Does Amazon make more from ads than AWS? | Hacker News Reader

Does Amazon make more from ads than AWS?

ben-evans.com

334 pointslxm133 comments

This is a good read. For a long time I explained to people that the value of search advertising was that it matched a buyer who had just asked a buying question with advertisers who wanted to answer that question. This was HUGELY different than previous Internet "banner" advertising. Think of it this way; I put a poster up in the metro that says "Best Deals on Rolex watches" and a million people a day "see it" but a vanishingly small number want to buy a Rolex. However, if Google gets a query "Where to buy a Rolex" or "best deals on Rolex watches", well you know that person is perhaps already considering buying a watch, so you want first crack at the deal.

The secret sauce was instantly pairing "commercial intent" (or the intention to do some commercial action, buy, sell, hire, Etc.) with vendors and firms who can satisfy that intent or "aspects" of that intent. (You might get an Uber ad on a query for "restaurants near me" for example)

Amazon has that in spades. If you are on the Amazon site, that is a huge signal that you are intending to make a commercial transaction. Either with Amazon, or perhaps you are reading reviews of things you can get locally, but the probability that you are about to do something commercial is much much higher. (So does Facebook along a different vector but that is a different post for a different time.)

I know from experience at Blekko that one way to piss off Amazon is to out advertise them :-). That just told me that they considered Ads strategic and selling stuff was secondary to that goal[1].

[1] I reasoned it this way, if selling stuff was primary then any additional exposure is "good" even if you have to pay affiliate fees for it. If advertising was the money maker, then someone doing it better than you (converting your ad revenue into their affiliate revenue) then you respond harshly to that.

  However, if Google gets a query "Where to buy a Rolex" or "best deals on Rolex watches", well you know that person is perhaps already considering buying a watch, so you want first crack at the deal.

I've found that my "buying questions" are never properly answered by those ads. The more commercial intent I display, the scummier the ads get.

I blame misaligned interests: I want to know if the product is good and to find good deals, they want me to give them money, now! And since advertising is a numbers game, they don't pull their punches when manipulating me.

No, thank you. Talk to my adblocker.

I get where you're coming from. As someone who has had ads be "basically" what I want, I've never really believed that someone wanting to take my money for a thing I might want is _that_ misaligned in terms of interest (kind of like recruiters or real estate agents wanting you to take basically anything: they might not have your best interests at heart but they do want to close a deal). This is, of course, a naive view of the world

A sort of value-neutral narrative I've come up with recently: the ad -> sale pipeline is basically super flexible (you put in X money to get Y clicks for Z sales, but don't really need much staff on it). So if you're super good at this kind of optimization you can confidently put in the maximum about of X money to get even just some profit. And of course outbid people who are less confident.

So in the auction-based ad market, the people who will "win" the ad space will be people _who are good at auction-based ads_! Everyone who really cares about sales funnels etc will be able to get better results!

Of course there's like.... misleading copy etc, but I'm more and more convinced that the reason for the awfulness is that the people who get the slot are getting it for a mostly unrelated skill.

> I've found that my "buying questions" are never properly answered by those ads. The more commercial intent I display, the scummier the ads get.

It’s a lot like walking into a car dealership. They aren’t aiming to match you with the best car for you, they’re aiming to sell you the car that makes them the most money.

Modern life is like being in a car dealership 24/7.

That is a sadly apt metaphor

I've never actually bought straight from a dealership so excuse my naivety here, but wouldn't they be better off trying to sell me the car that is best for me? Otherwise if it doesn't meet my needs, I won't be buying anything.

I recommend talking to some people who know about car dealer sales strategies. For many people buying a car is highly emotional and car dealers are trying to manipulate those. Some examples, if you come in with a budget of say $5000 they will always try to sell you something which is above your budget ("this is what you really want/need"), so never give them your true budget. If it comes to closing a sales deal they will often work on you with 2 people, which is much harder to resist psychologically. So best is to go buy a car with at least 2 people as well (and make a strategy before so they don't play you against each other). They also use strategies, like installing a mirror at the place where you drop the car back after a test drive so you see yourself in the car, which causes an emotional response.

...and this is part of the success of Tesla.

Car dealers are literally driving traffic away from them. They are about as popular as taxi drivers before Uber.

Psychological tricks are a predatory social behavior, not only to the buyer, but also because people (especially introverts) detect them quickly and grow tired very fast. That leads people online (for Tesla) or online shopping (who enjoys being harassed by vendors in every clothes shop?), or online because at least the price negotiation doesn’t depend on your face or your mastery of social games. When your psychological tricks lead to entire industries sidestepping you and letting you die with popular support, you know something is wrong.

And yet Saturn failed even though that was how their cars were sold also.

All true. And thankfully avoidable in 2021. I just bought a new car, from a large chain dealer, 95% by SMS.

