Hacker News new | past | comments | ask | show | jobs | submit login
The Illusion of a Profitable Amazon Prime Now and Prime Fresh (stratelogical.com)
61 points by stratelogical on Mar 30, 2018 | hide | past | favorite | 83 comments

The author dramatically underestimates the opportunities of scale and future tech in her model.

Amazon is already:

- Installing lockers into apartment communities https://www.amazon.com/b?ie=UTF8&node=6442600011 that make deliveries dramatically more efficient.

- Predictive logistics & ordering patterns are a core Amazon discipline and they are world class at it.

- Droning looks like a gimmick today, but why should it be in the future?

Imagine a not-far-out future where: - Predictive ordering algos keep right balance of product on hand - Order comes in, robot loads it from GIANT physical warehouse that is NOT in a city center - Order is sorted into a "locker" or a "single" end point - Locker end points are put onto a locker-loading vehicle that runs a daily route that ties into the 2hr timeline. - Single points deliveries are brought by automated heavy delivery truck to city center range from which point they are either droned or courier delivered

- - -

That future ain't that far off folks.

And in the meantime Amazon, who has their shareholders trained to reinvest every dollar of profit into their scale, is going to go and corner the multi-trillion dollar grocery & last mile market.

The lockers are amazing. If you live in an apartment building, the leasing offices which sign for packages are often only open during business hours. If you live in a house, there's a good chance that someone grabs your package off your porch.

So what do you do if you need a strip of LEDs for a project that you're putting together by this weekend and you don't have the right hours open in your schedule? (Or [insert random product here]?)

Just have it show up at the supermarket/drug store/gas station in a couple of days. Easy.

It's also cool that the apartment buildings can install the lockers and use them for packages which Amazon never touches, for the whole 'business hours' problem. But as Amazon's programs for offering free shipping to 3rd-party sellers are making it more of an 'Everything Store' by the day...well, it'll definitely be interesting to see what happens.

> If you live in a house, there's a good chance that someone grabs your package off your porch.

I don't know about "good chance". Small chance, perhaps. I've still never had a package stolen in 20+ years of ordering online. But many folks have packages delivered to work for this reason (and I admit I'm more likely to have expensive packages delivered to work). But the lockers solve a problem I don't really have, and suddenly make ordering online as inconvenient as buying stuff in-store.

It really depends on your area and living situation.

I have a place in suburban Minneapolis and I could (and have) left packages sitting on my doorstep for weeks while on vacation. Still there when I got back - only a single lost package over about a decade now.

I also have a place in Chicago. A package left outside at the door would be gone within hours as I live directly off a very busy street. Even if put inside, if the door gets left unlocked or ajar a bit you have a very high chance of the package disappearing. Packages get delivered to the office or not at all when I'm in town.

I agree re: having zero interest in driving to the nearest locker though - I'd just go to the store. It's possible they will reach the density in Chicago where I can walk a half block - which I'd be okay with - but I can't imagine that being cost effective for most areas of the city.

Isn’t that what Amazon’s IOT key is trying to solve?

> If you live in an apartment building, the leasing offices which sign for packages

Somewhat OT, but are these common in US (or wherever you are)? If so, out of interest, what other duties to they have?

In Finland, I've never seen such offices even in the larger apartment buildings with hundreds of dwellings.

Parcel lockers (both public and recently in apartment buildings) are a thing here as well, but they are mostly replacing pickup-from-post-office. Home delivery has always been a premium option when ordering online, not the norm. The network here is owned by Posti so any company or individual can send packages to them, or use them to send packages at cheaper rates than going to the post office.

Yes it's very common to rent from a company rather than individual landlords. Some of these companies are massive managing hundreds of thousands apartments. Many of the building are specifically built as rentals and all units are for rent hence there is often a leasing office somewhere in the building.

OK. Here such companies just have one central office per city/region, no building-specific ones. The largest companies here manage "only" tens of thousands of rental apartments, though.

How are they handling visiting the grounds/future apartments for interested renters? That's one of the main duties of the leasing office people in the US. When looking to rent one visits many such places and asks the leasing office people various questions while doing so.

Typically by making an appointment.

Scheduled public presentations, or appointments.

I think you would be hard-pressed to find an apartment complex without a leasing office. I have never seen that in the US. Mostly I think they are there to handle showing people around and finalizing leases. They also can handle maintenance disputes and forward complaints to the right people.

