The economics of exponential growth
We come now to one of the biggest issues about Linden Lab’s business model, which, not surprisingly, shapes the way residents also set up their own businesses. And you will quickly understand why things all of a sudden look so desperate!
Why have people been worrying about exponential growth since the dawn of SL? Why is Tateru Nino still worried about the low relative growth per month?
To understand this, we’ll have to take a look at two different business models that everybody will understand, and contrast it to Linden Lab’s own business model. Let’s see cars and deskjet printers.
Cars are expensive, luxury items. There is a high cost in manufacturing them, many pieces of the cars are processed by Centerless Grinding and there is a high risk of managing a car stand with Honest Joe trying to get you a good deal. Honest Joe has some running costs — renting the stand’s office space, managing bureaucracy, paying taxes, and so forth. He also needs to pay for the car upfront out of his pocket, and sells it to the client, hopefully with a margin.
Once the client goes out of his stand, however, he’ll not make any more money out of that client — until possibly they come back after a few years for a new car. So there is no much point in establishing a long-term client relationship (which explains why Honest Joe can be anything but honest — the probability of the client returning to make a difference to his sales is very low, and not worth considering). Many stands also have garages, and the client might come back to change the oil or whatever goes behind doors when you take your car back to get an overhaul. But although the service provided there is highly profitable (spare parts are cheap compared to the labour charged to the client), Honest Joe will make far more money selling cars than servicing them.
So Honest Joe is mostly focused on getting new clients, all the time. This is his core business: establish a reputation and make sure that enough new clients come to his stand and regularly buy cars so that he can have a good income on average. If he fails to attract new clients, he cannot wait for the old, faithful ones to come back “soon” — they’ll take years, not months — and so they cannot rely upon them to cover his expenses. Similarly, although providing car service is highly profitable, he cannot make a living from it. It might help to pay a phone bill or two, but his garage mechanics expect payment too. So his risk is to fail to attract new clients.
Now, all Honest Joes in the neighbourhood aggressively compete in the same business, and they have access to a limited, not-expanding customer base. What this mostly means is that car sales are local — on average, you aren’t going to travel far to buy a new car. It’s not something you’ll order via eBay (although you can certainly do that!). So it means that in a community — a town, a city, a county — there is a limit to how many cars you can sell per month. You can figure out averages based on how many people live there, how much money they earn, and how often they replace their cars. Accounting for unpredictable issues (like gas prices rising dramatically, new green legislation, or overall recession), and looking how many competitors are already in an area, a car sales manager can pretty much estimate how many cars they can possibly sell in a month or in a year. Census estimates can also give you an idea on how many new families are arriving — or leaving — your community every year. Possibly this is, say, something between 1-10% per year, not more. New families mean new potential clients. But by far and large, the existing community will be the primary target for sales.
Suppose that all of a sudden everybody is opening up their car stands, tired of Honest Joe’s incompetence, arrogance, and blatant lies. Every tiny little plot of land is used up by a car dealer. What happens?
Since the number of potential clients doesn’t grow, and unless suddenly everybody in the neighbourhood wins the lottery, the number of total cars sold in a month will, on average, not rise. Instead, each dealer will sell less cars, as clients go to the competition instead. Honest Joe, after living on a nice income by being one of the few dealers in town, is suddenly surrounded by aggressive competitors. Suddenly people stop buying his cars. He has to shut down his stand and go away (perhaps to another neighbourhood still not infested with car dealers). The alternative — that suddenly there is a massive influx of new families needing cars — is simply not realistic. He would have a choice at first: dropping prices to see if he could attract more clients that way. But there is a limit on how much he can play around with the price (its elasticity), since, after all, a car costs a certain amount when it comes out of the factory, and you cannot go lower than that.
This is, strictly speaking, basic economics. When supply is far above demand (ie. too many car dealers for the same number of families in the neighbourhood), and the price cannot be changed much (its’s not elastic), suppliers will fail to get clients, and they will go broke and close shop — or be forced to move to other markets instead.
Nothing special here — we all learned this at school 🙂
Let’s see how deskjet printers work. Their business model is quite more curious. A deskjet printer can be sold for, say, US$30 or 40 these days. There is no way this is the real cost for manufacturing it, much less selling it for a profit after a long chain of intermediaries. But that’s no problem for the deskjet printer company. They’re actually not in the business of selling printers. No, what they sell is ink!