Bought my last car this way - once I knew what make/model I want (which did require physically seeing a few), it was 100% email until it was ready on the parking lot to drive off and the papers were ready to sign. The dealers that asked to "come and talk", or "call and we'll discuss the options" just got "thanks but no thanks". Fortunately, in 21th century there are people that are willing to sell cars without bullshit, and they got the deal. And that was pre-lockdowns, now I imagine more people would be willing to work this way.

No thanks, delivery would eat all my data allowance.

What is 'best' is highly subjective and emotive.

At almost any price point there is going to be some model or feature that is a bit more expensive that you may not strictly need but that may be better in one way or other.

You probably can't tell which car is best for you, and salespeople aren't gonna help.

OTT but in line with your question In my part of Australia we have an established dealership who has been here since the 70's and I bought my 1st car from him in the mid 90's. each couple of years I would go back and ask what he could do for me as a trade and on what. every time, walking out with a good deal (he actually moved figures around to get the best price) larger part of my wifes family started going there and same outcome, come home with a car they like, met their expectations and got good trade-ins for their cars. recently I took mine and my wifes cars in to change over, we picked our cars, got fair to good trade ins on them, paid and then waited on them to be detailed so we could drive home. I asked the dealer principle which car would he have sold me if he wanted to make money. he said, plainly "they all make money, in differing amounts, no point trying for a high margin car that won't suit either of your needs" I asked what he could of pointed us at if he wanted more money out of us. his reply "I'd push you to a cheaper 2nd hand car or one of the LDV/MG/Import cars where I would make 3x as much as what I am getting from this sale" guy is in his 80's and doesn't mince words but is as honest as they come

It seems that it's common understanding in US that car dealer is BS. Is it cultural? because I don't see such common understanding outside US.

They will sell you the car currently in their lot that most closely matches what you need. Not what you actually need. Just the closest they have.

Doubt. Yes, the one in their lot; no, on best match. They'll try and sell to you according to their need - highest commission, dump a lemon, etc. If you do choose a vehicle they'll try to give you a terrible finance deal, wiggle out of a warranty they were offering, and such.

They just have to convince you it's good enough.

Even if they do pick a car largely based on your needs, they're never going to say it's anything but a perfect or near-perfect fit.

Car dealers may not be a frequent repeat customer like stores Word of mouth is diluted by type model needs, it’s not my friend was recommended dealer bar for their experience buying a Honda foo I’ll go and buy one

My experiences exactly, my ad experiences fit in to two categories “things i just purchased” and “terrifically bad suggestions”. A third category might be “creepy connections” where a random search leads to ads on vastly different platforms usually quite inappropriately.

Amazon are terrifically bad at the “things I have already purchased” advertising. They _know_ with 100% certainty that I bought a webcam 6 months back - because I bought it on Amazon. Yet they’re still sending me email and putting ads on webpages for webcams, including the exact model I bought from them...

(I totally expect some shitty AdTech startup to start buying remarketing ads for NFTs any minute now. )

Maybe the victim in this case is not you, but the company paying for the ads.

Think about it this way: You buy a webcam from some company X. The company X wants to sell more webcams, so they buy ads. Amazon shows those ads to you. At the end of the year, company X gets a report saying "we displayed your ads to bigiain, and bigiain bought your webcam" (the little detail that you bought the webcam before seeing the ads is not mentioned). Would this seem to the company X as money well spent?

Have you considered that Amazon is very good at what they do, and that perhaps people who buy a webcam are much more likely than then general public to buy another one?

Considering that the same thing happened to me when I last bought a movie on DVD from them and then got ads for that very same movie for weeks, no.

> Have you considered that Amazon is very good at what they do

Cynical-me considers Amazon to be the least-worst of all the companies trying to do what they do.

I'm sure their "re-advertise stuff people bought before" thing works great for some of what they sell - toilet paper or laundry detergent or cat food. So just like every "data driven" VC backed startup, they've "validated the business model" and are now pumping advertising dollars into "scaling" their proven toilet paper sales pipeline to webcams. They're probably busy running multivariate testing to find if the optimal repurchase advertising repeat rate for webcams is down near 4.3 weeks like laundry detergent, or way up at 9.1 weeks like toilet paper. 'cause the data doesn't lie, right? :sigh:

You'd be surprised how often the people who are supposed to be good at ads aren't. Organizational inefficiency explains a lot of it: https://thecorrespondent.com/100/the-new-dot-com-bubble-is-h...

That may happen in many companies, but I'd be very surprised if it happens with remarketing ads and Amazon? Performance-based advertising with solid experiments that appropriately consider both counterfactuals and long-term user value is one of their core strengths.

Are the "things I just purchased" suggestions actually bad though? I see people complain about them a lot, but I haven't seen anyone chime in with what the conversion rate is on them compared to other targeting strategies. Like I'd suspect that having already purchased an item is highly, highly correlated with purchasing it again. And that probably even applies to things that don't initially seem like you'd need more than one of.

When I have my ad-blocker off I see a lot of ads for Linode, AWS and DO. I already have a relationship with each of them.

Amazon still keeps showing me faucet valves and a kitchen-sinks. (I did a replacement two years ago).

And, every last ad for any/all PostgreSQL commercial add-on.

