Webvan, 20 years ago, had the right idea - don't pick from retail stores, use a proper fulfillment center. But they had 3% market share in 30 cities, instead of 30% market share in 3 cities, because their VCs thought the "first mover advantage" was crucial. So they had too much infrastructure for their volume.
This approach is much more expensive in the U.K. at least - I’ve done these sums for retailers (I’m a logistics consultant who specialises in e-commerce and grocery) and it’s hard to make a distribution centre cost competitive if you have an established store estate that you don’t want to get rid of.
In short, a distribution centre gives you:
* improved pick efficiency (depending on the level of automation), offset somewhat by maintenance costs and the cost of capital.
* increased delivery/transport costs (stores are generally closer to the customer than the cross dock point / outbase, vehicles are already going to the store for the main delivery)
\ flowing more product through existing retail points reduces waste (expiring products).
* increased property cost (you have to pay for a distribution centre at c£9 per square foot incl rates etc, while in retail the space is more expensive but you are already* renting it).
The one thing a fulfilment centre does give you is better service though. You can hold more range, better track dates/batches so you can give your customers the expiry dates on the website and increase availability.
So the honest answer is that delivery from a central point isn’t a slam dunk, and can definitely work but probably won’t be cheaper.
You see all this with Ocado in the UK, right? Distribution centres only, so great range, and a brilliant app … but as I understand it they make most of their money selling tech to other retailers.
Does the loss of normal customers factor in the calculation?
Tesco are the biggest grocery store in the UK and their from store fulfilment is beyond annoying for me. So much so that I avoid shopping with them now.
I'm guessing stores in the UK tend to be smaller. In-store distribution is mildly annoying in US supermarkets but no more so (in fact less so) than stocking that blocks aisles is.
Much more, since you have to navigate around 2 to 6 fulfillment trolleys in every aisle, as well as the usual restocking trolleys and random staff walking around and not appearing to have any purpose.
Which indicates fulfillment is best done by a company whose primary business is fulfillment, not brick and mortar retail. This is a classic disruption scenario.
I don't read that at all. I read it as there are pros and cons and, in practice, the relatively limited grocery delivery/pickup that happens seems to be from B&M stores. I expect that any delivery/pickup businesses that didn't thrive in the past 18 months aren't going to.
Not really - centralised distribution doesn’t mean it’s best done by a company whose primary business is fulfilment, and in fact most major supermarkets have the in-house capability to run these depots by themselves if they want to (for e-commerce grocery specifically there are advantages in running yourself).
While I never ran the numbers, that was my conclusion as well. Cost wise, use your existing retail infrastructure. It is easier so, to set up a separate fulfillment flow for home deliveries. It's a trade of, like always.
This is starting to change in the UK as the proportion of customers ordering online keeps increasing, for example Tesco have started setting up "Dotcom centres": https://www.tescoplc.com/news/2013/tesco-opens-sixth-grocery...
(of course that's best of both worlds, fulfil from your existing stores until the demand justifies a dedicated centre)
Don't they already have distribution centres? When covid started, some local store chains just started supplying consumers from the same distribution chain through which they (re)supplied all the many small stores they had.
Not the parent poster, but a distribution centre for stores has a different function to a distribution centre for individual customers. The former is about getting whole boxes of products onto large trucks to deliver to stores, the latter is about getting individual items from inside those boxes into bags in vans for individual customers. For that you need a different warehouse layout and many large businesses won't have that space available for the scale they need.
Click-and-collect from supermarkets' existing stores is likely here to stay. It's very appealing for people on strict budgets, because you don't get to the cash register then realise you have to put things back.
However, any retailer with existing stores wants to operate click-and-collect from those, as if you realise you forgot something on your order you'll go into the store.
> Click-and-collect from supermarkets' existing stores is likely here to stay. It's very appealing for people on strict budgets, because you don't get to the cash register then realise you have to put things back.
People on strict budgets typically also need to stretch that budget as far as they can; they likely can't absorb the markups on each item that pays for the in-store shopper to collect them.
on an item where quality parameters vary so much ( size, visual shape, ripeness etc), customers will not give up the value of touching and picking an item.
This applies to a cabbage or a fish.
In the case of a pair of socks which are standardized, they may not care as much to physically pick up stuff.
---
I have always thought that the disruption to costco would be an online warehouse, that could gather the customer's preferences ( patterns of buying - and consolidate them and send them to localized centers ).
Customers can then pick them up or have them delivered.
Discovery will purely be through the web.
I'm certain they will, at least here in Sweden. We have bought most of our groceries online for years. It is worth quite a bit to not have to use half a Sunday to drive to a big ugly supermarket surrounded by a huge parking lot and then run around trying to find whatever that isn't where I expected it to be.
It isn't common in the UK, where online supermarket shopping has existed since about 1997. (And not as some obscure thing, the biggest supermarket launched it then.)
Over 20 years later, this still happens to me every time I ever buy root ginger.
