David N. Welton
In this article, I claim that the market for web hosting services is a market for lemons. What, exactly, is a "market for lemons"? The term comes from a paper by economist George Akerlof dealing with information asymmetry. If that's a dry sounding term, the concept is actually simple: in some markets, the seller knows more than the buyer does, or vice-versa. Consider the classic example of buying a used car - you can't tell just by looking at it whether it's worth what the seller is trying to get for it. Shady dealers will try and sell low-quality products at a higher price, and sellers with high-quality products will face buyers who are less willing to pay high prices because of the risk they might get a "lemon".
This "adverse selection" makes for a less efficient market, meaning, in the end, that value is lost. Everyone knows that once you drive a car off the lot, for instance, that its resale value just went down far out of proportion to how much it's been used. It also creates problems for buyers, who have to take a risk, rather than simply being able to select the quality of product they want to buy, and paying the corresponding price. Of course, it's impossible to eliminate this disparity of information, but as we shall see later, there are ways of lessening its impact, so that trade does happen, even though it may not be under ideal conditions.
This same principle applies to many markets where information is "asymmetrical" - the seller knows more than the buyer about the quality of what they are selling. This is certainly true in the market for web hosting.
If you put "web hosting", "dedicated servers" and similar terms in Google, the search engine will come back with a huge quantity of pages advertising these services. Clicking on most of them will reveal a fancy looking page extolling the offerings of the provider. They give you a very good idea of what type of system with exactly what characteristics you can get for a particular price. However, what you, the consumer, can't really know is the quality of service offered. It's easy to write "24x7", "99.999%", and other such terms, but what really happens when you can't ping your machine at 1 AM Sunday morning? How fast will they respond? Is the technician knowledgeable enough to fix the problem if it's something that you can't do yourself remotely, or will they just suggest that on Monday, you can have a reinstall done for free. Many "web hosting companies" are just some guy reselling space from a big data center, others are serious professionals that know their stuff and can help you out in a pinch. This information is much more difficult to obtain - at best, you might find some comments in a forum from people using aliases like S0M3D00D saying "these guys are ok", or "these guys suck", but that's not very sound information to base a business decision on.
Of course, you can always shell out a lot of cash and go with one of the big outfits, but often the prices are far higher for the same hardware. That might be worth it if you are a big company that can afford the price, and is willing to do anything for some peace of mind. But if you're willing to do most of the work of running a system yourself, and are looking for a deal, things get trickier.
A year ago, I was looking around for someplace to place my domains. I wanted a dedicated system, so that I could manage the machine myself. After having been slashdotted a couple of times, I also didn't want to run on a 'virtual' system where the hosting provider might be cramming so many virtual servers on one machine that a bit of traffic makes them fall over. I don't need much in the way of assistance - just a minimal Debian install, and a guarantee that techs are on call to deal with anything that can't be handled over ssh, such as the occasional reboot, or hardware problems.
The first outfit I found seemed ok. They responded to my email enquiries quickly and professionally, and after receiving payment, got a server on-line for me very quickly. Nice! The problems started when the machine's disk started to go a few weeks later. It happened on a weekend, and apparently no one was around that was able to fix it, because the machine was not back on line until the following Monday. No email and all my websites down for a weekend... very stressful! After another misstep on their part, I decided that I'd signed up with a Lemon, and went looking for another provider.
The second place I found is my current hosting provider, layeredtech.com. They provide the same basic services on generic hardware that's ok for what I need, but the real difference is to be found when problems arise. They've always been on it right away, even in the middle of the night their time (I'm in Italy, they're in Texas).
As another example, a company I did some work for offered hosting services on their web pages. They were a real company, not just some guy with a phone number. However, they had no people or plans in place to deal with nighttime or weekend emergencies. And yet they charged more or less the same prices as other hosting providers.
As can plainly be seen, though, it is difficult or impossible to determine the quality of service prior to signing on. Same providers may wave a "service level agreement" or guarantee at you, but realistically, if they don't live up to it, your only option may be to find another place, if they simply say they aren't going to refund your money.
While discussing this article with some economists of my acquaintance, they leaned towards another explanation of problems people encounter with web hosting services, something economists call "moral hazard" - briefly explained as one party choosing to live up to their obligations, or not, after payment of the service. I do not believe this is the case, because it appears that some hosting providers simply are not able to provide the same quality of service as others. They are not equipped to do operate 24/7, and respond to problems in a timely fashion, so it is not necessarily a question of "getting lazy" once they have your money, but of hoping that things generally go well, which may be enough for some of their clients. If the price to the consumer for the service is the same, the provider spends less on personnel, and only loses a few clients who are unhappy with the service, while acquiring new clients at a rate more or less equivalent to other, higher quality providers, there is certainly a profit motive for this behavior. Whereas, if you have the capability to provide good service (you are already paying employee salaries), why would you not put that to use? Not doing so would lead to unsatisfied clients, and you would still be out the money spent on maintaining a 24/7 capable facility.
How does one mitigate the effect of a market for lemons? The best way is to eliminate the disparity of information. If it were known that one provider cannot respond to emergencies during the night, and another can, you would expect their prices to vary accordingly, which would be good for consumers - those wishing to save some money, and who are ok with the fact that their site(s) might be down for hours at a time in the worst case, would gravitate to cheaper services, and those needing more reliability would know that they were getting it when paying a bit more.
How to shine some light on the situation then? Currently, the combination of various forums dedicated to the topic, plus the ability to search with Google, provides an easy way to gather information - although it's not a very effective method. It is easy to post false information on forums - a sneaky 'lemon' provider could say nasty things about a good provider, or he could write off accusations against his own operation as the work of the competition, casting himself as the victim. In general, "reputation" is one way of getting around the problem of information asymmetry, and indeed, in the hosting market if you are willing to shell out a lot of money, you can go with one of the big, well known providers. You will be relatively safe there, but if your budget is tight, it might hurt to pay out that much more money.
Guarantees - a service level agreement in this case, are another way to deal with the problem. However as mentioned above, not much money is involved, so you're not going to get your money back if the provider decides they don't want to give you a refund.
For other, more mature markets, a good solution is often to utilize a neutral third party who can verify the quality of a given product. If you buy a used car without a guarantee, for example, it's usually a good idea to have a trusted mechanic check it out. The money spent is well worth it if it saves you hundreds or thousands of dollars in repair fees later on down the road. Unfortunately, I am not aware of any monitoring services that can tell you much information about the quality of a hosting provider's service - unlike a car, it's something you only see with time, and when problems come up.
In conclusion, we find that there really aren't any good solutions to the lemons problem in the web hosting market at the moment.