Some (many) items don't fit the repeat pattern and I'm confounded how, with all the Ai/ML crap this isn't known yet.

They (ad-sellers) would rather hustle the BS of "retargeting"

Or it fits better than you think, and/or the margins for certain types of products are high enough to justify bidding for ads even when they fit only a tiny proportion of the time.

E.g. you may have a relationship with Linode, AWS and DO, but presumably all three would prefer you ditched one or two of the others for more spend with them.

And you may have done a kitchen replacement two years ago, but how many percent need to do repairs or changes subsequently, or are unhappy about choices they've made, and end up spending enough on high margin additional products for years to come? That you've proven yourself to be willing to invest in improvements recently might well make you more profitable to advertise to even though you've already made the biggest investment.

I'm sure many advertisers are bad at targeting, and so I'm sure some of these ads are metaphorically burning money. But I'm also sure - because I've seen that too - that sometimes seemingly totally counter-intuitive ads makes sense after all with access to the actual underlying numbers.

My impressions about ads is they are usually not interested in selling you what you actually need, but they try more to get you to feel like you should buy things you don't really need. I suspect it is because it's easier and much more profitable (we would spend less money if you only bought things we want need).

Having already purchased an item is indeed highly correlated with purchasing it again, however, you are ignoring the temporal element. Yes, I will need to buy a new gaming monitor for my son, but hopefully not for another three years, and by then most of these ads will have disappeared.

But for all they know perhaps you want to purchase one for your daughter now or a second one for a dual-monitor setup. Ads in general have a very low success rate, so it can be worth doing even if the guess is wrong a very large percent of the time

Well, the most common ad of this type are for the exact item I purchased from the retailer I purchased it from. Usually for things which a person wouldn't need two of. I tend not to see these ads until after I have made the purchase.

Right? If anything that tells me what vendors spend more on ads than on their product; every dollar they spent on this annoyingly pervasive ad campaign was a dollar they took from buyers but didn't put into the products they're selling.

Yeah, it is annoying when I am searching for a specific product and the top ad results are places that don’t actually sell it. It is one thing to advertise to me the thing I want, but don’t distract me from what I am trying to do.

See, I have the opposite problem in that I'm too impulsive. I search for a mattress for 20 mins, then buy one.

The problem is that I then get mattress ads everywhere I click for the next 3 months. It's like I need a button to mark my query as solved.

A theory as to why the ads are so useless to people despite the above:

Who has the most to gain from being shown in these advertising spots? Those who would not be purchased otherwise. If you are the default choice, maybe some amount of people will be swayed by the competitors ad, but % wise, not a huge amount.

Who wouldn't be purchased otherwise? Those without recognition otherwise. This could be those who are simply too small to be known, or those who are "just another no name brand" with possibly questionable quality.

Of these two groups, it's the latter who will have more money to spend - the "small" sellers by definition don't have the marketing budget. So it's those selling crap in volume who have both the incentive and the resources to put the most into these ads, and why they end up showing up disproportionately among the advertisments.

It's also worth mentioning that "a buyer who just asked a buying question" can be almost entirely met through contextual search (no need to track the user, just the question ala duckduckgo) - and so you can let go of an awful lot of previously held behavioural data (tracking people across multiple sites) - partly for privacy but also I have not seen any evidence it adds (significant) value

This is true. The behavioral data is valuable to the bidding process. If you can provide context that this user previously bought things after clicking through, their demographic with respect to your target market, etc. Can be used by advertisers to change their "bid." (increase it generally) So providing behavioral data improves the CPC which improves advertising margins for the publisher. That is what drives harvesting all of that data.

DDG could increase their ad margins buy supplying identifying data, but perhaps they choose not to with the assumption that higher volumes will result in more gross revenue. I've never seen the books at DDG so I can only speculate.

The most prominent ads I see on Amazon are for the thing I just bought, that isn't a kind of thing most people buy repeatedly.

Well, you might think that. Did you consider that you might be wrong and maybe Amazon actually has lot of data that you don't and they know it better?

Just a couple of examples.

I might have bought a hard disk drive a couple of days ago and realize that I actually want more so I can do backups (or I want to build a cute HDD tower with a Raspberry Pi).

I might have bought ear plugs for home, and decide a week later that it would be nice to have those at work, too.

I could have bought zip ties, only to realize that I actually want more in different sizes and colors.

I bought a pijama, it was delivered and I want to buy 3 more of those, so I can throw out my old ugly pijamas and not worry about buying new pijamas for the next 5 years.

I got headphones but they aren't quite perfect. I didn't issue a refund yet, but I know I'd eventually send them back and want to buy a similar headphone asap.

Except the same-damn-product ads also come up for things that make very little sense for re-purchase: artwork, DVDs, books, and so on.

I can envision scenarios where you're buying multiple copies of these things - gifts, book clubs, etc.

Advertising doesn't have to work universally - it just has to work better than not advertising at all. I could very well imagine situations where it is cheap enough to advertise to everyone who has purchased those same items to get additional sales from those duplicate purchases and make money after the advertising spend.

