And every real-world example of price discrimination I know of works like this.
That might be availability bias, though. If you deliberately try to not waste money, you’re probably avoiding many of the ways that you can price discriminate up. Software as a service has many examples of price discrimination the other direction; here’s Patrick MacKenzieon the subject. (It’s a recurring subject, there might be a better place to jump in, but that’ll get the idea across.) Many people in companies are willing to pay 10X in order to say they’re getting the Enterprise version of software instead of the Hobby version of that software, even if the hobby version covers their use case entirely, for signalling reasons.
Software as a service has many examples of price discrimination the other direction
As I said, if you charge somebody more for something, there has to be some added value. Making a deluxe or enterprise version of something where the customer freely chooses to pay more for the extra goodies isn’t price discrimination, as I understand the term. That’s just “having a range of products”.
To be price discrimination, you need to be offering something that’s substantially the same, but you want to get more business by not selling it at just one price. Airlines, coupons, etc., as I mentioned above.
You can quibble whether this is the “true” definition of price discrimination, but at that point we’re arguing the meaning of words. This is what I mean by price discrimination in this discussion. I consider raising prices without adding extra value to be either “monopoly price gouging” or “correcting your dumb mistake of not charging enough in the first place”. ;-)
Many people in companies are willing to pay 10X in order to say they’re getting the Enterprise version of software instead of the Hobby version of that software, even if the hobby version covers their use case entirely, for signalling reasons.
Enterprise versions of software generally come with support contracts or SLAs, not to mention features that require more support (such as API access, single sign-on, integrations, etc.) that justify the higher price because they come with higher ongoing costs for the service provider.
Even if the software itself is otherwise identical, an enterprise offering is different, speaking from having personally been on both sides of that particular business. Even if most of the “enterprise” users don’t actually end up needing that extra support on a regular basis, it’s a legitimate added value to have the protection of choosing the more expensive plan that comes with those guarantees. They’re basically paying more for the option to get more help when they need it, or to sue you if something goes wrong.
But even if it were entirely for signaling reasons, that still wouldn’t make it price discrimination. It’s the buyer’s choice whether to take a lower-priced offering, and if the seller adds an enterprise package, it’s usually not because they’re trying to squeeze more money out of customers they already have. They’re doing it because they wouldn’t get the enterprise customers at all with their existing offer. (It’d be price discrimination if they needed some way to prevent the enterprise customers from buying the hobby offering.)
While SaaS and PaaS vendors may also have price-discrimination offerings in their lineups (closeouts, old servers, “free” tiers, etc.), their mainline plans are usually priced so that the vendor is indifferent to which pricing plan you buy: they’re making a decent living even if everybody picks just one plan, whether it’s the lowest or highest priced one.
Again, “price discrimination” as I’m using the term is something used to eke additional revenue out of excess supply without destroying your ability to get your regular price for your regular supply. Merely having a range of products isn’t. Even Wikipedia says:
Price differentiation is distinguished from product differentiation by the more substantial difference in production cost for the differently priced products involved in the latter strategy
This is a bit tricky, because as I explained earlier, “production cost” is a bit of a chimera, due to the question of allocating fixed costs and semi-variable costs. But if you offer an enterprise package with support, you have to pay for some extra support staff capacity just to have them available in the event that this is the month the customer decides to use it. Few of a business’s production costs are truly variable, unless you’re somebody like Uber, who can instantly hire/fire their drivers. (And they still have lots of fixed and semi-variable costs.)
Anyway, the point is that most SaaS/PaaS companies’ pricing tiers are based on varying costs between the tiers, which makes them product differentiation, not price discrimination per Wikipedia.
That being said, the Wikipedia article is actually pretty vague about the distinction at times and notes that only “some” economists argue that signaling premiums make something price discrimination rather than product differentiation. My personal opinion is that perceived value, not “production cost” forms a relatively bright line between price discrimination and product differentiation. If the buyer considers the higher-priced product more valuable, IMO that’s product differentiation. If they consider it the “same thing” (e.g. DVDs, airline flights), it’s price discrimination.
As I said, if you charge somebody more for something, there has to be some added value.
