This is the briefing for the week of November 21st. Highlights this week are Spotify's acquisition of Megaphone, new Apple chips, podcast host reads, Amazon vs Sears internal competition, junk data and super monkeys.
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.
This week's essay was on the relative value of reach vs frequency. The interview was with Peter Fader (Wharton professor) on "but until you die" (highly recommended and free for all).
Next week's essay covers the importance of convenience through looking at the entertainment industry.
- TikTok:A good argument for NOT following the news. TikTok has been given a stay of execution. There IS a problem with allowing the Chinese government to control the algorithm on a massive consumer app to influence consumer opinion any way they see fit. But how this was handled wasn’g going to solve that problem anyway. Related: TikTok’s ad revenue in China has hit $27.2B. It has passed Baidu to become the second largest digital ad company in the country. It is still relatively tiny in the rest of the world, but it’s coming fast….
- Spotify: Spotify is acquiring Megaphone for $235MM. Megaphone began as Audiometric a tool for hosting podcasts (as well as providing analytics and monetization tools). It was acquired by Panoply, a podcasting network spun out of Slate.com (which was in turn spun out of Microsoft). Panoply stopped producing its own podcasts, and pivoted entirely to the Audiometric business, and then re-branded Megaphone. The acquisition will allow Spotify to offer streaming ad insertion (SAI) into podcasts outside of the Spotify platform. This is a big step towards Spotify becoming the defacto ad platform for podcasts like Facebook is for their newsfeed, and potentially allow them to monetize podcasts the way FAcebook montetizes display ads (i.e., orders of magnitude better than the alternatives). This is one of the most interesting and important things happening in marketing right now. Podcast ads may become the platform that replaces traditional television as the advertising and brand building anchor. If you missed it I recommend “Spotify and Podcasts” and “Joe Rogan and Podcast Advertising” for more of my take on this.
- Lyft: Unlike Uber, Lyft does not have a GrubHub delivery competitor (i.e., UberEats). But last week on their earnings call John Zimmer, Lyft’s CEO said two things:(1) He plans to expand the company’s delivery offerings by creating alliances with more businesses, including restaurants; and (2) Lyft won’t build a stand-alone food-delivery service like Uber Eats. He said: ““These businesses want a partner—someone to help them move their goods from point A to point B, but one that does not step in between them and their customers”. This sounds an awful lot like “Shopify for restaurants”. It’s a smart move. Meal Delivery has a LOT of competition, so rather than trying to build another miner looking for gold, just open a shop selling pick axes and allow the restaurants to become their own miners. Too many metaphors?
- DoorDash: On the surface Doordash is a direct competitor to UberEats and not whatever it is that Lyft is planning to build. But in their S-1 last week, they wrote, “We believe that by starting with food, we have created the most sophisticated and reliable logistics platform for local businesses. And, while food itself is a category that has a long runway for growth, we believe the network we have built ideally positions us to fulfill our vision of empowering all local businesses to compete in the convenience economy”. So they want to be a logistics platform as well. When you want a high valuation and interest rates are at all time lows, it is all about maximizing the addressable TAM.
- Apple: Apple announced their new Macs last week that use their proprietary chips. James Allworth plots Apple vs Intel chip performance over timeand find a chart Clay Christensen would be proud of.
- Advertisement of the week: German public health has created a parody of the “old WWI veteran looks back on the war” in the age of COVID. Funny even with subtitles. Worth watching. Also good: Ryan Renolds and Hugh Jackman have a new spot co-branding with Sam’s Club.
- SEO Training: SEO touches many many different organizations in a company, and most of the people in those organizations don’t know how to do it. But training them on what they need to know is itself a marketing problem. Ian Lurie has build a guide on how to train an organization on doing SEO. Key is that different groups need different things (engineers, non-SEO marketers, executives, etc). It is a fast read and recommended.