No, I never, ever, consider that I might be wrong about anything. Doesn't happen.

Where do you see those ads?

The most prominent ads I see on amazon is when I search for literally anything, the top results are often whoever paid Amazon the most to be the top results. Sometimes I will literally search for a brand by name and some other brand will be shown before it. This kind of race to the bottom is something I've heard a lot of companies complain about, companies literally need to pay Amazon for the ads on their own terms to outbid competitors.

My guess is that they make sure that you don't doubt your purchase. You could send it back if a competitor gets into your head. Additionally, if you don't doubt, you become loyal to the brand and you will recommend it in your network.

The pricing for the click might kill off any benefit over less targeted advertising though. I poked around in keyword planner, and one click for some rolex related phrases costs $3-$10. I imagine you would have to do quite a lot of expensive experimentation to figure out how to make that profitable.

Rolexes sell for five figures. Even if someone has to spend $x,xxx on clicks before a single conversion, they are still making a killer ROI.

One of my favorite anecdotes from my time in comparison shopping (where north of 80% of our revenue was CPA deals) was looking at the variation of margins by product (electronics paid 1%; clothing more like 8-10% as a rev share) and seeing one outlier where we were paid a commission of 50%.

The product was a fake rolex, as in, the name of the item was literally something like "Fake Rolex", and while it was only something like $100, instead of $10,000, it was apparently still such a high margin that the seller would pay us a 50% sales commission.

Sorry what are cpa deals

Cost per acquisition

Versus CPC, where you get paid for anyone clicking through, or CPM, where you get paid per thousand views of the "ad".

The question is how many people search for Rolexes vs how many people actually end up buying one now (or at the very least, later but from the same seller, to offset the ad's cost)?

I think that for luxury goods, you can have a lot of people (more than for other expensive but otherwise boring goods) looking at them (maybe they are curious about the price) but very few that would actually follow through with their purchase.

Clicks show a lot more buying intent than impressions though. Might still be OK.

I think the great problem is "commercial intent". I search the internet multiple times a day, but I rarely want to actually buy something.

Amazon has a win here compared to facebook or Google because if I search for something on Amazon it's vastly more likely I am buying.

It's the difference between asking a sales assistant "do you have good coffee" (amazon), asking a librarian for a book on good coffee (google) and asking my friend if they like their coffee (facebook). Intent and context matter.

https://news.ycombinator.com/item?id=26535260

That, or I'm shopping for someone else to recommend it to them.

Where Amazon's got a major problem is that the counterfeits and the perception that any negative product experience might be the result of that, makes it increasingly less likely that I and others, go to Amazon looking for that product.

I have been thinking about why Amazon is pissing in its own soup, with counterfeit and other "problematic" goods.

1. Scale - it's just too hard to curate all those products. Meh - Amazon has hired like a billion people, it can hire product line owners.

2. Do you remember in the mid-90s Amazon had some big push for sellers to add their SKUs abs descriptions of goods (ie instead of Amazon holding a database of goods, and you saying "i am selling one of these" you the seller added it to amazon marketplace. This was a move to collect all the data about all the goods ever.

The current allowance of poor quality goods is I think the same game. Most dodgy goods sold are not "dangerous counterfeit" - as in selling medicinal drugs made from floor sweepings. Most dodgy goods are the things you find in Poundland and market stalls. A plastic knock off of "Mike-y Mouse" or running tops made from cotton not the high tech wick, or often the same goods as branded made on same machines, just not branded.

This set of goods is what the next billion consumers will buy for two decades. You cannot afford H&M prices but you can afford the knock offs pouring out of Chinese and Philipino factories - and yet how do you get the catalog for those factories?

The same way Amazon did in the 90s.

Anyway ..

I think we need to look at the impact of the recent Apple ID and chrome / firefox changes.

If it becomes harder for a the ad network ecosystem to track a person across sites, then sites that provide commercial intent (amazon) or provide first party access to self selecting groups (ie car magazines, Wall Street Journal articles on cars)

If apple can stop ad networks selling "people who read the FT" at half the price that FT charges, then mainstream media might be able to claw back what Google and facebook took.

"Amazon has a win here compared to facebook or amazon"

I meant google - corrected

The basic idea is in the GP - amazon is a shopping site so any search there carries a lot of commercial intent. This fades (linearly? exponentially?) as we walk down to google then facebook.

Would love to hear your thoughts on Facebook and how that’s similar / different

Rolex doesn't need to advertise on public spaces. They will just sponsor an event that attracts the wealthy, like a sailing race or a horse race. Contextual advertising is very powerful, and it's the one thing digital networks decided to abandon because it's way easier to automate everything and spy on users.

Part of the issue with this logic is that ads for things people were already going to buy are going to preform like gangbusters according to metrics, and will be having next to zero actual impact on future sales.

Not future sales, the choice, right now, that the buyer makes.