I think this is symmetric, though. The ‘added value’ of paying the non-coupon price is that you get the time you would have spent managing coupons and to conspicuously consume non-coupons; the ‘added value’ of paying the enterprise price is often just the conspicuous consumption of non-hobbyism. But if I want coupons to represent price discrimination, then it makes sense to see the enterprise tier as also price discrimination.
And if we use the ‘capture as much consumer surplus as possible’ definition, both of those are clearly serving the same general function.
Anyway, the point is that most SaaS/PaaS companies’ pricing tiers are based on varying costs between the tiers, which makes them product differentiation, not price discrimination per Wikipedia.
The Kalzumeus response to this is that it leaves immense amounts of money on the table to sell software based on varying cost rather than capturing consumer surplus by price discrimination. When he sells Appointment Reminder, he encourages people to pick a tier based on the cost of one missed appointment rather than based on their particular needs (i.e. the cost of serving them).
When he sells Appointment Reminder, he encourages people to pick a tier based on the cost of one missed appointment rather than based on their particular needs (i.e. the cost of serving them).
It’s insanity to ever ask a buyer to value a product based on your cost; that’s like an elementary principle of business. But there is a difference in cost, in that in the article you linked, the author talked about 25 extra hours of work he put in just to land that enterprise deal. That is a different cost of the offering, and has to be priced in.
if we use the ‘capture as much consumer surplus as possible’ definition, both of those are clearly serving the same general function.
There are LOTS of ways to “capture as much consumer surplus as possible”. Marketing and sales, for example, are the process of communicating the benefits of your product to a customer in such a way as to make them willing to pay more—thereby capturing more of their surplus.
Well-run business are always trying to capture as much consumer surplus as possible, so that’s not a differentiating factor for what makes something price discrimination, IMO. The portion of consumer surplus a business captures is called profit, so if you’re not trying to capture as much of it as you can, you’re probably doing something wrong. ;-)
I think this is symmetric, though. The ‘added value’ of paying the non-coupon price is that you get the time you would have spent managing coupons and to conspicuously consume non-coupons; the ‘added value’ of paying the enterprise price is often just the conspicuous consumption of non-hobbyism. But if I want coupons to represent price discrimination, then it makes sense to see the enterprise tier as also price discrimination.
The difference is that in the case of the coupons and most other discounted offers, the differentiation is acheived by making a worse product: a factory second, something you have to take more time for, stay over Saturday for, etc.
I think what I’m trying to say is that if your higher-priced customers are trying to bypass your tiering to pay a lower price, then you have price discrimination. If your higher-priced customers are happy to pay more, you have product differentiation. This isn’t a perfect bright line, because some customers will always want the better deal. But a good question to differentiate is whether you created your new product by tacking conditions, restrictions, and other annoyances onto a perfectly good product to keep your higher-end people from wanting it, or whether you added premium features onto an existing product to capture people with lower price sensitivity or greater desire to signal their consumption. The latter IMO is product differentiation, the former price discrimination.
There’s also an important point you’re missing about “enterprise” deals. Yes, enterprises do engage in a certain amount of conspicuous consumption… though it’s usually at the multi-milion dollar level. (As the linked article points out, $5K is just a rounding error for an enterprise.) And yes, enterprises can overspend due to it not being the decision maker’s money.
BUT… there is a good reason for that. An enterprise values stability over optimum performance, because it relies on lots and lots of mediocre people rather than a few talented ones. The public enterprise’s “real” product in some sense is an attractive investment for institutional investors, and that requires actions and systems that look stupid on the surface to people who don’t have their money invested in the business (as opposed to say, their time).
It helps to realize that an enterprise is not about making the most money, but making its bosses happy… and the real bosses are the investors, who want their investment to be safe as much as they want it to perform. The downside risk of a bad purchasing choice for a large organization can be immense compared to the upside benefit of the best possible choice. That’s why they do what they do, and it’s ultimately the investors who are paying for that extra insurance.