- Podcast Host Reads: More data that host reads outperform polished ads on podcasts. They are “higher fidelity” and fidelity matters a lot in marketing channels.
- Peloton: Peloton has signed Beyonce as a celebrity “partner”.This is smart. It gives them a celebrity face, but they have integrated the sponsorship beyond the superficial level. They have created themed workouts, promoting Beyonce’s latest album and providing free memberships to college students at historically black universities/colleges (HBUC).
- Facebook:Only 6% of Facebook’s posts are political in nature. Minority of the content causing majority of the problems.
- Amazon: A viral slide from a Social Capital (VC firm) slide deck that shows how Amazon took its cost centers and turned each into profitable businesses. The problem with internal groups is they can become complacent (who responds faster - your accounting firm or your internal accounting team?). Amazon solved that by pushing their internal service groups to provide their services outside of Amazon - to be “tested by the market”. This forced those groups to be high performing, but it also created service businesses that had the back-stop of a very very large customer (Amazon itself). This allowed these businesses to achieve scale very quickly. When it didn’t work right away, Amazon just bought a new business to become an additional “first best customer”.
- Sears: The flip side to Amazon’s choice to make their businesses compete in the “marketplace” is what happened at Sears. Sears made their divisions compete as stand-alone businesses, and ended up with situations like this: “Sears’s widely trusted appliance brand, Kenmore, was divided between the appliance division and the branding division. The former had to pay fees to the latter for any transaction. But selling non-Sears-branded appliances was more profitable to the appliances division, so they began to offer more prominent in-store placement to rivals of Kenmore products, undermining overall profitability.” Getting the balance right to keep your business competitive is Small differences are the difference between being the next Amazon and the next Sears.
- Fraud:If you want to be perfectly safe you should never leave your house. There are trade-offs in everything, and for each individual (and society) there is a theoretical “optimal level of safety” that is not 100% safe. If you are running a business there is also an “optimal level of fraud”. Too much and it destroys your business. Too little and you have made things too inconvenient for your true customers. Megan McArdle (author Upside of Down and columnist at the WaPo) has a great story about how trying to drive employee fraud to zero at a company she worked at doomed the firm. It involved ball point pens.
- Metrics become goals:When metrics become goals the metrics stop measuring what you think they do. But one way to solve that is to select AGAINST the metrics. Here is a story about science funding where the investors took a three-step process:(1) Identified flaws in the current system, (2) Estimated who would be negatively affected by those flaws, (3) Start writing checks to those people. This is a “good enough” solutions. Assume that you can’t identify who is great vs good, so instead just fund the people who are discriminated against in some way and assume that will give you a small edge vs everyone else. Don’t look for the best, just look for “good enough” that everyone else rejects.
- McDonalds:McDonalds is launching a loyalty program. Loyalty programs generally destroy value, but this one looks to be based on the Starbucks program that creates a TON of value (just not through Loyalty). I wrote about it here in case you missed it.
- Distribution:The Economist has an article on a Brazilian e-commerce company which delivers to remote villages along the Amazon river. Increased distribution is the most powerful lever in marketing, and increasing it is often really hard (like building the infrastructure to both deliver along a river, and the financing to allow these people to buy without traditional banking: ““Amazonians who lack savings or credit cards use zero-interest loans starting at 150 reais ($26) to finance their purchases; a whopping 85% of Bemol’s online sales are paid for this way.”)
- Managerial Incentives: A continuing theme at Marketing BS is that individual incentives matter more than company incentives when predicting what an organization will do. But here is an example of how banking analysts over-recommend companies that are large net issuers of their own stock.Those companies tend to under-perform, but they generate over-sized fees to the investment banks that employ the analysts. While the analysts are not supposed to set ratings based on the fees generated by their employers, they clearly are. This may be an example of the exception proving the rule and it is worth considering what the analysts TRUE incentives are…
- Legibility: You can realize something is good or better than an alternative without being able to fully explain WHY you believe it is good or better. But over time in most successful industries and companies legibility - the ability to articulate of quantify the “goodness” - increases. Rohit Bhargava writes about the perfume industry and the growth in products that smell like specific things (like outerspace, IKEA or Powells bookstore). The idea is that perfume can trigger emotions and shift moods, and rather market that idea with images, it can tell you specifically what feelings it will try an invote.