I suspect it is most effective when you have a "product" which several sources in China make under what ever trade name you want. (There seems to be lots of electronic test equipment like that) And your typical FBA business. Advertise you gear when the right keywords come up and get a click ahead of the guy selling it at the cheapest price.

If all the products look the same and Vendor B advertises to put it at the front and pays a service for a few hundred 5 star reviews, I would not be surprised if it added 10 to 15% to the net margin on sales.

> I know from experience at Blekko that one way to piss off Amazon is to out advertise them

It is a shame (at least to me) that Blekko was acquired by IBM - it showed a lot of promise and I was a fan at the time.

>then you respond harshly to that

How harshly? I've got the popcorn ready...

>So does Facebook along a different vector but that is a different post for a different time.

That was six hours ago. Now is a different time...

Google and Facebook are going to get screwed. They'll get hit by antitrust action (even if just more fines and settlements) while also under the radar losing market share to Amazon via market forces, which ironically is an argument against taking antitrust action against the formers.

Amazon makes more revenue from ads than Youtube does (and Youtube makes 100x what Reddit does). That's nuts.

Chart of each ad/marketing channel by spend for context: https://www.rightpercent.com/b2b-guides/which-marketing-chan...

When I am already on a marketplace trying to buy something, the intent value and timing of it makes it infinitely worth more than random videos, at best re-targetted because i previously visited something.

I understand this is probably against HN[PolicyNo], but I just came across your comment in another thread that is locked from being ?too old?. It was the Google FloC Post where you stated you wouldn't work on gambling or alcohol ads. This is an area that I recently had to personally struggle with my own personal morals. I wasn't working in ads specifically but marketing. I guess I just wondered if you would be willing to share any thoughts you have on how you find that line of what is acceptable regarding feeling ok to work on something or walk away. Specifically my dilemma is less that about caring about what someone's vice might be (it's not up to me to moralize or decide for people what is 'good'/'bad') but more so that in my opinion we as a humans can be exploited through *technically sophisticated* approaches no matter how much we each think we are immune to that kind of influence; so under that premise, how do I know when some proposal is evil or at least I won't feel good about it when delivered.

Makes sense given Amazon visitors are already pre-disposed to purchase something while the vast majority of YT viewers aren't.

RightPercent: B2B AGENCY FOR FACEBOOK AND GOOGLE

I'm not sure what data that shows, but I'm confident it's not showing anything but the "public data"...which is a simple excel spreadsheet, hosted on Google by "someone".

If you search for the doc id (16HWY3ytQY0kMEFCoJKAwkjO2dr6x0-Nm146-taLpf7A) it's a couple RightPercent posts...probably because they pulled it out of their ass for their own sales pitch.

I worked in the ads org at Amazon an year ago and yes I can tell you the revenue generated by ad products across their services is insane. It's a well executed operation and it is only growing rapidly year over year. From a customer perspective they are also very relevant most of the time.

The remaining part is making sure they don't upset customers way too much since that would conflict with the idea of "customer satisfaction" the company has.

> From a customer perspective they are also very relevant most of the time.

I wonder just how that is measured, and how much of a bullshit metric only calculated and reported on to get someone promoted?

I’ve managed to get myself into the “buys webcams” demographic, but buying a webcam. I now suspect all the pointless webcam ads Amazon send me are all highly ranked for “relevance”, even though it’ll probably be 5 years before I buy another one...

Doesn't matter if it's going to be 5 years, they still know what you might be interested in. To a lot of people it is much shorter than 5 years.

It's sad how "We aim to be Earth’s most customer centric company." turns into "not conflicting too much" real quickly when there's money on the table.

Surely "we aim to be Earth’s most customer centric company" is still a means to the end which is money, not some altruistic goal?

The reason why webshops got big isn't because they have better customer service but because they are cheap. With online shopping it always ends in a race to the bottom.

I've always been stunned by how irrelevant the ads were on my Kindle, until I became so annoyed I paid the fee to turn them off. Ads for Kindle ebook titles I had already purchased, ads for household cleaning products, ads for books in genres I have never read (romance), etc. My spouse still gets them on their Kindle so I can see things have not improved one iota.

I actually found it a bit offensive, as in "how stupid do they think I am to be tempted to buy a book I already bought from Amazon", mixed with being puzzled and confused over why a company like Amazon would do such a disservice to its shareholders by doing advertising so very poorly and leaving money on the table.

Can you put some numbers around "rapidly"?

> This week the FT reported the EU is having trouble building a competition case against Amazon: its theory is that Amazon unfairly steers customers away from Marketplace vendors towards its own products, but one of Amazon’s arguments is apparently that Marketplace is more profitable, so its incentive is the opposite!

It's the Alibaba'ification of Amazon. The ideal - if you're Amazon - is to end up as a floating brand that owns a giant billboard mall platform. You want to leave the sellers to deal with most of the problem of not having any margin, as is typical in retail; with Amazon then having ~70% margins on advertising on their platform. And if you're going to bother with traditional retailing, push upstream to owned-brands, which gives you better margins, which is what their Amazon brand build-out is all about.