Also, for the sake of those outside the US:

Heavily weight the first impressions of the leasing / office staff to determine whether you WANT to live in an apartment complex: As stated above, these will be the people you will deal with for anything from clogged drains - to being unable to pay rent on time.

In short: you want to make sure they are friendly, professional and courteous -- or it may ultimately detract from the perceived quality of living in that Apartment [0].

[0] - I lived in 6 apartment complexes from time I was a broke student to young professional with a family.

Commonly apartment complexes here (Finland) are owned by building-specific limited liability housing companies that do not make profit, with shares corresponding to apartments. Those shares are then owned by various individuals or rental companies, so a building office makes less sense.

But if a rental provider company owns the complex outright, they still don't have building-specific offices, but just one central office per city/region.

That arrangement would be considered a co-op in the United States.

While such an arrangement is possible and legal everywhere something like 95% of all co-ops are in New York City.

What's the upside for the company that owns the building?

It's fairly common in newer and/or larger buildings; walk-ups and buildings with a single landlord don't usually do things that way. But a lot of people moving to big cities around here are probably familiar with the idea. They also handle maintenance requests, showing people units, and sometimes things like community events.

The USPS does also have package lockers - larger boxes with keys that get left in your mailbox - but they aren't universal and companies like Fedex and UPS rarely have access to them. You can also have things held at the post office.

It's not that there are no other ways of solving the problem, it's just that the lockers are convenient and requires very little planning or forethought.

Many states require an on site property manager for larger buildings. So it is normal for them to handle collecting rent, screening tenants, signing leases

Is this the same thing as a concierge? I've seen a few apartment blocks with them in the UK, but it's rare.

I'm an Amazon seller and Amazon can't even keep track of products in their own fulfillment centers. They've added $10k+ worth of random inventory into my account and I can't even find someone who cares enough to fix the mistake. My support ticket is 2+ months old now. At least I'm on the better end, I know someone else with the exact opposite problem who is magically losing inventory at Amazon's fulfillment centers... Based on this, I want to say that that future is off but it's probably a different team... From the top level, figuring that out probably isn't hard even if Amazon still has far to go when it comes to execution.

From the perception of someone who's currently learning how to code to build a restocking software for Amazon sellers.

>Droning looks like a gimmick today, but why should it be in the future?

It's loud. And dangerous.

I'll be the devil's advocate here: Cars are loud and dangerous - we figured it out as a society.

We "figured it out as a society" by sacrificing 40,000 people/year on the altar of the automobile, paving huge portions of our most valuable real estate, and adding car debt as a necessity for our under-class to function in society.

Maybe we'll be smart enough next time to think through the costs of our revolutions.

We've become significantly more risk-averse as a society though. If cars were invented today, it seems extremely unlikely that the intimate mixing of pedestrian and vehicle realms would be tolerated in the way it is now. We generally only accept dangerous transportation, structures and substances if they are a legacy from an earlier age. And even then you see high pressure to reduce or eliminate them, e.g. cigarettes.

No, we haven't. The more civilized among us, like the Dutch, have gotten rid of them as much as they can, and other places promote public transport too.

That's a fair point, but have you actually heard a drone? It's really loud. It will be incredibly hard to counteract public pressure.

I actually just bought one last weekend. Haven't flow it yet, so I suppose I'll find out very soon!

I believe you that they're crazy loud, but I think the technology will improve and we'll figure out regulation, fly zones, etc.

Of course, this is all predicated on us agreeing as a society that we want this thing integrated into our communities. This want is unclear and I think that's the meat of the point you were making which is a fair one.

They are not able to be heard (at least in a common city setting) above 100ft or so, in my experience. Maybe less.

I imagine any drone network would fly at a couple hundred feet for most of the journey, and only be a "nuisance" for the period of time it has to descend straight down to the delivery point.

Cars exhaust system is louder, it was fixed in the process.

Also, they have the potential of being semi anonymous, captured, suffering mech failure while over people, etc.

Eventually, Amazon could also combine their predictive ordering with their lockers to stock common goods before they're ordered (some combination of a vending machine and a pickup locker).

There must be a lot of slop still in the system, because I had 3 separate small packages (from the same order) delivered by 3 different vehicles to my door within the span of 30 minutes (on the same day) last week. All three items would have easily fit into the smallest box that they delivered. Presumably they had enough deliveries in my neighborhood to make this worthwhile.

I would add Robot deliveries to this list as well. Yes, if these take off, things will look different.


