This is the briefing for the week of October 27th. Most important links this week are the data putting blogs on subdomains vs folders, the breastfeeding study (which has far reaching implications for start-ups relying on research), and Amazon's move into collecting third party reciepts.
Startup Curated subscribers can get access to these briefings on four days early (on Tuesdays), as well as a long form essay on an important marketing topic, and exclusive interviews with successful CMOs for free by subscribing here: https://marketingbs.substack.com/startup
This week's essay was on the importance of Inputs vs Outputs when building a business. Next week's essay covers what matters in manage product and company reviews.
Onto the briefing:
- Quibi: The short-form, mobile-only, professionally produced paid subscription video company (not to be confused with the short-form, mobile-only, amateur produced free subscription video company called TikTok) is shutting down. Unsurprising. But many huge successes would have been called “unsurprising” if they had failed (Airbnb, Uber and Google were all thought to be ridiculous ideas when they started). The difference here was the size of the bet before any form of the product was put in front of customers. I think it’s called Hubris.
- Google Antitrust:The justice department has initiated antitrust action against the search company. While earlier reports threw the kitchen sink against Google, this salvo is extremely targeted. They accuse Google of using its monopoly-like position to pay for exclusivity agreements to dominate distribution. Google pays Apple $8-$12B/year (10-15% of Apple’s annual profit!) and Mozilla $400-$500MM/year ((90-100% of Mozilla’s revenue) to be the default search engine on Safari and Firefox respectively. The justice department argues that Google can afford to do this because they monetize better than any other search engine, and then they use that data to improve their product, making it harder for any competitor to get the data they need to compete. Google has released their counter argument:
- They pay to promote their service (like everyone does). This is like buying shelf space more than "exclusivity agreements" since they pay only to be the default - anyone can change the default to be whatever they like without any difficulty
- "competition is only a click away"
I am not a legal expert, but it seems the Justice Department has a reasonable case here. But what are the ramifications if Google loses? Are they not allowed to pay for distribution (and therefore Apple and Firefox switch to Bing - at a much lower price point)? Is no one allowed to pay for distribution (Apple’s profit takes a huge hit. Mozilla effectively goes under)? It feels like Google will pay a fine and then the world will continue without any significant change.
- Facebook Research:Facebook is still taking heat from allowing a researcher to have access to user data (with permission from the individuals collecting the data, but not from the friends of those individuals), but then passing that data to Cambridge Analytica. Facebook was made to promise not to let this happen again. On Friday the WSJ reported that a researcher at NYU was collecting user data (with permission from the individuals collecting the data, but not from the friends of those individuals). By all accounts the NYU researchers have good intentions - they are collecting data about political ad-targeting. But if they are allowed to proceed, the only real difference between this and the Cambridge Analytica scandal are those claimed intentions. Facebook shut the researcher down (as required by their commitment). Now many of the same people who were upset about the privacy violation in 2016 are upset at that Facebook is preventing the privacy violation in 2020. As I regularly say, “Privacy concerns are never really about privacy”. Benedict Evans: “Facebook must control privacy! It must stop people stealing our data! Also, Facebook must let any random academic scrape whatever they want”
- Utah: The opposing candidates for Governor of Utah teamed up to run public service announcementsasking people to work together after the election, “no matter who wins”. This makes sense. Whoever DOES win may find an ad like this makes it easier for them to govern. The one who loses still gets the halo on their personal brand for whatever it is they do next. It may not be good for the losing PARTY, but this is an example of individual incentives overpowering group incentives, for the betterment of overall society.
- Facebook Gaming: Facebook has launched their own cloud-based, free, no subscriptions, no controllers gaming platform (available on most platforms - except iphones)
- Subdomains: Should you put your blog on a subdomain or a subfolder? i.e., blog.company.com or company.com/blog. Google says it does not matter. But many SEO experts will tell you based on experience that the right choice is a subfolder NOT a subdomain. Stephen Kenwrightshows some recent data from a client where he moved a blog from subdomain to subfolder and says “...I have never seen the opposite happen when that switch is made...” I haven’t either. Never trust what Google says…
- TheDodo/PetPlan: TheDodo is a leading content site that produces social-friendly pet and cute animal videos. They traditionally monetize by selling brand advertisements to companies that want to reach pet lovers. Last week they announced they are taking an equity stake with PetPlan pet insurance. PetPlan will be relaunched as “Fetch by TheDodo”. Content sites are good at building traffic and bad at monetizing. Traditional companies are good at monetizing and bad at attracting and holding an audience. Expect to see more and more of these types of “mergers”.
- YouTube: The largest advertisers in America are Comcast, AT&T, P&G and Disney. You can find similar lists for Facebook, paid search, podcasts, OOH, etc. Neoreachtracks the biggest spenders on YouTube and updates it weekly.
- YouTube 2: Apptweek tried to estimate the impact of YouTube influencer sponsorship(i.e., direct ads, not through the YouTube platform) on app downloads. They found, “no correlation between downloads and a YouTuber’s number of subscribers or the number of views per video”. Be careful.
