Hi,

Please excuse the long gap between newsletters...there have been a few things going on recently. I do want to briefly discuss the twin crises of our day.

I hope you, your family, friends, and broader communities are getting along well during the COVID crisis.

I am dumbfounded that wearing a mask during a global pandemic has, for many, become a political tribal marker instead of a public health decision. We need to find our way back to establishing agreed-upon facts, determining who has expertise, and finding the best solutions.

With respect to the protests in the aftermath of George Floyd's death, I recommend reading Ta-Nehisi Coates' book Between the World and Me. Short and sweet, it's beautifully written, like an updated James Baldwin. Coates speaks about the truth of America and race relations with an undeniable honesty. And the truth is just as horrible and compelling as when Baldwin spoke it 50 years ago. In the 1960s America was moved by the evident injustice to aspire to do better and enact legislation in that direction. We should be so inspired again today.

Hopefully we can catch up soon, if not in person then by phone or the ubiquitous Zoom.

Friends & Family Spreadsheet Help Offer

I regularly get calls from folks who are stuck on some Excel problem. They range from very basic to quite advanced, and I enjoy helping out. I’d like to extend that offer to you, since you subscribe to the newsletter. I like hearing from people randomly and I find the issues to be like a little puzzle I get to solve before getting back to work.

Eight years ago I wrote about QuiBids.com, a so-called “penny auction” website, which I thought of as “flim-flam in its purest form.” The Washington Post went further, calling it “the evil stepchild of game theory and behavioral economics.” It’s no wonder it makes for a delightful exploration.

It turns out that I can check my modeled estimates, because Quibids drew academic attention. The fine folks in Computer Science at Brown scraped data and performed an analysis (link to paper; link to final cite) with surprising results.

I had thought that the gambling-like quality of the game led people to make bad decisions and give the company a high profit margin. It turns out that the core game may be close to zero margin. The company makes its money from meta auctions where they sell bids, which cost them nothing to produce.

We'll explore the details behind this. But first, let's review the basics of QuiBids.

The Basics:

Brilliant Propaganda

It’s surprisingly difficult to describe QuiBids succinctly without using their propaganda terms. Let’s try it their way first and then break it down. QuiBids runs “auctions” on their website, e.g. a $500 TV. The “bidding” starts at $0.01, and, similar to eBay, there is a “clock” which announces the time until the “auction” ends. When the “clock” hits 0:00, the winning “bidder” gets to purchase the item at the final price. The appeal is that the TV may sell for $25.00 at the end of the “auction.”

Obviously, there’s a catch, right? Actually there are two.

Catch #1

What you realize when you sign up for an account is that “bidding” is, unlike a typical auction, not free. In fact, each “bid” costs $0.40 (but was $0.60 back in the Brown study days). So when you get an account, the first product the company wants to sell you is a package of 60 bids for $24.

Catch #2

Another unusual aspect is how the “clock” works. Unlike eBay, where auctions end at a specified time, on QuiBids the timer gains an additional five seconds upon each new “bid.” When you watch an “auction” you’ll see it cycling between one and fifteen seconds for long periods of time as the “highest bid” increments.

The Brown paper, omitting the clock factor, describes it as “a form of an ascending auction in which, in addition to the winner paying its bid to acquire the good up for auction, each bidder pays a fixed cost for each bid it places in the auction.”

Profitability of Basic Auctions

My original analysis modeled the business as having an 80% gross margin. That turns out to be overblown, as Brown found that basic auctions have a 47% gross margin. Further, the variance between auctions is interesting. The median auction loses money. Quibids makes 43% of its gross profit on only 2.5% of all auctions. In general,the higher value the item being auctioned, the higher the margin, so the art of running this site includes knowing what high-value items bidders will find most attractive.

Impact of the “Buy Now” Feature:

From my 2012 newsletter:

An interesting twist is that the losers also get an option, but a less valuable one. If an entrant spends $60 on “bids” to try to win the $500 TV, he or she is permitted to spend $500-$60=$440 to buy the TV. Thus, the company argues, you can’t lose because you never need spend more than retail but have a chance to get it for far less.

I thought Buy Now would increase gross profit by having higher revenue, albeit at a lower margin. The Brown study found surprising results. First, I had assumed the markup over product cost was 67%; Brown found it was 21% above retail. They assume Quibids buys their products at retail, but I consider that an absurd assumption, so I’ll continue to assert that Quibids’ margins are higher than Brown calculates them.

But the real surprise about Buy Now was that if buyers behave rationally (bid only if they want the product at the retail price and exercise the Buy Now option if the value of their bids exceeds the markup over retail) that Quibids’ gross margin falls from 47% to -3%. The model below shows how this can be the case.

Customer Retention

Of course, one of the big problems these sites have is that there is regular negative reinforcement. Aside from the QuiBids company itself, only the winner gains; all other bidders experience loss. This led to customers abandoning the site: 86% left quickly. To solve the customer retention problem, QuiBids hit on a variety of solutions, including giving “badges” to customers, e.g. a badge for spending one hour of time on an auction or by placing 100 bids in a single auction. It’s amazing how little positive reinforcement it takes to keep people engaged.

Down the Rabbit Hole: Voucher Bids

One brilliant aspect of Quibids was that 30% of the auctions are not for physical goods, but for packages of bids, called Voucher Bids. Of course, these are free for the company to produce. Since most bids are not spent by the winner, they cost the company nothing. The gross margin of these auctions is 97%. And, here the Buy Now feature works in Quibids’ favor, since they are always willing to sell you more bids, which can only serve to increase customer engagement and still cost nothing to produce.

In the end, almost all of Quibids’ profit comes from Voucher Bid auctions. And if that isn’t meta, what is?