If you assume a driver’s wage is at least the Seattle minimum wage of $15 per hour, add nominal fulfillment and delivery costs of $5 per order (the contractor’s fuel, car etc.) and nominal sales and marketing and general administration costs of $5 per order, Amazon needs to fulfill close to eight minimum order ($35) deliveries per hour to breakeven just on the variable costs

Across 25 associates, the average number of orders was only 3.

There's a third option for how this service survives: increase adoption so that:

- variable SG&A drops from $5 per order

- stops are closer so delivery costs drop from $5 per order

- drivers can deliver more orders per hour

It seems like the person who wrote this analysis assumed unchanging variable costs, rather than recognizing that they will decrease over time.

1) Where did they get a $5 marginal SG&A number? This seems high to me and no source is given.

2) There is no way that, at scale, delivery costs are as high as $5/order.

The IRS standard rate for mileage in 2018 is 54.5 cents/mile. (Maybe not the most accurate metric to use, but a decent rule of thumb.) So this analysis assumes that, at scale, for the average order a driver travels 9.17 miles. This routing would be incredibly inefficient - Amazon should "batch" deliveries so that in a 10-mile "loop" a driver can drop off 3+ orders.

Furthermore, I'm not sure if Amazon drivers are employees or 1099 workers. If they're independent contractors, the delivery costs are irrelevant to Amazon, since they're passed through to the drivers.

3) Naturally, drivers utilization rates will increase as demand increases.

Yes, you are right. But increasing adoption across different zipcodes won't help that much (although, yes, it’ll bring SG&A per order down). The adoption has to increase in the exact same area (your point #2 has to be true, i.e. stops should be so close, like an apartment building) for drivers to make more orders per hour, and for this service to be profitable in the long run.

Also critical is the that customer tips reduce the wages Amazon pays to the driver.

He did very little analysis on how the actual Prime subscription price contributes to the profitability of Prime Now.

You cannot evaluate Prime related services in isolation because the whole point of Prime is that you get a bundle of services that make your life easier and allow you to get a bunch of things for a flat subscription rate. Some of those services could be forever unprofitable, but if you remove them from the bundle, you could be destroying more value and making the whole program unprofitable.

The author fails to evaluate at least two key points:

1) Prime is still in its infancy. It's clear that Amazon's goal is to reach a total share of the market that can only be compared to paying for your phone bill, except in this case the phone bill is Prime, and Amazon is the only provider. That changes the whole picture regarding economies of scale and flywheel effects.

2) The author did his pseudo-experiments (asking Prime Now drivers how many deliveries they did last hour ) in Seattle (I presume since he mentions he is a PM at Amazon).

Of course Prime Now it's not going to be profitable in all markets but it's also not that far from breaking even. The real profitability will come from highly concentrated markets like NYC and Singapore.

I'm rather surprised to see an Amazon PM openly slagging their company's own product, or at least its profitability. I wouldn't have expected Amazon to tolerate this sort of thing, and this seems like a career-limiting move pretty much anywhere.

Given the number of employees and what we know about working at Amazon I'm sure there's a subset at any given time who are trying to get out and might fish for a dismissal if it comes with severance. Rather than just quit.

I tell you, this service has been a godsend. Emergency trip to the hospital, no food at home? Prime Now.

Living alone sick, no food in the home? Prime Now.

Parents feeling poorly and I'm 3 states away? Prime Now.

>Emergency trip to the hospital, no food at home? Prime Now.

> Living alone sick, no food in the home? Prime Now.

> Parents feeling poorly and I'm 3 states away? Prime Now.

I hope that any of those three events don't happen to you (or to anyone else) very often.

I mean, I perfectly understand the utility of the service, in a given set of "emergencies", but the profitability (at scale) would come if the service is used consistently, every day, not in a handful of (hopefully) not often recurring cases.

> Stuck at work late, need kitty litter but no local groceries are open by the time you leave? Prime now.

> Need kitty litter but don't have a car and don't want to hulk 80 lbs of litter home? Prime now.

> Need a random ass torx screw driver and don't want to drive to home depot across town? Prime now.

> Forgot the ground beef but don't have time to get it because the oven is already on, cooking for the company coming over? Prime now.

> At a beer festival and sick of standing but didn't bring your camping chairs because you don't have any? Prime now.

> Standing in line outside in the rain but didn't bring an umbrella? Prime now.

> Ran out of Christmas cards? Prime now.

> Nieces birthday and no time to go shopping? Prime now.

