Hacker News Reader

Comment by finnh | Hacker News Reader

This has all been pretty obvious for awhile.

Take Instacart: I pay $100/year for unlimited grocery delivery? Which I'm doing once per week, so that's < $2 per delivery. To deliver this service takes the Instacart shopper/driver at least 30 minutes, probably closer to 60. Assuming a $15/hour wage to the driver, that's $13 of loss per order. Which, again, I'm doing 50+ times per year.

And, at the store I order from, prices are exactly the same as in-store so there's not much (if any?) profit they could be extracting on that side.

Let's all just enjoy the VC money subsidizing our lives for this (second) brief moment. Just like Kozmo.com back in the day.

View on HN

Yes. In many ways this seems like a (admittedly less severe) version of dot-com happening again. There was also a piece a week or so ago about how a number of these companies are trying to branch into offering these services for businesses.

One of the problems both then and now is that there are a lot of convenience services that people would, in general, like. However, in many cases, there aren't (enough) people willing to pay the cost of delivering those services to make them work as businesses.

Grocery delivery can work. Some Whole Foods were offering delivery pre-Instacart and Peapod (Stop & Shop) has been around for over a decade. But you pay about $10 per delivery, which I assume is priced to be roughly breakeven given that the grocery chain is still making money on the food. I think it's fair to say that the market has shown the average person (especially one with a car) isn't willing to pay that--especially given the other real or perceived issues associated with ordering groceries online.

Grocery delivery only really works as a business if you're shipping the orders from a distribution center. I believe Peapod (and others like FreshDirect) use this method: they have a giant warehouse full of food, and they load trucks up for deliveries every morning with the day's deliveries.

Instacart is honestly a wholly inferior product. I find myself using it whenever I have a craving for something specific and I'm willing to drastically overpay for it. The few times I have used it for grocery delivery, it was way overpriced and the shopper couldn't find half of the things on my list (so I wasn't able to make either meal I had purchased the ingredients for). The only benefit I see from Instacart is 1-hour delivery (most grocery delivery services require you to order at least the day before).

"Grocery delivery only really works as a business if you're shipping the orders from a distribution center"

In Montreal a lot of grocery stores offer delivery for a couple of dollars as a norm, although the orders would usually be placed over the phone, you also had the option of shopping at the store and having your groceries delivered in a few hours. It was only after I left that I realised grocery delivery was not the norm in most places.

Example: http://www.supermarchepa.com/eng/services/delivery.html

Sorry; I should have said grocery delivery only scales past a single neighborhood without a distribution center.

You're certainly right; grocery delivery was more or less a regular service before the advent of the massive suburban supermarket.

I didn't love Peapod either when I used it admittedly a long time ago because they regularly didn't deliver some items--even some I thought would have been very common--and made weird substitutions. I only used them while I had a broken foot and was on crutches because, although I could drive around and even go to the store easily enough, doing a full grocery shopping would have been a pain.

I do know people in SF without a car who do use Instacart regularly but, for me, it's easy enough to hop in the car and go to the store. I also buy mostly unprepared food so I'd just as soon look at it anyway. I don't actually have Instacart available in my area anyway.

Your spot on that some stores are offering grocery services. Our local Hyvee is doing this and has been a life saver [1]. For this to work you already need a distribution imo.

Orders over $50 are free for pickup, just drive there and they load it into your car. To house deliveries are free over $100.

I don't know if they are making anything on the delivery or just taking it out of the per item markup. Just makes one realize how much markup are on the items. I spoke with the delivery guy and he said they are doing past 300 delivers a week.

[1] - https://www.hy-vee.com/grocery/

>Orders over $50 are free for pickup, just drive there and they load it into your car.

Yeah, there's a local chain around me that just started doing this as well. I'm not really the market for it, but I could absolutely see it for parents with small children or people with mobility issues.

I imagine there's a fair diversity in how these deliveries are paid for between explicit delivery charges, probably at least some markups, and simply not including in-store specials.

I enjoy the subsidy, but VC money is rarely "VC money" it's mostly LP money from pension funds, endowments etc., and when they fail there is usually a bail out from a large publicly traded corporation which again is spending everybody else's money. Certainly, wealth is extracted along the way by the VC's, bankers, stock traders, laywers, etc; however, I think at the end of it the joke and the bill are coming at the expense of the average person.

This is really wrong thinking and a good example of misunderstanding unit economics. You're calculating the unit economics of a delivery assuming there's only one "unit". You have to look at the aggregate. That driver isn't just delivering to you. Not every customer is ordering 1x week.

as a reply to his post, this reads like a magic wand wave

While the post you're responding to doesn't explicitly state all the detail you were looking for, I do think you're being a bit too dismissive.

Not sure on the accuracy of the claims, but the "hand waving" is saying that the estimated time spent per delivery may be a large overestimate if that driver is delivering multiple orders per trip, thus reducing the amount of time that a single customer pays the driver to drive out to a specific area. Point number two was to imply that, yes, if they priced a delivery at $2 per delivery, they probably are operating at a loss, but not everybody gets 50+ deliveries in a year, thus bringing the average cost of delivery up. Heavy users have existed in other systems and they didn't become suddenly untenable (Amazon prime seems to handle it just fine).

It needs a bit more detail. Is the assumption of groceries once a week really weak? (literally everybody I know gets groceries once a week)

Is the density of drivers and customers such that drivers deliver at a rate <<15 min?

