Twitter Community Notes, Google Ads, and Long vs Short term
What cannot be measured is sometimes still very important
I have nothing witty to start with today. Let’s jump into the post, but before we do…
This Week’s Sponsor
Time is Money: Maximize Efficiency, Minimize Costs - Outsource the mundane!
We recently found out that founders spend 30-45% of their time on back-office operations. That’s 8-10 hours/week NOT spent on growing the business. As a founder you probably don’t enjoy filing payroll tax, sending employee contracts, I-9 verifications, vendor payments, contractor management etc. Learn more here.
Back in May, during an interview with CNBC, Elon Musk volunteered that Twitter (X) had lost $40MM from advertisers who had cancelled their business specifically because they had “got community noted” and that Twitter (X) refused to take the community note down (see an example above).
Community Notes launched on Twitter (back when it was still called Twitter) in December 2022. Individuals can sign up to be “contributors” who provide “notes” on any tweet (post) to clarify or contradict the original content. Other community notes participants vote on the contributions, and if there is consensus the note is added under the tweet (post). It is unclear when the community notes began appearing on advertisements, but it seems to have been either immediately upon launch of the feature or very soon thereafter.
There has been much discussion about how Twitter (X) is insane to to allow community notes on advertisements. Some easy to find examples from the first page of Google results on the topic:
Vice: “Twitter has an ads problem”
Social Media Today: “That can’t be great for business”
Performance Marketing World: “Naive at best”
WSJ: “The possibility of receiving a Community Note is one more reason for advertiser caution”
Musk himself positions Community Notes on advertisements as an example of his free speech absolutism. But is he really sacrificing the market capitalization of Twitter (X) for this particular principle? I would argue he is not, and that the critics of Community Notes on ads are missing the bigger picture.
A key theme here at Marketing BS is that, while it is important to to track metrics and be “data driven”, it is also important to understand that most things that can be quickly measured are short-term in nature. It is relatively easy to measure the value of a paid-click on Google search. It is much harder to measure the impact of a great TV advertisement on long term price elasticity. Just because it is harder to measure the latter, does not mean it does not create value.
Google ads use something called “Quality Score” to change your “effective bid” and determine your placement in the search results. If Google thinks your ad is very good and relevant you will pay less per click than a “bad” ad, and may even appear above that ad (and get more clicks). The best thing you can do to drive your paid search performance is to optimize your ad account, ad copy and ad targeting to obtain high quality scores — you will get more traffic for every given dollar you spend.
Mostly quality score is based on click-through-rate [CTR]. This is short-term profit maximizing for Google. Advertisers pay based on [spend per click] x [number of clicks], so Google does not really care about cost-per-click [CPC], they care about CPCxCTR. It makes sense, even in the very short term, for Google to reward advertisers with high CTR and charge them lower CPC and move them to the top of search results.
But Quality Score is not JUST CTR. In addition to that short term profit maximization, you can get your quality score higher than other advertisers with the same CTRs. You do that by providing more value to the user (in a way Google can measure). So Google looks at how relevant your ad and landing page is to the search term the user typed in. They look at how engaged users are once they click to your website. They look at how likely it is that someone checks out your website, but then comes back to Google to do the same search again (a sign that your website did not meet the user’s need). All of these things do not help Google in the short term, and in some directly cases hurt Google’s revenue (Google would make more money on a given search if someone could not find what they were looking for and searched a second time — Google would get twice as many ad clicks! But Google penalizes those advertisers rather than rewarding them).
Why would Google do that? I think we have moved well beyond the idea it is because they don’t want to be “evil”.
I think the reason is, is that Google is trying NOT to optimize revenue for the individual search. Instead they are trying to maximize revenue for the individual searcher. If you search and click on an ad and have a good experience, you are likely to click on an ad in the future when you search. Conversely, if you click on an ad and have a bad experience, you are less likely to click on an ad the next time you search. If the experience is bad enough, you may even go in search of a new search engine entirely.
It is possible that Google can measure that long term impact today, but is is very unlikely they had the ability to do so fifteen+ years ago when they first opened up their ad auction and implemented quality scores. They originally designed the system NOT to be short-term profit maximized. Instead they optimized the short term the best they could, and then chose some proxies for how they believed long term value could be created, knowing that those decisions would hurt short term revenue.
Better Ads are Better for Everyone
Community Notes on ads on Twitter will clearly drive away some advertisers. My only surprise is that is was ONLY $40MM in annual spend. I expect that number is under-stated and represents two very distinct advertisers that TOLD Musk that they were cancelling because of the ads. I expect there were many more who stopped or reduced advertising and never told anyone (or told someone in the organization, but it was never relayed to Musk).
But that $40MM+ in cost comes with at least two benefits:
Consumers can now trust the ads they see on Twitter (X) more. Both directly because they can see the Community Notes, but also indirectly, because the bad-acting advertisers will have left the platform
The remaining advertisers now have the halo effect that their ads are more trusted by the users. Users will be a little less skeptical of the ads, resulting in, on average, higher CTR and higher intent to purchase
Both of those results should create more value on the Twitter (X) platform, allowing the company to both charge higher ad rates, and to obtain more ad impressions.
Can all of that be measured? Definitely not directly. The cost comes immediately when the advertiser cancels. The benefits accrue over time, and are co-mingled with all the other things that are changing, both inside and outside Twitter (X). If Twitter’s traffic drops over the next year or ad prices per click decrease, that does not mean that Community Notes on ads was a bad decision. It LIKELY means it was still a good decision, but something else drove down company performance. Unfortunately there is no real way to know for sure, and people like me that think Community Notes on ads are a good idea will always be able to tell a story we were correct — my belief is almost by definition unfalsifiable.
But I will continue to believe it anyway.
Keep it simple,
Edward