I read an interesting article in The Lever titled, Everybody Has a Price – And Corporations Know Yours. The article discusses how the massive data collection of our personal shopping and browsing habits, along with machine learning algorithms, has allowed marketing firms to paint an individual profile of everybody. The article conjectures that this might lead to personalized pricing where we are each offered prices for products and services based on what the algorithms think we are willing to pay.
This is something that economists study in school as a thought experiment because individual pricing is the way for companies to extract the most revenue from the public. People who can be convinced they are getting a premium product will pay more, but the folks who aren’t willing to pay full price can still be lured to pay something. This concept is now known as personalized pricing. The idea with personalized pricing is to make more sales and increase overall profits by extracting what each buyer is willing to pay.
Having differing prices has always been around. When I lived in the Caribbean, locals were quietly charged less than tourists at farmers’ markets and bars. Until the industry shifted to listing firm prices for cars online, everybody went to a car dealership knowing they had to haggle to buy a new car. Countless businesses give discounts to selected customers but not others.
The first widespread use of varied prices was introduced by the airlines. On the typical flight, people are paying dozens of different prices. The airlines have sophisticated algorithms that constantly change the price of the next seat based on the specific seats that are left to be sold and the time remaining until the flight date. This pricing has increased profits and has allowed airlines to get more people on each flight. This pricing model has been expanded to other industries and is used for concert tickets, seats at sporting events, hotel rooms, and other common commodities that don’t have standard publicly known prices.
There is evidence that this has been tried in broadband. In 2022, the California Community Foundation conducted research that looked at the online prices offered by Charter in various neighborhoods in Los Angeles. They were surprised to find widespread price discrimination where people on streets with higher incomes were offered lower prices than people only a block or two away where housing prices and incomes were lower. Charter not only offered different prices, but it offered different broadband speeds to different neighborhoods. Before this study, I assumed that online prices offered by a big ISP like Charter would be the same nationwide. However, at least in Los Angeles, prices were pinpointed based on address and presumed demographics and the perceived willingness of folks to pay different prices.
Many ISPs would never consider this. A lot of ISPs are sold on the idea of one price for everybody. It’s a major challenge for customer service reps to talk to customers about pricing when everybody is being offered different prices based on where they live or some other factors in their profile.
Personalized pricing works best when done online, and that probably means it fits best with a big company like Charter and not smaller ISPs. Any ISP considering this has to be aware of the new FCC rules against discrimination since personalized pricing runs headlong into those rules. But this is an interesting concept, and if personalized pricing is shown to increase ISP profits, it seems likely we’ll see more of it, at least from big ISPs.
What many people fail to take into account is that the cost to do business in different places is often wildly different.
Newer neighborhoods with more recent builds tend to be high income areas and the cost to deliver 300Mbps is lower for a cableco than it is a mile away in an older neighborhood. The quality and condition of the wires and the age of the plant in the given areas should make this obvious. And investment is going to go for new business before old business because it’s more profitable.
There are some unfortunate truths that businesses have to take into account as well. Low income areas are going to have far more late and non-payers. More efforts to collect equipment and bills. More write downs of uncollectable invoices. There are more CSR hours spent on those neighborhoods. More truck rolls. More calls for bad service because ‘cheap’ wifi was bought at walmart. The cost to provide services goes up as the income brackets go down for many services. Lower income areas tend to also have dramatically more use. I’m an operator, I see this data live, this isn’t presumption.
As a person running a business, it’s very difficult to choose not to build into a new (read higher income bracket) neighborhood with the cashflow I have because there’s an older (read poorer) neighborhood that needs service. That costs me money and profits. If I were to go to that poor neighborhood first, I might not actually be able to afford the new one. If I build the new one first, I can afford to come back and build the old one a bit later. It doesn’t seem fair, but economics is rarely ‘fair’ and that’s just how the spreadsheet presents.
There are so many factors that make newer neighborhoods must more cost effective to build into, with customers that are much more likely to pay bills, and that means there’s more competative pressure to keep the cost down. Poor neighborhoods that are older, more difficult to build into, and far less interesting so get less competition.
In Economics class, I believe they called that third-degree price discrimination…