- Billboards:It feels like the most creative advertising these days is happening in OOH. I loved this one:
Market Research (new category this week)
- Leaders of Social Change: 19% of marketers want to be leaders of social change. Only 9% of customers feel the same way. But marketers THINK 24% of their customers want to be social change leaders.
- Focus Groups: A chilling behind-the-scenes account of running beauty focus groups
- Breast Feeding:Data shows that babies that are breast fed are healthier. But this type of analysis suffers from significant selection effect. Rich mothers are far more likely to breast feed than poor mothers (for example), so how much of the fact breast fed babies have better outcomes is from breast feeding and how much is from having rich moms? (Emily Oster says it is almost entirely rich moms). New research finds there is a third driver. When research says something is "good", then smart/educated/informed people are more likely to change their behavior to do the "good thing" - which then effectively INCREASES the impact of the selection effect. Publishing research findings cause smart people to change their behavior more than dumb people. So when follow-up studies look to see the impact it seems to be even bigger than the initial study.
- Privacy:I’ve long argued that while people SAY they care about privacy, they actually do not (and they don’t seem to make any decisions consistent with a belief that it is important to them). Now their beliefs seem to be catching up with their actions…
- Amazon: Is paying customers $10/month to share ten purchase receipts from non-Amazon merchants.The best targeting comes from lookalike audiences built off purchase behavior. Amazon has a LOT of purchase behavior, but not all of it. I expect this is an experiment to see how much better their targeting capabilities will be with the incremental data. My bet: A little better, but not enough to continue the experiment. Related: An introduction to advertising on Amazon’s sponsored display platform.
- Gaming Metrics:Financial analysts have figured out that when specific (positive) words are use on earnings calls they predict future stock gains. So now CEOs are gaming the system by training themselves to use those words.When the metric becomes the goal it stops being a useful tool example #678.
- Unions: I came across this old Paul Graham essay. When you are growing fast you don’t care about optimizing expenses. When $1 in spend gets you $100 in impact, trying to optimize to spend $0.90 instead is both a bad use of time, and a way to slow down the (profitable) growth of the business. But as a business or marketing channel gets to scale and “matures” you can (and should) try and optimize to reduce the input costs. Graham’s argument is that fifty years ago (or so) MANUFACTURING was a growth industry and it stayed a growth industry for a long time. So the right way to deal with employees (input cost) was "pay them whatever it takes to keep growth going. Don’t optimize on the $1 that is making us $100”. But eventually manufacturing became “mature” and the “right” choice was to start optimizing costs. The result was growth in unions when businesses did not care about employment costs, and then a decline as cost reducing became the most important driver. What was unique was how long the growth cycle lasted.
COVID and the New World Order
- Netflix: Growth slowed down in Q3. Stock dropped 5%.
- P&G: Had its biggest sales increase in 15 years, and has responded by hiking marketing spend by $100MM. So many companies treat marketing spend as an output instead of an input…
- Unilever: The P&G competitor also so significant sales growth in Q3(+4.4% globally, +8% Americas). And they too responded by "boosting" marketing spend in Q3, and will do so to a "greater extent" in Q4.
- Transit: Some good data from Governing.com on how transit systems are hurting in this COVID era. Indianapolis usage was down 43% YoY in August. Chicago down 98% at the end of March. These systems make ~30% of their operating costs from fares and many were operating on the edge before COVID.
- The Flu: Some data from The Daily Shot on the rate of influenza in Australia by month since 2015. Basically no flu season in 2020
- Apex Technologies: Is looking for a Chief Growth officer. The company builds “Supply Chain Technology” including self-serve automation. They created the “Pizza Portal’ for Little Caesars. The role is responsible for global revenue and gross margin delivery including marketing, sales, channel development, pricing and contracts. Role could be based anywhere in the US.
- Onboarding: Alex Danco, a “cyrpto-person” at Shopify, has written “Lessons from his first 6-months”. While some of it is Shopify-specific, other lessons are applicable for those working at any big company. I particularly agree with “get inside the heads of everyone you will be working with, since you are not going to be effective actually doing your job initially anyway”. When I started at Expedia I spent the first 90 days going to lunch with 90 different people around the organization. It paid dividends down the line when I needed their influence to get things done outside of my own organization. Sometimes you can influence with data or authority, but often “appeals to friendship” or “appeals to principles” work better - and that only works if you have become friendly and understand their principles.
- Halloween costumes: Searchengineland shares data on the most popular Google costume searchesthis year. Top “trending” include CobraKai & Dungeon Master. Top kids, Supergirl (Nevraumont family 2017) and Flamingo. Top couple costume Bonnie and Clyde. Good to double check so you don’t end up being the third Mandalorian to join you Halloween Zoom Party.
- The Future:From a 1962 Italian magazine imagining what the future will look like. The claim on Reddit was it represented 2022,but Scopes believes no specific date was claimed.