As somebody who lives in the city without a car and nowhere near a home depot or a grocery store that is open past 10pm, prime now has been an absolute godsend. The fact that I can order a ton of items that I just _cant_ find locally no matter how hard I look, but also don't want to wait for shipping from Amazon/Ebay/whatnot is an absolute godsend.

I don't think you need to cook up an emergency to have a good reason to use Prime Now. I have tons of 'micro emergencies' every month where I'd gladly tip my driver a bit to bring my items within a few hours.

>I don't think you need to cook up an emergency to have a good reason to use Prime Now. I have tons of 'micro emergencies' every month where I'd gladly tip my driver a bit to bring my items within a few hours.

Yep, your examples are a lot less "emergencies", the previous examples were all connected to unpleasant events, but - with all due respect - yours are all connected to the same "base issue", forgetting to procure in time (or in advance) whatever you need.

A very handy solution, in those cases.

I just relocated to Alabama from NYC and one of the only things I miss is Prime Now. I’m happy to pay a small premium to get something “now” without having to leave the house. Cables / electronics / “just thinking of you” presents for my wife / “crap, we’re out of X and the kids are actually playing happily...”.

I would frequently get heads down on a project, realize I needed widget X, would order on Prime Now, continue my work, and have the widget when I needed it.

As soon as it’s introduced to Birmingham my suburban life will be complete...

You could say the same thing for UberEats ;)

[edit] Disregard, can't ask an earnest question on HN either.

>Is this a common state of affairs? Who doesn't have at least a week's supply of food

On the morning of the grocery shopping day, I'd guess most families do not have 7 days of food left to eat.

In America, the average # of grocery trips per week was ~2.[1]

That frequency doesn't seem so strange if the number of trips is affected by: (1) family size such as 4 people including 2 parents + 2 children, (2) refrigerator size of typically less than 20 cu ft (not the luxury 28+ cu ft or extra refrigerator in the garage), and (3) preference for fresh & perishable food instead of canned items.

[1] https://www.statista.com/statistics/251728/weekly-number-of-...

Yeah my dude, I'd almost venture to say that you are the outlier here (echo/bubbles, yay!). For many that I know, food is an on-demand lazy loaded resource. Sure, sometimes they buy groceries (usually when they run out of pop-tarts) ... But the odd day when they actually cook dinner, it's a trip to the store for just those ingredients that are necessary.

Is it horribly wasteful and inefficient? yes. Does that get factored into the equation? no.

Efficiency aside, that feels like awful planning to me. Growing up in the Bay Area, the threat of a significant earthquake has been ever-present. My family always had several dozen bottles of water and a reasonable amount of canned/jarred/dried food around.

Yeah, but do you want that sort of food when you are sick? Plus I don't dig into my emergency supplies when there are other options available.

You need to rotate your supplies.

I put a note into my calendar with the expiry dates for the rations. I eat the rations when I have purchased a replacement.

Emergency food is emergency food.

Yep! It's terrible planning ... but that doesn't take away from the fact that it's an absolute reality for a ton of people.

My goodness. Most people I know have refrigerators full of condiments. I have that and two bottles of seltzer.

Living alone and sick, I don't want to cook. I want cereal, orange juice, ice cream, easy to prepare food, etc. These are the sorts of things I don't normally have around.

My emergency supplies have food, but I don't eat the emergency food unless there is an emergency.

> Is this a common state of affairs? Who doesn't have at least a week's supply of food and water in their home?

I cant tell if this is a serious question? I definitely do not have more than 1 weeks worth of food in my home.

Hell, the 5 cases of Soylent would last a long time, not to mention a fridge and pantry full of food. Maybe not stuff I want to eat, but they're calories in a pinch.

I suppose if you take it literally, this isn't too common a thing. Having a week's supply of _food_ is easy, but if you want to have a weeks supply of _meals_ that's a different situation.

There's a difference between literally no food at home and no fresh food. You could of course eat a can of beans that has been sitting there since Obama was in office, but you'd really rather a fresh meal.

Anecdotal, but I don't have that. And I don't feel like I'm an outlier.

I like how you're getting completely opposite answers about what kinds of food different people have available, but you're getting downvoted anyway.

The article estimates certain costs (e.g. $5 paperwork/overhead) without much factual evidence. Relatively small reductions in those costs could make all the difference in profit or loss on any particular order.

The author is a pm at Amazon

Amazon Flex will eventually encompass more than just delivering goods. Delivering people (uber/lyft). Retrieving packages from households and shipping them, maybe even last mile delivery?