I always want a lump / knob / finger / whatever of ginger, and without fail forget that the unit of measure is a kilogram, which would take me months to get through.
How much is the scenario helped by significantly reducing the number of SKUs? e.g., customers are choosing from two toothpaste options rather than 10+?
If a shopper wants milk, bread and scented candles, and you don't sell scented candles when your competitor does, they'll go to your competitor for candles and pick up their milk and bread at the same time.
So if you want to get a customer's main weekly shop, you can't cut your range too far.
Of course, there's a space in the market for gas station convenience stores with a smaller range, higher prices, great convenience and no aspiration to get the customer's main weekly shop.
I would say that a primary difference is the investment runway the successful players have enjoyed. Some of them are still not objectively profitable to this day, but their massive topline revenues, scales of operations, and levels of name recognition makes many of them remain attractive investment targets nevertheless.
> I would say that a primary difference is the investment runway the successful players have enjoyed.
No. The market size expanded exponentially for each of these opportunities - changing the economics of the business models and making them viable at scale.
The internet was nothing in the 90s. A place for mostly nerds in first world countries.
Today we've connect 4 billion+ humans around the world.
For better or for worse. When you say "connect", that means get them in front of a couple of hundred platforms (globally, comes down to couple dozen in the west) to transact on and generate revenue for those market players. It's like we took the 32bit color animated GIF internet and made it a 128 color static PNG of mostly similar shades of 10 colors.
> No. The market size expanded exponentially for each of these opportunities - changing the economics of the business models and making them viable at scale.
Maybe. Labor in China and the Chinese market absorbing inflation also helped.
My expectation is grocery delivery just barely missed the boat, because of import malfeasance crackdowns since April that is affecting everything.
Actually it reinforces that the idea actually is the easiest part of the equation. Execution is the key and the winners are the ones who executed the best. You can't compare Webvan to Amazon.
Amazon ended up owning and operating Webvan.com.[1] The people who developed Webvan went on to develop Amazon Fresh.[2] The founder of Kiva Robotics, which was acquired by Amazon, was from Webvan. Amazon's grocery business is directly descended from Webvan.
Yes, it was a descendent. Instacart is a descendent of Webvan too. The idea is the same but Amazon already had preexisting technology and market dominance. You can't compare Webvan to Amazon.
That might be a factor, but given that they all failed at about the same time I think there's a good chance that timing is just as important. Adding billions of people to the internet and mass adoption of smartphones may have been necessary precursors to the success of Amazon/Facebook/Google/Uber.
I remember in high school learning how to build little websites and thinking, “it sure would be cool if we could build a little online yearbook for our school! Everyone could have their own little profile page!”
Did you also think “l can then use their personal information to target ads and make $billions!” I wager you did not, because you’re probably not as depraved as some.
Doesn’t even feel like some of these were necessarily ahead of their time, almost like investors didn’t take as many risks as they do now with burn rate etc. of course, the dot com bubble burst in a bad way, so perhaps this theory isn’t very sound
The model that I think is going to work would be the equivalent of traderjoes.com
TJs apparently does not want to do it. It makes no sense to package items to be placed on shelves and then have people go pick the items off consumer shelves and deliver. There is so much vertical optimization that could be done.
I switched to FreshDirect during the pandemic, and I am probably never going to depend on a large grocery store again. FreshDirect is my weekly delivery, and the nearby bodegas have what’s needed when I need a couple things.
Every large NY grocery store consistently seems to have either poor-quality or extremely expensive produce, which was drastically exacerbated last April: if you don’t go soon after they stock, people pick over, scratch and generally take all the edible-looking choices. Whole Foods was terrible for delivery: the pickers seemed to just choose what was available, which inevitably tends to be what in store shoppers left over. I’d much rather on average get an average apple or onion, without having to spend time figuring out whether it’s “good”: average is almost always good enough.
My total time spent grocery shopping went down, while the quality of the food went up. Their website does a really good job of focusing me on what’s fresh, in season, and otherwise likely to be good this week, and it seems like there’s a lot less waste in this model. I can also order more than I can walk home myself, which is pretty valuable when you don’t own a car.
The situation is completely reversed outside of a NYC like dense area where everyone has cars. Once you have a car, it makes economic sense to use it and you will likely drive near a grocery store at some point in the week. The grocery stores are also not constrained on space so the quality is much higher and rent or land costs less so prices are lower too.
Although, I do have to say grocery stores in London and Toronto were much nicer than NYC in my experience, so it may be a cultural/customer demand thing too.
+1, I've been very impressed by FreshDirect. I have a Morton Williams across the street (crowded, bad produce, iffy selection) and a Whole Foods around the block (pretty nice, but long lines and I didn't really want to go in before vaccination), but FreshDirect has just worked.
Their site and app are glacially slow, though. Not sure what's going on there. And for a while they've had some text-entry glitch that reverses text every now and then. But that's just griping.
Also, with kids it's nice having the weekly 3-4 gallons of milk just show up rather than schlepping them home from the bowels of Whole Foods.