"Not everybody gets 50+ deliveries a year" can sound as sensible as "Not everybody brushes their teeth every day" or as pedantic as "Not everybody sleeps every day" depending on the underlying statistics.

It's a reasonable response: that driver isn't delivering groceries to you, they're delivering to your whole neighborhood. As the density of delivery orders goes up, there's more and more opportunity to batch nearby orders like this.

In addition: if you pre-order a bunch of deliveries, you're giving the delivery service more cash upfront. That usually allows for greater efficiencies in how the cash is spent and/or invested in the business.

Of course, if they never reach the break-even density, the whole thing is doomed to slow failure.

> And, at the store I order from, prices are exactly the same as in-store so there's not much (if any?) profit they could be extracting on that side.

They could easily be taking a cut from the shop's profits - margins at some grocery shops are pretty good, and at least in theory Instacart are bringing in custom that they wouldn't otherwise have.

Yes - if they're reducing the costs (real estate, cashiers, etc) of the store, they can get money from that side of the transaction. The same is true for food delivery: restaurants save a lot of money if they're not providing space, servers, dishwashers, etc.

This isn't to say that many won't go out of business. Many will, and the market will consolidate.

Gross margins are probably pretty good for some items at grocery stores, but in general their income percentages are razor thin.

Margins at most grocery stores are abysmal. Not even Whole Foods, which has the highest margins in the industry, could subsidize that delivery model.

> They could easily be taking a cut from the shop's profits

If its easy, why are they not doing so? Because there's only so much pain you'll tolerate as a supplier before setting up shop yourself to provide that service directly.

When I looked into Instacart last year, there seemed to be considerable uncertainty about whether their customers paid higher prices than in-store grocery customers. While they claimed at times that the pricing was the same, some of their customers claimed this was not the case. I don't know what the company's current claims are, but I'd be curious to know. This could give them a much easier path to sustainable revenue.

Also, if Instacart customers cannot take advantage of certain sales or club card pricing, Instacart could split this difference with grocer. At some stores, the difference is huge—my Safeway receipts boast that I've "saved" 25 or 30 percent. I haven't—I just avoided overpaying by that much, by purchasing items that are currently on promotion. Hopefully some current Instacart customers can weigh in on how the pricing lines up.

"Take Instacart: I pay $100/year for unlimited grocery delivery? Which I'm doing once per week, so that's < $2 per delivery. To deliver this service takes the Instacart shopper/driver at least 30 minutes, probably closer to 60. Assuming a $15/hour wage to the driver, that's $13 of loss per order. Which, again, I'm doing 50+ times per year."

Except Instacart marks up most items on their website, so it's not really $2 per order (http://www.nytimes.com/2014/05/22/technology/personaltech/on...)

Most items on Instacart cost the same as they do in the stores. They tell you whether prices are the same as in-store prices on a store by store basis. In NYC, only Costco items are marked up, while items from everywhere else (Fairway, Whole Foods, PetcoNow, and Breads Bakery) aren't market up.

Except they don't for the store in question, as stated in the very next sentence of my post:

"And, at the store I order from, prices are exactly the same as in-store so there's not much (if any?) profit they could be extracting on that side."

citation: http://www.centralcoop.coop/page.php?PID=1042

The point of VC-subsidization is only to last until the company can build up both a brand and a more efficient system.

For example, shopping at Whole Foods is inefficient. However, once Instacart has their own (automated) warehouses and can cut out the middleman supermarket, as well as real estate/cashiers/etc, it can potentially shift to being profitable. Add in self-driving cars, and they're set.

(Also, don't forget that for many stores, Instacart marks things up 10-20%. So they're making +$20 on a $150 order.)

Instacart are absolutely not subsidizing until creating their own automated warehouses. Pick and pack for groceries is really hard due to the high variety and expensive and the margins are really small.

Instacart are trying to capture the top-of-the-funnel via marketing efforts and branding, which is capital intensive, but from a variable cost instead of fixed cost perspective.

So basically a business that isn't going to be feasible for the next 5-10 years.

That's not a business model then

> once Instacart has their own (automated) warehouses and can cut out the middleman supermarket

So now they're running a delivery-only grocery chain that has to compete with the well-established brand and distribution network that is popular among their not very price-sensitive customer base.

>Add in self-driving cars, and they're set

Don't hold your breath for door to door that also somehow has to drop off bulky and potentially fragile packages.

Uh no. At that point you're like a UPS delivery man, pack the truck vlby delivery order. Then the car drives to each location, you get the package inside and the car moves on

Doesn't Instacart also charge a thick margin on top of each grocery good you're buying?

I don't think it's only costing you $100 a year, in fact it's probably much more than that.

Not in my case, as I stated above:

"Except they don't for the store in question, as stated in the very next sentence of my post:

"And, at the store I order from, prices are exactly the same as in-store so there's not much (if any?) profit they could be extracting on that side."

citation: http://www.centralcoop.coop/page.php?PID=1042

Unless they can "batch" up deliveries enough. I.e. that 30 minutes actually involves 8+ deliveries to people near you. Hard to do obviously.

I agree that the Instacart economics don't quite work out.

However, to be fair to them, it is also expected that you'll tip.

& I do ... but the tip goes to the driver, not Instacart.

Reply on news.ycombinator.com