Is there waste? Sure. But it’s a form of insurance, which is also waste if you never actually need it. The thing is, you don’t know (institutionally) whether you’re gong to need it, so it’s a bad idea to take the risk. Especially since none of the people actually working in the organization get to partake of the upside of a good decision, but will personally be penalized for the bad decision. These factors have zip-all to do with conspicuous consumption. Plenty of employees within a company will happily pay the hobbyist price, but are bound by policies that e.g. require support or SLAs for all software purchases, etc. Or else they have requirements like auditing or whatever it is that trips the “enterprise” flag.
IME, you have to get up to a somewhat higher level of management to have conspicuous consumption as part of turf wars or signaling personal importance, and you’re talking six/seven/eight figure deals there, not four or five. The closer you get to the trenches, the less people care about signaling to other managers vs. getting something to help get themselves through the day.
That might be availability bias, though. If you deliberately try to not waste money, you’re probably avoiding many of the ways that you can price discriminate up. Software as a service has many examples of price discrimination the other direction; here’s Patrick MacKenzie on the subject. (It’s a recurring subject, there might be a better place to jump in, but that’ll get the idea across.) Many people in companies are willing to pay 10X in order to say they’re getting the Enterprise version of software instead of the Hobby version of that software, even if the hobby version covers their use case entirely, for signalling reasons.
As I said, if you charge somebody more for something, there has to be some added value. Making a deluxe or enterprise version of something where the customer freely chooses to pay more for the extra goodies isn’t price discrimination, as I understand the term. That’s just “having a range of products”.
To be price discrimination, you need to be offering something that’s substantially the same, but you want to get more business by not selling it at just one price. Airlines, coupons, etc., as I mentioned above.
You can quibble whether this is the “true” definition of price discrimination, but at that point we’re arguing the meaning of words. This is what I mean by price discrimination in this discussion. I consider raising prices without adding extra value to be either “monopoly price gouging” or “correcting your dumb mistake of not charging enough in the first place”. ;-)
Enterprise versions of software generally come with support contracts or SLAs, not to mention features that require more support (such as API access, single sign-on, integrations, etc.) that justify the higher price because they come with higher ongoing costs for the service provider.
Even if the software itself is otherwise identical, an enterprise offering is different, speaking from having personally been on both sides of that particular business. Even if most of the “enterprise” users don’t actually end up needing that extra support on a regular basis, it’s a legitimate added value to have the protection of choosing the more expensive plan that comes with those guarantees. They’re basically paying more for the option to get more help when they need it, or to sue you if something goes wrong.
But even if it were entirely for signaling reasons, that still wouldn’t make it price discrimination. It’s the buyer’s choice whether to take a lower-priced offering, and if the seller adds an enterprise package, it’s usually not because they’re trying to squeeze more money out of customers they already have. They’re doing it because they wouldn’t get the enterprise customers at all with their existing offer. (It’d be price discrimination if they needed some way to prevent the enterprise customers from buying the hobby offering.)
While SaaS and PaaS vendors may also have price-discrimination offerings in their lineups (closeouts, old servers, “free” tiers, etc.), their mainline plans are usually priced so that the vendor is indifferent to which pricing plan you buy: they’re making a decent living even if everybody picks just one plan, whether it’s the lowest or highest priced one.
Again, “price discrimination” as I’m using the term is something used to eke additional revenue out of excess supply without destroying your ability to get your regular price for your regular supply. Merely having a range of products isn’t. Even Wikipedia says:
This is a bit tricky, because as I explained earlier, “production cost” is a bit of a chimera, due to the question of allocating fixed costs and semi-variable costs. But if you offer an enterprise package with support, you have to pay for some extra support staff capacity just to have them available in the event that this is the month the customer decides to use it. Few of a business’s production costs are truly variable, unless you’re somebody like Uber, who can instantly hire/fire their drivers. (And they still have lots of fixed and semi-variable costs.)
Anyway, the point is that most SaaS/PaaS companies’ pricing tiers are based on varying costs between the tiers, which makes them product differentiation, not price discrimination per Wikipedia.
That being said, the Wikipedia article is actually pretty vague about the distinction at times and notes that only “some” economists argue that signaling premiums make something price discrimination rather than product differentiation. My personal opinion is that perceived value, not “production cost” forms a relatively bright line between price discrimination and product differentiation. If the buyer considers the higher-priced product more valuable, IMO that’s product differentiation. If they consider it the “same thing” (e.g. DVDs, airline flights), it’s price discrimination.