I've gotta say, groceries in NYC seem uniquely awful. We stayed with friends in Greenpoint for a week and it was grim. The supermarket that was around was dire, the bodegas were fine, but really expensive for not great quality or range, and then you could walk half an hour to Whole Foods in Willamsburg, which was eye-wateringly expensive and rammed.
Compare this to where I live in a London, which is a similar PT distance to central London as Greenpoint. We do all our shopping at a brilliant greengrocer just the round the corner, there's specialty delis and such in walking distance. There's also a number of Sainsbury's and Tesco locals. Then, about 25 min walk away, there's two major supermarkets (a Waitrose and a Morrisons). I guess people in NY don't cook?
Amazon Fresh does really well in NYC, has been a positive experience except that their baked goods are frequently sold out. Wal-mart wants you to go out to one of their stores to pick things up - great strategy in a city where nobody drives. Target’s master strategy wavers between being out of everything and not telling you they can’t deliver something until you try to check out. So I’ve mostly used peapod to purchase items from Aldi - that has been the most reliable way to get groceries delivered.
> I've gotta say, groceries in NYC seem uniquely awful. We stayed with friends in Greenpoint for a week and it was grim.
What you experienced are the fruits of gentrification in NYC. Greenpoint is an overpriced neighborhood which used to be a Polish enclave (still is to a degree but on life support from what I see). People down there shopped at local ethnic markets, many of which are gone due to gentrification which drove real estate prices through the roof.
Young people and/or "up&coming" neighborhoods in NYC in small apartments (ie - Greenpoint, Bushwick, E Village, etc).
Go to UWS/UES/Greenwich Village and you will find very nice, very expensive, very overcrowded groceries.
I recently needed to buy a propane heater. Ordered it online for pickup at Walmart because there was only one left in stock.
Showed up at the Walmart thinking I could just pick it up right away. Turned into a huge big thing. The staff only goes around 1-2 times a day to fulfill these sorts of orders so I would have ended up waiting.
I ended up buying the last one on the shelf, which was mine anyway, and cancelling the online order. I wonder what would have happened if someone else bought it first.
I see this more as iterating on an idea to provide this service, and what I experienced, is the MVP.
WalMarts inventory system preemptively orders based on sell through rate, on-hand, and historical trends. Backroom stock is logged so a pick order can be fired when shelf count gets too low.
Source: worked in a few Walmart back rooms picking and binning among other things. The was about 8 years ago so things are likely a little bit different. Hopefully better since shelf max counts were always way off or we'd get flooded with freight and lose stuff in the bins.
I order for pickup all the time nowadays, at various retailers (Walmart, Target, Home Depot, Walgreens, Staples), and I have only had 1 time I can recall that one of the retailers reported they did not have the item I ordered. I must have picked up at least a 100x by now, and I feel like as long as I wait for the email that says it is ready for pickup, I am good to go.
I actually order online for pickup to ensure no one else buys it if there is only a few of the item in stock.
Why is a real fulfilment centre better? I can see an Amazon centre with robots etc being more efficient, but other warehouses seem no more efficient than going through shelves at a walmart.
I would think there might be lots of reasons, but one that comes to mind is that supermarkets are physically designed to get the customer to spend as much time and money as possible in the store as opposed to finding the items on their list(s) as efficiently as possible like a grocery delivery service would want its workers to be able to do.
Amazon just recently started to use robots. Most existing operations still use physical labour. FCs are more efficient, as single entities, because they are designed as such. They have proper infrastructure, tools, people, processes.
For grocery fulfillment, I don't think they are better at scale. Because they are for from the customers, they are not everywhere. Existing stores, so, are there. They use established replenishment from existing warehouses. Which means you have to use existing staff and store processes (and software) to fulfill orders. Which is hard. So it's either expensive or hard to do (which could expensive in itself). Over simplified.
Supermarket floors aren't designed for pick and pack efficiency, they're designed from the ground up for phycological manipulation.
Warehouse trumps Supermarket for efficiency by multiple factors easy.
Just as an example imagine the 50 most ordered items. I can put these in a tiny area closest to my packers in a warehouse. How far to you think you'd need to push a cart between those 50 points in an average Walmart?
A good fulfillment system tracks its inventory and communicates with the ordering system, so you can't order items they can't deliver. That's the big problem with ordering online from grocery stores - about 80% of what you order shows up, which is a headache for both buyer and seller.
Mostly because markets deliver a less consistent experience (delays, cancellations due to vacations), and you face situations where your employees and customers are fighting over the last bag of chips. It also becomes harder to plan logistics because most stores (at least where I live) have limited real estate.
The semi-automated ones have different issues. Only Amazon really has the tech fully figured out to a degree where issues are resolved almost instantly. Amazon can also use less but larger scale fulfillment centers because they deliver mostly small objects at an arbitrary time, whereas you can't leave 6 bags of groceries at a front door.
At Amazon scale, though, that approach works.