Third-party food delivery and arbitrage

A short follow up on Gad’s post from earlier this week on Uber’s interest in Grubhub. There’s a blogpost on The Margins about pizza arbitrage via Doordash (Doordash and Pizza Arbitrage, May 17) that has gotten some attention this week. In a nutshell, the story concerns a pizza shop run by the author’s friend who discovered that (a) Doordash had posted the shop’s menu and started placing orders without telling the pizzeria and (b) Doordash had posted the wrong prices. Wrong as in Doordash was selling a $24 pie for $16. So if the pizzeria owner were to have an accomplice order ten underpriced pizzas through Doordash, he would quickly pocket an extra $80 relative to selling the same pizzas to a real customer. Since Doordash was doing this on their own, that would be $80 of Doordash’s finest venture capitalist supplied cash.

It’s a fun read and worth checking out if only to enjoy the line “Was this a bit shady? Maybe, but fuck Doordash. Note: I did confirm with my friend that he was okay with me writing this, and we both agreed, fuck Doordash.” But there is also an interesting analysis of whether Doordash and its ilk have a feasible business model.

How did we get to a place where billions of dollars are exchanged in millions of business transactions but there are no winners? My co-host Can and my restaurant friend both defaulted to the notion “delivery is a shitty margin business” when discussing this post. But I don’t think that’s sufficient here. Delivery can work. Just look at a Domino’s stock chart. But, delivery has been carefully built as part of a holistic business model and infrastructure. Maybe that’s the viable model.

After the start of this pandemic, my friend actually launched in-house delivery at one of his restaurants. He said he’s starting to get a sense of the economics and explained he’s starting to get a sense of the volume required per location to make the economics reasonably work. That’s what is so odd to me about third-party delivery platforms. The business of food delivery clearly is not intrinsically a loser. Domino’s figured it out. Every Chinese restaurant in New York City seemed to have it figured out long before any platform came along. My friend is figuring it out.

Read Full Post »

Shipping from stores in a pandemic

We have had a number of posts over the years on retailers filling ecommerce from brick-and-mortar stores (see, for example, here and here). From the perspective of inventory management, treating what’s in the stores and whatever is in a fulfillment center as one giant pool of inventory makes a lot of sense. In theory, there is no reason to turn down a web order just because the fulfillment center is stocked out if the needed item is sitting at some mall. The reality, of course, is more complex since picking and packing at a store is going to be more costly than doing the same work at a dedicated facility. Additionally, there is the question of how taking items to fulfill online orders impacts in-store customer behavior.

Now add to those concerns how shipping items from random locations impacts the logistics provider who has to collect and schlepp those packages. Apparently, FedEx has had enough and is working to rein in retailers shipping from stores (FedEx, Strained by Coronavirus, Caps How Much Retailers Can Ship From Stores, Wall Street Journal, May 14).

Read Full Post »

Developments in e-commerce fulfillment

There have been a handful of interesting article over the last week or two about online fulfillment centers. The first is from Bloomberg (Amazon’s Robot War Is Spreading, Apr 5) and discusses how many firms have followed Amazon’s lead and have invested in robots to help run their fulfillment centers. (Recall that Amazon bought Kiva, which we have discussed before.)

One interesting point made in the article is that the automation may not be quite what you think. For example, these robots are not reaching and grabbing items from the shelves (form something like that, see here). Rather, they are working with human pickers who load them up and then let the robot carry items from storage shelves to the packing area. That is, the robots takeover time-consuming schlepping so that humans can focus on identifying the right item on the shelves. Check out this video.

Read Full Post »

The benefits of supply chain speed

Here is an interesting graph. It comes from Goldman Sachs by way of Quartz (A new generation of even faster fashion is leaving H&M and Zara in the dust, Apr 6).

It is showing that how sales growth relates to lead time. And while I am obliged to say that correlation is not causation, it seems pretty clear that it is good to be fast; firms with shorter lead times have distinctly higher sales growth.

The focus of the Quartz article is on Boohoo and Asos, two British web-based apparel retailers that target young shoppers. As seen in the chart, their recent performance has been smoking everyone — even Inditex, the parent of Zara. An obvious consideration here is that both Boohoo and Asos are younger, smaller firms so it is easier for them to generate rapid growth than older, larger firms. It also seems that at least Asos has done some things recently to juice its sales that are independent of its operational expertise. For example, the Financial Times reports that they took advantage of a week British pound following the Brexit vote to cut price in international markets (Asos cuts its cloth for growth but leaves less margin for error, Apr 4).

But it is still an interesting question of how a web-based retailer can benefit from its distribution structure to execute fast fashion faster.

Read Full Post »

Managing in-store inventory for omni-channel retailing

As we have posted about before, retailers using store inventory to fulfill on-line orders is a thing. It is also a thing that raises an interesting question: At what level of store inventory should a retailer stop using store inventory to fulfill on-line orders? That is, should everything be available first-come, first served or should some store inventory be held back only for those customer that wander into the store? According to the Chicago Tribune, different chains are following different strategies on this (With Hatchimals scarce, who gets dibs — online shoppers, or those in the store?, Dec 13).

Target ships online orders from 1,000 of its stores, up from 460 last year. To avoid empty shelves, Target will turn off the order pickup or ship-from-store option on some items when a store’s stockpile falls below a certain threshold, said Target spokesman Eddie Baeb. Stores that ship also get extra inventory.