I feel like this article completely misses the point about the Amazon philosophy. It doesn't matter if Prime Now / Amazon Fresh are profitable. They aren't supposed to be profitable. They are building out the infrastructure of Amazon Flex for themselves AT COST and then they will open it up to other services and uses that will undercut any competitors (rideshare is the first one that comes to mind)

The focus here should be on Amazon Flex and how/where it is going, because that service will be the one that turns the actual profit. Just like AWS.

Yes the author is kinda silly. People have been writing these sorts of silly articles about Amazon for nearly 20 years. Yes, Amazon is wasting investor capital. No, investors don't care as long as the innovation and growth train continues.

Still, as an Amazon investor, I do wonder why Amazon has to do everything themselves. It makes me nervous that they are essentially building or their own ride sharing network rather than leveraging existing networks.

There's a revolution coming to urban logistics that people like this author cannot understand. What Uber has demonstrated is that cities are much more efficient than anybody previously imagined. The day is fast approaching when on demand transport of everything will be cost effective and when that happens cities will be remade. All that retail space can become storage. Restaurants lose their tables and become industrial kitchens. Office space becomes highly fragmented with possibly just core teams of 10-20 people requiring co-location. Schools fragment; students will go to specialized teachers and facilities wherever they happen to be rather than betting everything on monolithic school systems. (Education "regresses" into large scale tutoring.) Hospitals ironically grow more centralised as technology becomes every more sophisticated. Tying all of this together will be extremely sophisticated and dynamic transport networks that can indeed move anything anywhere in the city in under an hour for very low cost.

This is already happening in China btw. This sort of urban hyper-on-demand culture has already taken hold. The author will look back on this article on ten years and regret that they failed to see the revolution.

> Still, as an Amazon investor, I do wonder why Amazon has to do everything themselves. It makes me nervous that they are essentially building or their own ride sharing network rather than leveraging existing networks.

When Prime Now started, there were no other companies that were doing what Amazon wanted to do. Let alone, keep up with the expansion that Prime Now has realized over the past few years. In addition to all that, Amazon loves to optimize every little detail. Flex is the logistics infrastructure that didn't exist to support Prime Now and other delivery programs such as Fresh, .com packages, etc.

None of what you listed is happening in China. It also will not be happening anywhere at any time. Restaurants won't become industrial kitchens because that is not what restaurants are for.

Some restaurants are becoming just kitchens: https://www.theguardian.com/business/2017/oct/28/deliveroo-d...

These might take some of the business of physical restaurants - probably not a large amount, though.

Someone once joked with me that Amazon Fresh was never intended to be a profitable business. It existed to scare UPS and FedEx by saying 'hey look, we can deliver stuff ourselves!'.

I'm not sure how much they were joking.

If you want a look from the other side of the equation, here's what it's like to be a contractor delivering for Amazon.


Am I missing something or does the analysis assume every order is a minimum order? Personally my orders average around $100.

Have not been to a physical grocery store in over 6 months because of Amazon Prime Fresh. Would be willing to pay much more than I do now after being accustomed to saving so much time.

> It also works like magic for Amazon shareholders, by pulling a Houdini on them and making their invested capital disappear.


Author is pulling per order estimates out of their hat, and seems to have the missed the fact that driver tips reduce the wage Amazon pays the driver until the minimum wage is met.

Very poor analysis with both bad assumptions and weak imagination.

You fail to take into account reduced price sensitivity and price manipulation, for instance Whole Foods prices in my area, Tucson, are 30-50% higher than Sprout’s or Trader Joe’s, and that nominal sales and marketing and general administration DOES NOT cost $5 per order, but functions more as a fixed cost.

Remember how Wal Mart used to move into small towns with low prices, put the local shops out of business ,then raise their prices?

Amazon's doing that to the entire world right now.

Is the Prime Flywheel a term Amazon actually uses?

Very good analysis on Amazon

Very good analysis of Amazon

I think that the price to the vendor of FBA was not taken into account.

It does not take much research to figure out what Amazon is charging the vendors for Prime under FBA. You can look at the rates for Fulfillment by Amazon online. Remember that Amazon doesn't pay for shipping to their warehouse.

For instance, an $18,1 kg roll of 3d printer filament shipped via FBA nets Amazon about $8 by my calculations. Yes it has to be shipped and fulfilled, which involves taking it off the shelf and adding a shipping label. What are Amazon's costs for shipping/delivering 1kg small packages? Far less than $8.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact