn or violence), but generally are much more willing to buy less attractive inventory that has higher ROI. Classic examples here are Wish (or most ecommerce), SaaS, etc. Sales meetings and calls with advertisers happen almost only with brand advertisers because it is their business where decisions cannot be made simply by looking at the metrics. This is a lot of spend but, importantly, it's usually inefficient spend from an ROI perspective. If these folks exit the market, the direct response folks will see lower ad auction prices and proportionately increase their spend. These feedback loops take a few weeks/months to play out, but this means the impact of decreased brand spend is usually much less than appears on day 1. Twitter is a bit unique in that they probably have one of the worst direct response conversion rates, attribution, optimization, and overall performance. That's really where Elon should invest a significant amount of resources because that problem can be improved a lot with better adtech, rather than salesmanship. Google is generally the best at this ML based adtech optimization (people searching for products also tend to want to buy that product making their problem a little easier), followed by Facebook, followed by a bunch of 3rd party display platforms, followed by Twitter (in my experience as an ad buyer, but obviously not universal). If I were Elon, I would: 1. Create a new ad unit that can only be shown to verified users or adjacent to verified tweets - this should solve most ad fraud and content rule concerns. Ad unit should be launchable within a few weeks (harder technical problem because it requires context of what's adjacent in the feed, but can be hacked with rebates for improper impressions). Spend the next week selling this to brand spenders. 2. Spend 50% of my twitter energy hiring/empowering adtech experts (this is a playbook that has been done several times over) 3. Relaunch vine with trending videos on twitter (video ads) 4. Ignore search - technically hard problem with low 1yr ROI 5. Start decreasing organic reach for non-verified users in feed + upsell boosted posts -- this is Facebook Pages playbook and is super easy to implement technically (feed value model adjustment)