An online customer likely doesn’t care which store or warehouse handles their purchase. The shopper already walking the aisles does. Exactly how many items Target holds back depends on the product and how quickly it typically sells. …

Other retailers, like Toys R Us, don’t try to guess how many items to hold back for in-store customers.

Even on Christmas Eve, the retailer doesn’t bump back online orders to help procrastinating brick-and-mortar holiday shoppers. Purchases, whatever the format, are first-come, first-served, said Toys R Us spokeswoman Jessica Offerjost.

Read Full Post »

The logistics of fulfilling holiday web orders

It’s the end of the year so it is clearly time to see what is up with how retailers are handling holiday logistics. A useful starting point is this graphic from the Wall Street Journal (As Web Sales Spike, Retailers Scramble to Ship From Stores, Dec 1).

 

This shows how Toys R Us fulfills its web orders. And, yes, that says that over 40% of the web sales were fulfilled from stores. (To put that total in perspective, the company’s revenue last year was $11.8 billion.)

Read Full Post »

Surge pricing for storage space at Amazon

There is a good chance that the last time you bought something on Amazon’s website, it wasn’t actually sold by Amazon. It instead came from an independent merchant, and Amazon just handled the logistics of getting the item to you. That arrangement has an implication that I never considered until a recent Wall Street Journal article (Amazon Prods Its Sellers to Free Up Warehouse Space, Nov 4): By inviting in the additional sellers, Amazon is giving up control of just what is in its fulfillment centers. If a merchant wants to sell miscellaneous crap, that is their business. At the same time, however, that miscellanea potentially ties up space that Amazon needs — or at least could use more profitably on other items. This is particularly true as we head into the holiday season when Amazon should reasonably expect business to be booming.

What is a poor e-commerce giant to do?

How about a little surge pricing?

Read Full Post »

When the last mile is on a college campus

Many years ago, I worked as an office assistant in my college dorm. One of the responsibilities was letting residents know if they had received a UPS package. Anything sent via regular mail went to the campus post office where we were all assigned a box. Since UPS couldn’t deliver to the post office box, it came to the dorm where someone would get around to writing out a paper form so the student could be notified. It wasn’t a particularly efficient process, but once you got past the start of the school year, there weren’t too many packages to deal with.

But times have changed! The Daily Campus (of the University of Connecticut) reports that UConn’s mailrooms are getting overwhelmed (Mailroom backup continues as officials search for solutions, Oct 5).

The number of packages received by the university for residential students has been increasing drastically in recent years, in large part due to the rise in online purchases, Assistant Director of Building Services Tracy, told The Daily Campus in an earlier article. As a result, there is a growing need to find long-term solutions beyond hiring more staff.

According to Cree, more than 100 student workers are involved in work related to the residential mailrooms.

In order to fix some immediate problems, plans have been made to modify existing mailrooms in the next few weeks. These alterations are intended to allow packages to be processed much more efficiently. …

In addition to these immediate changes that are to be put in place, UConn is also considering other ways to efficiently deal with the influx of packages and alter the current infrastructure to better reflect the needs of students. …

“We have been discussing the possibility of creating central locations for sorting, but also discussing distribution systems to get the packages delivered sooner,” [executive director of Building Services Logan] Trimble said.

The earlier article referenced above claims that UConn is receiving 3,000 packages per day. The Business Insider article on UConn’s mailroom woes notes that the school has about 31,000 students. Some of those presumably live off campus and have their Amazon sent to their apartments so we are probably looking at over 10% of on-campus students getting packages every day. I am glad my dorm assistant days are behind.

Read Full Post »

Giving everyone an address three words at a time

What’s your address? For most readers of this blog, that is a pretty easy question to answer. You have a street name and a unique number. Throw in a postal code and maybe an apartment number, and you are good to go. For much of the world’s population, however, things aren’t so easy. Whether because they live in rural villages or poorly planned, rapidly growing cities, many people in developing nations don’t have a standard address. This creates a variety of problems. In particular, it cuts them off from many parts of modern commerce. How do you deliver a package to someone who can’t easily write down where they live? Note that this matters for a developing nation. If a country has an under-developed retail market, fostering an e-commerce industry is likely a better solution for many products than building out physical locations — but that cannot happen without some way of locating customers.

Solving this addressing problem is the goal of what3words, a start-up firm profiled in a recent BBC article (Giving everyone in the world an address, Apr 30). Their plan is to match every three-meter-by-three-meter square on the globe with a three-word triplet. Under this scheme, the house I grew up in becomes collapsed.networking.farm — which would only be better if it were collapsed.networking.firm.

The argument is that it is easier to remember three words than, say, a set of random numbers.  The goal then is to come up with words that are simple and unambiguous to use. Here is how their website explains the process.

Each what3words language is powered by a wordlist of 25,000 dictionary words. The wordlists go through multiple automated and human processes before being sorted by an algorithm that takes into account word length, distinctiveness, frequency, and ease of spelling and pronunciation.

Offensive words and homophones (sale & sail) have been removed. Simpler, more common words are allocated to more populated areas and the longest words are used in 3 word addresses in unpopulated areas.

How does this play out in practice?

Read Full Post »

Zulily, delivery times, and operations strategy

About a year ago, we had a post on Zulily and how they managed their order fulfillment. It featured a nifty graphic from the Wall Street Journal showing just how much longer their delivery times were relative to other interet retailers. Now, the Journal has another story — with a spiffy updated graphic — discussing how their delivery times have gotten even worse (Zulily Nips Business Model in the Bud, Mar 23).

Read Full Post »