Tuesday Jan.12, 2021

📦 Amazon's anti-return policy

_Roku's Golden Arm_
_Roku's Golden Arm_

Hey Snackers,

The best products to come out of the Consumer Electronics Show (CES) so far: a digital scroll, a rollable phone, and a wine-pouring robot.

Stocks dipped as investors worried about the likelihood of more Big Tech regulation following the US Capitol riots. Meanwhile, House Democrats introduced an impeachment article against President Trump, eight days before he's set to leave office.

Unslip

The anti-return policy: Walmart and Amazon don't want your stuff back

Returning your returns... So meta. The new approach to online returns: don't do them. From Amazon, to Walmart, and Chewy, companies are starting to say “keep it,” according to the WSJ. They’ll refund you for that tight-fitting beanie — but they don't want it back. Why so nice?

  • It's a price thing. Some returns are more expensive to process and ship than the products actually cost — especially cheap or heavy items. Companies use AI to do the math.
  • Returns can cost companies $10 to $20, excluding transportation. So it makes more economic sense to just refund you for that $7 PopSocket (iPhone was allergic).

Do "one-size-fits-all" fuzzy socks count?... The number of ecommerce returns in 2020 jumped 70% from 2019. More than half of that was driven by soaring online sales. A fourth was because you didn’t want to return online orders at physical stores. And a smaller part was because... clothes didn't fit after gaining the Quarantine 15.

  • For customers, this is clearly a win. Your pile of "return" items has been growing for years since the psychological hassle is just too much to bear.
  • For companies, making customers happy and losing less money are wins. But it could be a slippery slope ("Amazon return fraud" is apparently a dark-web career).

Score one for brick-and-mortar... Ecommerce takes all the spotlight nowadays, but we often overlook the big, hidden costs of online shopping: returns. According to Shopify, online purchases are returned at least twice as much as physical purchases. According to others, they're returned three to four times as frequently. That's wasted time, effort, and money for companies. As return volumes hit record highs, that model just doesn't make sense for some businesses.

Streamy

Quibi gets a new lease on life (and Roku gets a great, ad-tastic deal)

Gone in six Quibis (and back in three)... In October, Quibi announced it was shutting down just six months after launching. The short video streamer had raised a ginormous $1.75B to bring star-studded original shows to your phone (Liam Hemsworth, Idris Elba, Chrissy Teigen — all Quibi'd). The hype was real, but the cruel reality was realer (it flopped). While Quibi is dead, its content is being reincarnated... in the body of a streaming dongle.

If content is king... then Roku just got a royal deal. The popular streaming device company has acquired the rights to 75+ Quibi shows, to be streamed exclusively on its free, ad-supported "Roku Channel." Sure, Quibi had some questionable content (see: show about a woman with a golden arm, show about a sex doll come to life). But it splurged on HBO-level production quality and A-list stars, so Roku's getting a sweet deal:

  • $1.1B: How much Quibi spent on content in the first year, topping out at $100K per minute for the priciest shows.
  • Less than $100M: How much Roku reportedly paid for all that Quibi content.
  • For reference: Netflix spent more than $17B on original content in 2020 alone (you could probably tell from the Bridgerton set design).

Roku is thriving on both sides of streaming... As "the People's Streaming Device" (aka: cheaper alternative to Apple TV), Roku is the leader in streaming hardware, covering over 50M households. That's likely why its stock has nearly 3X'd in value over the past six pandemic months. But more than 70% of its sales actually come from ad sales and subscriptions — and those are growing faster than hardware sales. By beefing up content on The Roku Channel — which has 40K+ movies and shows — Roku can make more ad money and sharpen its double-edge over streamers like Netflix.

What else we’re Snackin’

  • Sleek: Chinese EV-maker Nio unveiled a luxury sedan to compete with Tesla's Model S.
  • Boot: Parler, a conservative-friendly “free speech” app, went offline after Amazon booted it off its AWS cloud.
  • Granny: IAC’s ANGI Homeservices just bought Umbrella, a company that connects senior citizens with $20/hour “neighbors.”
  • Suspend: Marriott and Blue Cross suspend donations to US lawmakers who voted against Biden's certification.
  • Clogged: Crocs stock jumps on the CEO's forecast that the ugly shoe company will deliver the "strongest sales in Crocs history" for full-year 2020.
  • Crypto: Bitcoin pulled back from its epic rally, falling as low as $30K.