- Junk Data: In science research, “junk data is not just noise but disproportionately where interesting claims come from”. And: “In a 2003 study, 19 percent of teens who claimed to be adopted actually weren't, according to follow-up interviews with their parents. When you excluded these kids (who also gave extreme responses on other items), the study no longer found a significant difference between adopted children and those who weren't on behaviors like drug use, drinking and skipping school.The paper had to be retracted. In yet another survey, fully 99 percent of 253 students who claimed to use an artificial limb were just kidding.”
COVID-19 and the New World Order
- Bank Deposits: COVID has made it hard to spend money in many categories, which has led to people savings more than ever. When the vaccine hits, people will have lots of disposable income. Is your company ready?
- Traffic Accidents:There are a lot fewer cars on the road these days. But traffic fatalities were UP 30%. It seems with less cars, people drove faster and took more risks.
- Flying under COVID:The data is suggesting more and more than flying is not one of the dangerous activities you can be participating in (that said: Our family with three kids under 6 has decided driving to Southern California is a safer bet than flying). The airlines are doing a great job at cleaning the air to reduce chances of infection. But it seems there is an Achilles heel in the process: When the plane has to de-ice.So maybe no flying to Chicago this winter?
AI, Machine Learning and GPT-3
- AI and DoD: CSET surveyed US AI industry professionals for their views of working with the US defense department. The finding: “Most AI professionals are positive or neutral about working on DOD-funded AI projects”. Beware the loudest employees…
- Alexa: Amazon has made some significant improvements to Alexa to attempt to predict “what else” you might be interested in given the question you asked. They call this “latent goal discovery”. Think of this as “related products” for voice, with the added complication of figuring out when it is appropriate to offer a related product at all. Some examples:
- You: What is the weather at Seaside beach?
- Alexa: Here is the weather, would you like to use EarthCam to view Seaside beach?
- You: How long does it take to seep tea"?
- Alexa: Five minutes is a good place to start. Would you like me to set a timer for five minutes?
The blog post also gives examples of some of the challenges with early versions of the technology (when asked about a chicken recipe it responded with an offer to play chicken sounds. Which sounds like something my dad might do)
- BlockFi:BlockFi is a fintech start-up that offers loans backed by cyyptocurrency. They are looking for a CMO based in NYC.
- Changing Employees: From a Paul Graham Essay on colleges:“At one point they [the executive leadership] worried Lotus was losing its startup edge and turning into a big company. So as an experiment she sent their recruiters the resumes of the first 40 employees, with identifying details changed. These were the people who had made Lotus into the star it was. Not one got an interview.”
- Race and Gender in Silicon Valley: Researchers at Stern looked at the hiring rate of applicants by gender and race at Silicon Valley tech firms: “we find that women are 9-10% more likely to receive a callback compared to men, whereas Black, Hispanic, and Asian applicants are 8-13% less likely to receive a callback compared to White applicants. These outcome gaps do not cancel-out in the later stages, as female and White applicants are more likely to receive an interview and offer.”
- Super Monkeys: Scientists have identified the genes that caused human brains to diverge from apes. They have grafted those genes onto monkey brains, resulting in: Increased size of neocortex; folding of the brain, similar to how a human brain is folded; The relevant pro-generative cell type, which produces neurons, was increased; and it increased upper-layer neurons, which are the neurons that increase during evolution. They decided NOT to allow the fetus to be born which could result in a second intelligent species and lead to humans downfall ala Planet of the Apes. But the technology is out there now…
Keep it Simple,