I'm surprised Shopify hasn't (seemingly) figured out this is the direction they have to go, full stop, if they want to keep that comical market cap from collapsing whenever sanity returns to the stock market. Scraping by on very weak margins a few pennies per transaction along with modest subscription fees isn't going to cut it (their operating income margin on trailing four quarters of $2.9b in sales was a horrific 3%). I'd bet on Shopify trying to aggressively leverage their position in tandem with advertising to try to create a fountain of cash. When they inevitably do it, turn to the dark side, it'll gradually make Shopify into a terrible platform.

So the counterargument would be that shopify realizes it will become a horrible platform if it moves towards the 'higher margin markets'. And therefore doesn't.

Maybe they are looking at the long term. I doubt it, and even if they do so today, tomorrows VC or hedgefund may change this. Still. It is not unthinkable.

I think it's definitely a nice thought, that they'd choose that path, even as their growth slows and shareholders get irate as their market cap implodes (it's impossible to support a $129 billion market cap with their existing business model, they'll be lucky to support 1/3 or 1/4 that). Canada always has an outsized bubble stock with each great bubble; for the dotcom bubble it was Nortel; for the real-estate bubble era it was BlackBerry/RIM; for this one, it's Shopify.

My brain wants to think: hey, the Canadian market might be different, they'll act different. And that sounds great, except by the time this is a critical issue (by the time their current business model runs out of its high growth), there will be immense outside influence over the company and the inside owners and their control will have eroded. As that market cap plunges toward a more realistic fair value and growth slows, the calls for change will be extremely loud. They probably have no choice in the end, they'll need to maximize their position at the center of all of those stores, as the platform so many stores depend on, and that will mean forcing an ad network at the center of it all that they try to milk for margin.

And in terms of supporting that silly bubble market cap, when Amazon was worth just a bit more than Shopify is today, back in mid 2014 (when interest rates were sitting at 0% also), they had ~$80 billion in sales, 27 times larger than Shopify is today. Shopify is either going to crash, go sideways for many years, or they better find billions in profit asap.

Interviewed with Amazon's ad product org a few months ago (ultimately both decided it didn't make sense at the time.)

Amazon is crushing it with ads and will do even better over time. The main value of ads is to steer shopping dollars towards a better product for the customer. EG: I was about to buy product X but an ad taught me that product Y is better (or just as good, but cheaper.)

Amazon is already the best place for this kind of advertising because people are on the site 100% of the time in shopping (or at least product research) mode. Such ads are WAY more valuable than showing me ads when I am trying to read the news or some other non-shopping activity (although amazon is in that business as well.)

Over time, Amazon is becoming more and more the start of people's product search. That means Google neither learns about the customer's interest nor earn the ad revenue from (most likely) sending the shopper to Amazon, which makes Amazon's ads even more valuable in comparison.

I have no idea HOW valuable the business is but - very.

Maybe this is working for the masses but I HATE searching for things to buy on Amazon. Last time I tried, the results pages are overwhelmed with sometimes literally hundreds of the exact same product sold by gibberish seller names. Ratings were obviously useless (the same product will have thousands of 5 star reviews by one seller, 10s of more mixed reviews on others).

Maybe it's better now but I can't remember the last time I tried to find something on Amazon because of this.

The thing that I hate the most about Amazon search is that they return partial matches. So if I search "evaporative humidifier" half of the results don't have the word "evaporative" in them and I am just wasting my time scrolling though and having to carefully read the descriptions.

It seems that they try their hardest never to return an empty page, but this is extremely frustrating to a consumer because I would much rather you just tell me that you don't have something, or only have 10, rather than spreading those 10 matches over 20 pages of junk results.

I guess their thought is that the results may be "close enough" but they aren't! If I am ok with compromising I'll remove some conditions from my search upon seeing an empty result list. (here is an idea: recommend which terms I can remove to find the most results) But they just ignore random terms, even if some terms are negotiable for me they will get me to waste so much time scrolling though the items that don't match the critical terms that I no check other sites first.

Amazon’s search is terrible. If you sort by price, then the top pages of results are almost always complete garbage items that are completely unrelated to what you are looking for, no matter how specific you are. Almost like Amazon don’t want you to find the cheapest items...

So much this. Its actually amazing how it can be this bad. It also is a good example how you don't have to make a good product to rise to the top.

Amazon = eBay + house brands + ad network

It's a marvel that any third parties remain on Amazon.

+ shipping and returns expectations + chance of refunding your small purchase 3X to lie for a seller in your review.

Ya, Amazon Prime was pretty smart. Ok, Bezos did two new things.

It's because Amazon's ad platform has been the best ecommerce ad platform for the narrow and profitable purpose of direct sales basically since the day that it launched. Google let Amazon take a giant bite out of Shopping because it was complacent with its also profitable business model of absorbing giant corporate black hole ad budgets and earning multi-hundred dollar clicks off of people like personal injury lawyers.