Tuesday

Authors of this Snacks own shares of: Amazon, Walmart, and Shopify

ID: 1475186

Get Your News

Subscribe and thrive

Snacks provides fresh takes on the financial news you need to start your day. Chartr provides data visualizations on business, entertainment, and society. This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

Latest Stories

Go Deeper with Market Depth

Nasdaq TotalView powers the need-to-know data serious investors rely on.

Scuba Diving in the Wild Blue Yonder in French Polynesia
Power

World out of balance: It costs the US 3¢ to make 1 penny

The cost of producing the US penny rose 13% in fiscal 2023 to 3.07 cents. Yes, that means that Uncle Sam loses more than two cents for every cent it produces. (And no, you can’t make it up on volume.)

For the record, that’s the 18th-straight year the penny’s face value has been below production costs, fueling calls for abolishing the lowest value denomination coin. Canada started to phase out the penny in 2013, joining Australia, Brazil, Finland, New Zealand, Norway, and Israel, according to Smithsonian Magazine.

3.07¢

Your inbox is ready

Subscribe and thrive

Snacks provides fresh takes on the financial news you need to start your day. Chartr provides data visualizations on business, entertainment, and society. This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

Business

Netflix is going to stop sharing subscriber numbers

After posting subscriber numbers that beat expectations today, Netflix says it’s no longer going to share those numbers starting in the first quarter of 2025. That’s a big deal since subscriber numbers have long been one of the main metrics that investors have looked at.

“In our early days, when we had little revenue or profit, membership growth was a strong indicator of our future potential,” its shareholders letter read. “But now we’re generating very substantial profit and free cash flow.” The company said that it will focus on revenue and operating margin as its main financial metrics, while it will look at time spent on the platform to gauge customer satisfaction.

Another way to read this? They’ve hit market saturation and just aren’t going to be growing that much anymore, and they thought they’d end on a good note. Going forward they’re focusing on how to get more money out of the customers they do have.

They’re doing so by cracking down on password sharing and charging for extra members. They’re also pushing people to ad tiers, which are more profitable than non-ad tiers.

“Scaling ads to become a more meaningful contributor to our business in ‘25 and beyond,” Netflix said.

Netflix’s ads membership grew another 65% in Q1 over the previous one, after rising 70% the quarter before, and 40% of signups in ad markets continue to be for those ad plans.

Tech

Meta’s not telling where it got its AI training data

Today Meta unleashed its ChatGPT competitor, Meta AI, across its apps and as a standalone. The company boasts that it is running on its latest, greatest AI model, Llama 3, which was trained on “data of the highest quality”! A dataset seven times larger than Llama2! And includes 4 times more code!

What is that training data? There the company is less loquacious.

Meta said the 15 trillion tokens on which its trained came from “publicly available sources.” Which sources? Meta told The Verge’s Alex Heath that it didn’t include Meta user data, but didn’t give much more in the way of specifics.

It did mention that it includes AI-generated data, or synthetic data: “we used Llama 2 to generate the training data for the text-quality classifiers that are powering Llama 3.” There are plenty of known issues with synthetic or AI-created data, foremost of which is that it can exacerbate existing issues with AI, because it’s liable to spit out a more concentrated version of any garbage it is ingesting.

AI companies are turning to such data because there’s not enough good, public data on the entire internet to train their increasingly greedy AI models. (Meta had reportedly floated buying a publisher like Simon & Schuster to satisfy its insatiable data needs.)

Meta, of course, isn’t the only company that’s tight-lipped about where its AI data is coming from. In a now infamous interview with WSJ’s Johanna Stern, OpenAI’s chief technology officer Mira Murati was unable to answer questions about what Sora, OpenAI’s video generating app, was trained on. YouTube? Facebook? Instagram — she said she wasn’t sure.

What is that training data? There the company is less loquacious.

Meta said the 15 trillion tokens on which its trained came from “publicly available sources.” Which sources? Meta told The Verge’s Alex Heath that it didn’t include Meta user data, but didn’t give much more in the way of specifics.

It did mention that it includes AI-generated data, or synthetic data: “we used Llama 2 to generate the training data for the text-quality classifiers that are powering Llama 3.” There are plenty of known issues with synthetic or AI-created data, foremost of which is that it can exacerbate existing issues with AI, because it’s liable to spit out a more concentrated version of any garbage it is ingesting.

AI companies are turning to such data because there’s not enough good, public data on the entire internet to train their increasingly greedy AI models. (Meta had reportedly floated buying a publisher like Simon & Schuster to satisfy its insatiable data needs.)

