23 comments

[ 3.5 ms ] story [ 59.2 ms ] thread
Looks like they use a JS library, so users will be able to detect the usage of this service in most cases.

Few will do so, but those that realize it will be your most vocal detractors. Consumers despise dynamic pricing.

Dynamic pricing means the company is trying as hard as possible to give you the worst deal possible, and people resent it because it's insulting.

Basically, instead of "thank you", it's "fuck you".

It could work the other way. How many times have you seen a product or service found it interesting and saw the price and was immediately turned off? Or if you're shopping for a product and you could convey that information to the retailer, they might lower their price. This is the same idea as searching for coupon codes to put in during checkout.
I’d like to second this comment. The only reason you can fly economy for so cheap is because first class is so expensive.
First class and economy are different experiences. Airlines don’t change your price for economy based on your personal information.

Would you like to have to pay more because you live in an affluent zip code, even if you aren’t affluent yourself? Because of your age, relationship status, income, wealth, education, friends, and more?

There are plenty of budget airlines that only offer economy
They have both a js snippet you can include and call their api (I guess it's intended for websites with fewer ressources), and then a backend integration that would make it way harder for detection.

I'm guessing this qualifies to be added to one of uBlock lists.

For the consumer products even without a JS library it won't take long for people to realize the pricing is dynamic and the pushback will follow.

For the enterprise products it will take longer, but it will happen too. Plus people there have thicker skins, they've seen worse shenanigans and if they really need the product, they will just call the sales directly and negotiate.

> We process the IP address and user agent for each new visitor to your website, then we predict the maximum price level where each visitor is willing to buy.

Whoa. I think if my customers found out about this I would have a real problem on my hands.

And since my customers are all IT admins, it seems like they might figure it out.

Customers are much more tolerant of there being a fixed high price, and some customers being offered a voucher code for 'loyalty' or whatever.

People are fine with being given a discount, but don't like to be charged a premium.

Industries like car insurance have done differential pricing for decades, and consumers have come to accept it.

In my opinion they accept it because they are forced to. I live in an area where the public transit is not built out and my economic opportunities would be drastically limited by not having a car so once a month I unhappily pay more than my peers because of some teenage mistakes.
Criteria for selection includes but is not limited to:

Traffic: > 100K uniques/month

Test period: 6 months minimum

Existing pricing strategy: static

Can anyone shed some light on whether using a service like this may be a legal concern? I recently read that at least in the EU dynamic/personalized pricing can be illegal if it discriminates groups based on certain criteria.
IANAL but have worked a lot with GDPR. If the pricing is based on, for example, user browsing habits then that would constitute usage of personal data. At a minimum this would force the company to disclose that this kind of processing is going on in a clear way to users (if the company claims it has a legitimate interest to do so). The other case would be that explicit consent would be needed and I guess no one would be consenting to that. With the upcoming ePrivacy regulations I suspect many of these legitimate interest cases will also become illegal.
Yes, this is what economists simply call price discrimination and for any normal business it is definitly illegal.

This is a grey zone now, but it will be decided that either it is illegal now, or it will be made illegal eventually.

So I'd be careful with this kind of stuff.

It's pervasive, and nobody's made an effective case that it's illegal outside of basing it on protected classes of people.

For instance, you can detect whether someone browsing your website on their phone is in your physical store and if so, show them the same higher prices that are in the store. But if someone is not in your store, you can show them lower prices to compete online. I've noticed retailers doing this. At least sometimes, they try to defuse it by matching the lower price if a customer notices the game.

I would like to see a system able to give me the derivative of price tolerance vs various factors.

For example, browser type, advert viewed, various on-site AB tests, etc.

It could collect all this data via two prices (eg. $12.99 and $13.99) and looking at the conversion rate for each.

I could totally collect this data myself, but I would like someone else to build a system to measure it, and present it in graphs with confidence intervals.

Pretty much this service, but in 'advise only' mode.

Discriminatory pricing. What could possibly go wrong?
I believe in dynamic pricing based on supply and demand for fixed assets, but using it for zero marginal cost goods like software is insulting.
This seems like it's pretty unethical. I'm not a fan. Surge pricing only really (sort of) makes sense to me for resources that can be temporarily at capacity, as a means of prioritization of customers.

Even that is kind of scummy.

Here is the wikipedia article on the theoretical basis: https://en.wikipedia.org/wiki/Price_discrimination

Many forms of price discrimination, like menu pricing or 2-part tariffs, as well as group pricing, are used every day. This, however, is an attempt at first-order price discrimination enabled by near-perfect tracking of users. It is then unclear to what degree this may or not be illegal, or will be in the future.

Important to keep in mind is that it "hurts" the consumers in the sense they can no longer get a theoretical surplus. Therefore consumers would have to change their mindset about purchasing things. Goods would have to be bought more with a mindset of filling "necessities", because on average we, the consumers, would lose surplus compared to firms.

So naturally, this has the potential to piss off consumers, either if they have the intuition, or understand what price discrimination does.

Edit: As for the legality, I can give the following advice. First, be very careful about how you classify consumers, because many group-based discriminations (such as gender) that are implicitly used by ML algorithms, are definitly illegal.

Second, price discrimination in the "normal world" often points to having monopoly power. Currently, there is no good eqivalent for the internet world, because many monopoly tests fail by default (for example, entry). That is why I believe the EU or the US will eventually legislate against this form of personalized pricing to protect consumers.

It is a grey zone now because, for example, the EU legislates that illegality necessitates a) a monopolistic position and b) using it to hurt consumer surplus or block competition. For internet companies, with some big platforms being the exception, the test for a) fails by default.

This differs from country to country, but because the algorithms can only infer purchasing power based on characteristica, I believe that they will be made illegal (in the EU) based on the respective constitutional laws that make discrimination based on these properties illegal, and I do not think monopolistic postitions will be necessary.

That means even a small internet start-up should be careful.

"Second, price discrimination in the "normal world" often points to having monopoly power"

Something I think is underappreciated, or at least I don't notice people mentioning explicitly, is that the state of being a monopoly intensifies the shorter the time horizon for making a deal. If you want something within the next 100 years, nothing is a monopoly. If you have an online shopping basket that you're about to order, even the most abundant commodity has a privileged position relative to the competition.

Companies are increasingly realizing they can utilize discriminatory pricing in proportion to the amount of commitment demonstrated by a customer. For instance, if you are in a store, you will likely pay a significant premium rather than go to another store, even if they were on equal footing when you were at home. If you are checking out, you will likely pay a premium rather than start all over. If you have a membership... and so on.