I just went through a rabbit hole of sites about adwords for injury lawyers and my mind is blown by how obvious and expected it is but it’s still insane to think of a single click being billed 200$. But it makes perfect sense, adwords is far more suited to for this purpose and it’s one case of everyone getting what they want (except the dude getting sued) so why not? Google is getting healthy competition in the ad space so within that dimension it’s great? If we now just ensure the Amazon workers can unionise and put amazons more physical manifestations on check then this is a better situation overall than whatever we have now.

I run a commercial insurance campaign with clicks from $25 to $45 each but it works and provides the whole company with enough leads the renew the next year to make it worth it but it costs $120k+ a year. Google has figured out how to make the organic list as irrelevant as possible.

Amazon ads are contrary to the company's mission statement of being the "most customer centric company." They could improve the customer experience at one stroke by eliminating pay-for-placement.

It is more complicated than that. Advertisers are customers too and advertising creates incentive to use Amazon and enter into larger deals with Amazon that ultimately lower prices for customers. Additionally, revenue from advertising can go directly into funding supply chain innovations that result in an even better customer experience. Amazon doesn’t typically do anything unless the total downstream impact is shown to be positive; while customers do get annoyed with ads, they appreciate the nice things ad presence provides for them to a greater extent on average.

It's pretty depressing that the worst parts of Amazon (the shady marketplace and ads that don't look like ads) might be the most profitable because it means they won't be reformed or eliminated.

I noticed that almost the entire front page of results now on Amazon are sponsored results.

It’s winner winner take all, all the way down.

Yip they have their hands in all the pies with the exception of bank accounts, though with Amazon gift cards almost treated as currency - they are not far from that already.

Are we talking about ads on the Amazon app/site for products on the Amazon store? Or ads on other sites on the internet for Amazon products?

Ad revenue that Amazon receives as a publisher, which comes from a diverse set of ad products served both on Amazon-owned sites and externally via their DSP.

So basically they have figured out a way to keep a larger part of the pie from e.g. those randomly named PRC sellers?

I may be slow here, but couldn't they just instead have raised the percentage they charge for sales? (Are those "external sales" really that large?)

My guess is that it’s a tax scheme.

They probably run everything through different subsidiaries in different jurisdictions and make the capital intensive retail business as bad as possible, and shift losses or depreciation around to write down any earnings in the US or Europe.

If the commerce division owes a big commission to Amazon Isle of Man, LLC, they “lose money”.

They lure in 3rd parties to the marketplace with relatively low fees before the third parties realize no one sees their product unless they buy ads. By that point amazon has a bundle of their merchandise and a relationship so instead of pulling out they just buy the ads.

That's not really 100% true. I sell a single product on Amazon (USB oscilloscope I designed a couple of years back), and have no discovery issues at all.

It's the top product when you search for "USB Oscilloscope", and on the front page if you just search "Oscilloscope". I don't pay a cent in ads.

That said, I did notice a huge rankings boost when signing up for FBA (as opposed to shipping the goods from overseas).

EspoTek? Because Siglent comes up first in AMZ organic searches for me. Not trying to bust your chops, just interested in this field.

Yeah, EspoTek. Interesting to hear that from you - Siglent appears third for me, and that's using an incognito window too (so it would be safe to assume any personalisation is off?).

Just for a data point, in a private tab, when I search Amazon for "USB Oscilloscope" my first three results in order are: EspoTek, Hantek, and then Siglent. It lists EspoTek as "Amazon's Choice".

Same results as when I'm signed in. I'm in the US.

You come first of the organic results for me on amazon.com (with an "Amazon's choice" label), but is nowhere to be seen on amazon.co.uk

> so it would be safe to assume any personalisation is off?

Depends on whether Amazon is using fingerprinting to target ads

This is basically Amazon getting sellers on its site to pay money to be ranked higher in search results, and since result ranking is so competitive, companies basically are forced to buy out the keywords for their own products/brand name to avoid other competitors outbidding them. Sometimes, searching stuff like "Dyson" or "Roomba" will give me results from cheap knockoffs before the actual brand I'm even looking for.

I've already heard it said that if you have the notion that Amazon is not competing with your business then you've probably just not been following its activities lately.

Just to name a few, Amazon is not competing with: banks, cars (rivian is _an investment_, it's not owned by AMZN), almost all direct manufacturing eg. semiconductors or displays, apartments.com, vacation planning, etc.

They're at least partially in most of those markets.

Banks: Amazon Lending

Cars: Zoox

Semiconductors: Annapurna Labs on the design side. They aren't entering the fab side, but who is anymore?

Vacation planning: They've vaguely entered it a few times, but withdrew. They used to offer travel bookings directly on Amazon in partnership with Expedia. That disappeared, then they made tentative steps towards bookings with Amazon local. That's not to mention how they already own a significant chunk of the retail side of travel.

I think GP’s comment was more of a rule-of-thumb about Amazon’s breadth.

But the fact that a list of business _not_ engaged in is that short, _and_ that a bunch of subsequent comments dispute a number of them, is freaking insane.

I woud not put cars past them. One profitable area that will experience the autonomous driving revolution is logistics.