I think this is symmetric, though. The ‘added value’ of paying the non-coupon price is that you get the time you would have spent managing coupons and to conspicuously consume non-coupons; the ‘added value’ of paying the enterprise price is often just the conspicuous consumption of non-hobbyism. But if I want coupons to represent price discrimination, then it makes sense to see the enterprise tier as also price discrimination.
And if we use the ‘capture as much consumer surplus as possible’ definition, both of those are clearly serving the same general function.
The Kalzumeus response to this is that it leaves immense amounts of money on the table to sell software based on varying cost rather than capturing consumer surplus by price discrimination. When he sells Appointment Reminder, he encourages people to pick a tier based on the cost of one missed appointment rather than based on their particular needs (i.e. the cost of serving them).
It’s insanity to ever ask a buyer to value a product based on your cost; that’s like an elementary principle of business. But there is a difference in cost, in that in the article you linked, the author talked about 25 extra hours of work he put in just to land that enterprise deal. That is a different cost of the offering, and has to be priced in.
There are LOTS of ways to “capture as much consumer surplus as possible”. Marketing and sales, for example, are the process of communicating the benefits of your product to a customer in such a way as to make them willing to pay more—thereby capturing more of their surplus.
Well-run business are always trying to capture as much consumer surplus as possible, so that’s not a differentiating factor for what makes something price discrimination, IMO. The portion of consumer surplus a business captures is called profit, so if you’re not trying to capture as much of it as you can, you’re probably doing something wrong. ;-)
The difference is that in the case of the coupons and most other discounted offers, the differentiation is acheived by making a worse product: a factory second, something you have to take more time for, stay over Saturday for, etc.
I think what I’m trying to say is that if your higher-priced customers are trying to bypass your tiering to pay a lower price, then you have price discrimination. If your higher-priced customers are happy to pay more, you have product differentiation. This isn’t a perfect bright line, because some customers will always want the better deal. But a good question to differentiate is whether you created your new product by tacking conditions, restrictions, and other annoyances onto a perfectly good product to keep your higher-end people from wanting it, or whether you added premium features onto an existing product to capture people with lower price sensitivity or greater desire to signal their consumption. The latter IMO is product differentiation, the former price discrimination.
There’s also an important point you’re missing about “enterprise” deals. Yes, enterprises do engage in a certain amount of conspicuous consumption… though it’s usually at the multi-milion dollar level. (As the linked article points out, $5K is just a rounding error for an enterprise.) And yes, enterprises can overspend due to it not being the decision maker’s money.
BUT… there is a good reason for that. An enterprise values stability over optimum performance, because it relies on lots and lots of mediocre people rather than a few talented ones. The public enterprise’s “real” product in some sense is an attractive investment for institutional investors, and that requires actions and systems that look stupid on the surface to people who don’t have their money invested in the business (as opposed to say, their time).
It helps to realize that an enterprise is not about making the most money, but making its bosses happy… and the real bosses are the investors, who want their investment to be safe as much as they want it to perform. The downside risk of a bad purchasing choice for a large organization can be immense compared to the upside benefit of the best possible choice. That’s why they do what they do, and it’s ultimately the investors who are paying for that extra insurance.
Is there waste? Sure. But it’s a form of insurance, which is also waste if you never actually need it. The thing is, you don’t know (institutionally) whether you’re gong to need it, so it’s a bad idea to take the risk. Especially since none of the people actually working in the organization get to partake of the upside of a good decision, but will personally be penalized for the bad decision. These factors have zip-all to do with conspicuous consumption. Plenty of employees within a company will happily pay the hobbyist price, but are bound by policies that e.g. require support or SLAs for all software purchases, etc. Or else they have requirements like auditing or whatever it is that trips the “enterprise” flag.
IME, you have to get up to a somewhat higher level of management to have conspicuous consumption as part of turf wars or signaling personal importance, and you’re talking six/seven/eight figure deals there, not four or five. The closer you get to the trenches, the less people care about signaling to other managers vs. getting something to help get themselves through the day.