Meta, of course, isn’t the only company that’s tight-lipped about where its AI data is coming from. In a now infamous interview with WSJ’s Johanna Stern, OpenAI’s chief technology officer Mira Murati was unable to answer questions about what Sora, OpenAI’s video generating app, was trained on. YouTube? Facebook? Instagram — she said she wasn’t sure.

Today’s earnings: Who’s making money edition

Here are some some notable numbers out this morning, as earnings season gathers steam. Thursday’s main event will be Netflix after the close of trading. (Keep an eye on its advertising business.) But until then...

7.13%

The 30-year fixed rate mortgage is back above 7%, according to weekly numbers from the Mortgage Bankers Association, the highest level in four months. High borrowing costs are creating havoc for would-be buyers, as affordability lingers at the low levels not seen consistently since the late 1980s.

Business

Amazon’s spy ops on rivals: shell companies, printed docs, and a fake Japanese streetwear brand

Some companies check out rivals’ websites, stores and public filings to stay abreast of the competition. Amazon made its own fake shell company and brands, transacted hundreds of thousands of dollars per year undercover on competitors’ platforms, and kept its intel operation a secret for nearly a decade even from others at Amazon, according to a fascinating investigation by the Wall Street Journal.

Working as a seller called Big River, a secret group of Amazon employees gained access to rival platforms, including Walmart, FedEx, and Alibaba. They used Big River email addresses and went to seller conferences as Big River employees. They even stayed hidden within Amazon itself. These employees would take screenshots of competitors’ systems that they would then show others at Amazon in person to avoid an email paper trail.

Perhaps most strange of all, the company created a fake Japanese streetwear brand called “Not So Ape” (clearly a play on A Bathing Ape) and continues to sell products from the brand on a Shopify store, presumably as an attempt to learn the inner workings of the shopping platform. Of course, copying is old hat for Amazon.

In meetings where they’d use this clandestine information to inform Amazon’s own business practices, the group resorted to literal paper. “[T]he team avoided distributing presentations electronically to Amazon executives. Instead, they printed the presentations and numbered the documents. Executives could look at the reports and take notes, but at the end of the meeting, team members collected the papers to ensure that they had all copies."

Working as a seller called Big River, a secret group of Amazon employees gained access to rival platforms, including Walmart, FedEx, and Alibaba. They used Big River email addresses and went to seller conferences as Big River employees. They even stayed hidden within Amazon itself. These employees would take screenshots of competitors’ systems that they would then show others at Amazon in person to avoid an email paper trail.

Perhaps most strange of all, the company created a fake Japanese streetwear brand called “Not So Ape” (clearly a play on A Bathing Ape) and continues to sell products from the brand on a Shopify store, presumably as an attempt to learn the inner workings of the shopping platform. Of course, copying is old hat for Amazon.

In meetings where they’d use this clandestine information to inform Amazon’s own business practices, the group resorted to literal paper. “[T]he team avoided distributing presentations electronically to Amazon executives. Instead, they printed the presentations and numbered the documents. Executives could look at the reports and take notes, but at the end of the meeting, team members collected the papers to ensure that they had all copies."

Crypto
Jack Morse
4/17/24

Worldcoin pivots to the blockchain… with a 'humans only' discount

Worldcoin, the “proof of personhood” crypto project launched by OpenAI’s Sam Altman, said it plans to launch its own ethereum layer-2 (L2) blockchain dubbed World Chain. The pitch: a blockchain where it’s both easier and cheaper for people to transact than bots.

Worldcoin has made waves for its iris-scanning metallic orb that promises a future where people can mathematically prove they’re real humans and not AI bots.

But it’s run into trouble: the orbs have been banned across Europe and Africa, and the associated WLD crypto token has plunged 50% over the past month.

For project insiders, who reportedly received a token allocation of 25% of supply, that could equal significant losses. 

Which is what may make World Chain attractive. Crypto exchange Coinbase launched its own L2, Base, last year. Base has since seen rapid user growth — activity that’s generated the exchange millions of dollars in weekly fees

Worldcoin could benefit from similar revenue if its L2 is adopted around the world.

But it’s run into trouble: the orbs have been banned across Europe and Africa, and the associated WLD crypto token has plunged 50% over the past month.

For project insiders, who reportedly received a token allocation of 25% of supply, that could equal significant losses. 

Which is what may make World Chain attractive. Crypto exchange Coinbase launched its own L2, Base, last year. Base has since seen rapid user growth — activity that’s generated the exchange millions of dollars in weekly fees

Worldcoin could benefit from similar revenue if its L2 is adopted around the world.