Almost none of those are on amazons e-commerce platform either

Given

> if you have the notion that Amazon is not competing with your business

you would think OP is talking about industries in general, much like how AMZN is moving into a bunch of other industries like online payment processing (Amazon Pay), grocery (Whole Foods, Amazon GO), etc. They certainly aren't moving into every industry, probably because these industries already are low-margin or don't have a clear path to growth under Amazon disruption.

Your point stands, but they arguably do compete with banks through their credit cards/financing. Not sure how it looks in other countries but here in Japan it’s not insignificant. Right now they do it through a JV or sth with SMBC but surely it’s just a matter of time until they get serious there.

> it doesn’t make money, it sells below cost, it’s subsidised by investors and in particular it’s subsidised by AWS

Who makes this claim? It’s idiotic on its face, and disproven by their filings.

> Who makes this claim?

Matthew Yglesias did in 2013.

https://slate.com/business/2013/01/amazon-q4-profits-fall-45...

> That’s because Amazon, as best I can tell, is a charitable organization being run by elements of the investment community for the benefit of consumers. The shareholders put up the equity, and instead of owning a claim on a steady stream of fat profits, they get a claim on a mighty engine of consumer surplus.

And while it was a good quip it might not have became so famous if Bezos didn't quote him unattributed a few months later in a letter to shareholders. https://slate.com/business/2013/04/amazon-as-corporate-chari...

WalMart steals from suppliers, governments, and employees.

eBay & Alibaba learned to steal from customers (fraud, counterfeits).

Amazon also stole from investors. Like many before and since.

Bezos' single core innovation was somehow persuading Wall Street to go along with the scheme. He never tried to hide or obfuscate the plan.

Amazingly, uniquely, that gamble paid off.

How did Amazon steal from investors?

Whether it's right on wrong, it's not disproven by their filings, at least going back a few years. For a very long time Amazon was running at break even or thereabouts. Additionally, while the argument that Amazon had some very profitable lines of business seemed reasonable, it was never clear that Amazon could just choose the profitable ones and leave the losing ones. Loss leaders is a thing. AWS has subsequently made Amazon look amazing and they are obviously worth an ungodly sum of money, but if it weren't for AWS it would still be unclear how good of a business Amazon really is. Even ads isn't a business they get to be in for free. They couldn't just shut down Amazon's direct lines and be all ads + marketplace.

> Does Amazon make more from ads than AWS?

I didn't need to read the article to know. No, Ads don't make more than AWS.

Adspend has been utterly gutted since the early 00's as analytics have shown how ineffective it is. There's still very large companies that uses a nonsense correlated-purchase-after-seeing-an-ad attribution model. This is fantasy to justify existing jobs and to tell other big companies how amazing they are.

Digital ads are not going to "take off" anytime soon either, as the hyper-concentration of capital makes spends more regular and "chunky" (block of 5mil spend to each target platform from the big boys, to the big boys) which hasn't even kept up with inflation. A bearish stance when factoring in population growth slowing and political pushback/infighting/adblocking, is pretty reasonable.

Eh, you might not want to confess that you haven't read the article before arguing against it, that's pretty bad form.

Why don't you read it and argue against the arguments laid out in the article instead, so the rest of us might learn something instead of reading empty opinions from someone who doesn't even listen to others?

Ofc I read it, but the "article" barely manages to make a point with hyperthin facts (AMZ disclosures). The Benedict Evans chart (Amazon: more profit than AWS?) answers concisely, making the "discussion" rather pointless.

> However, we can make an informed guess. Google’s core business had 2020 operating margins

> To repeat - this is just an informed guess, and ads will of course change other things

Lots of handwaving to support a strange theory for no apparent reason other than to portray Amazon as evolving into the same business as Google (a pretty face on an Ad business) for a greater narrative. This feels like establishing facts, with a weak foundation, to reference in the future.

OFC this kind of information has been talked about before: https://marketingland.com/analysts-say-amazons-advertising-b... and now Amazon has figured out how to maximize by focusing on other "big brand" items, rather than the ever pitiful mom-and-pop (https://www.innovell.com/it-is-just-amazon-advertising-but-w... - all sponsored focused) but the market is only so big (eyeballs). If AMZ is only 10% of retail, it doesn't make sense to have the same spend as on Google or other retail-focused (Target, Walmart, BestBuy, etc).

I just don't see how this made a good case.

I think you're misunderstanding where many of these ads _are_. They're on Amazon itself. Search for anything on Amazon, and you're going to see sponsored listings in the search results. Look at a product page, and you'll see competing offerings.

I've got some novels up on Amazon and it is totally accepted within the indie author community at this point that Amazon ads can absolutely be worth it.

In fact, pay to play has been the requirement for some time now. Organic discovery is long gone. It also effectively lowers net book margins.

Got me to thinking that I'd like a way to signal Google that my intent is either info or buy. Now it's all buy and the results suffer.

Nothing in your comment would come close to refuting Amazon makes more from ads then AWS. Read the article.

Reply on news.ycombinator.com