Ink is insanely cheap to produce. I’ve read some fascinating numbers — like “a few cents” for tons of ink. But they are quite expensive when packaged as printer cartridges. In fact, as everyone knows, after printing a few thousand pages with your printer, you have spent way more in ink than on your printer! Packaged printer ink is actually one of the most expensive items on the planet, if you take a look at its price-per-weight-unit. It’s more expensive than gold, diamonds, or anything else you can care to mention.
So the industry has moved very cleverly from selling expensive printers (with little margin) and keeping ink costs low to a different model, where printers are pretty much dumped on the market (sold at manufacturing costs or even below), but selling inexpensive ink at insanely huge profit margins is their core business.
What is so good about this model? Once you’ve grabbed a client for your brand, they will continue to provide you with a regular income for the rest of their lives. They will always need to buy more ink, even if they never come back for a new printer (and to force them to do so, it’s easy: launch a new model which uses an incompatible cartridge, so they’ll have to come back when they can’t find any more ink cartridges for their old printer…). So even if you exhaust the market (ie. there are no more people around to buy more new printers), you’ll still get a revenue from them as they continue to buy ink. More fascinating is the concept that even on a saturated market (ie. one where you’re not making any more sales), your income actually grows from continued ink sales (since the profit margins are staggeringly higher) in that same area. Put into other words, if HP could pull it off, they would just sell ink and never launch a new printer in the market, because you can always sell more ink to the same people, while printers require innovation, research, concept, design, production, manufacturing, distribution, marketing, and promotion. Ink is just ink.
Using the car analogy, Honest Joe would make much more of a profit if, in a saturated market, he’d just sell gas (switching his core business) instead of new cars. Alas, gas has also relatively low margins, unlike ink 🙂 But it would be a much better business for him in a saturated market. Also, gas doesn’t require promotion and marketing and a sales pitch: to use a car you need gas, like you need ink to use a printer.
Both products and business models actually enjoy a similarity. They’re not dependent on how the market grows or expands. This basically mean that Honest Joe can continue his sales for all his life, assuming nothing drastic happens to the neighbourhood he lives in. It’ll grow slowly over time — a few percent per year or decade — but Honest Joe will not lose money because of that. Of course, if suddenly everybody starts moving into your neighbourhood, Honest Joe will have an advantage: he’ll sell more cars. If suddenly everybody starts opening their car stands to compete with Honest Joe, while the population grows exponentially, everybody will get new customers and make profits.
So during a period of exponential growth, the market grows, thus allowing more people to supply that market — enjoying the comforts of expanding demand (more clients willing to buy cars). Exponential growth can be hard to manage (Honest Joe might need to hire more car dealers if he can’t take so many calls in a day…) and push companies to extreme limits (as management becomes a mess), but it also means that “everybody enjoys the economy boost”.
In the real world, however, this hardly happens. Neighbourhoods don’t grow exponentially, at least not since the 1850s in the Wild West or Australia.
Except, of course, for the Internet.
Linden Lab’s business model is one of the high-risk ones. It’s not because of competition (there isn’t really any other virtual world with user-generated content), and not even because of technical glitches (SL has improved so much since 2004 that it could be a different product). It’s not because of bad press, or people grumbling all the time. All those factors might have just limited influence in the business by itself.
Instead, it’s because LL’s business model relies on exponential growth. Now, RL businesses never ever do that for two reasons: first, because exponential growth outside the Internet is very rare. Secondly because once the exponential growth phase is over — and normal growth steps in — the business model is not sustainable. Almost all dot-com burnouts experienced this in the flesh. They only made money while their product/service was growing from zero users to a few millions. Once it stagnated in attracting new users — they still had millions, they just didn’t get dozens of millions — their model was suddenly not working any more and they had either to change it and adapt, or simply abandon it.
What is a typical business model that is based on exponential growth? Well, almost everything that relies on startup fees to pay all running costs — while not worrying about monthly fees charged to customers (those are “niceties” to get you an extra income, but not your core business). These models forfeit long-term recurring fees (like Honest Joe, who doesn’t care about car service and repair — he still provides it, but it’s not his major source of income), keeping them low and attractive, and place their focus in attracting as many new clients as possible. New clients, like with car dealers, is all what counts.
The difference between selling cars and LL’s business model is, however, that Honest Joe can live by selling cars at a certain rate, to a population that never grows much, for years and years. He doesn’t rely on more new clients per month to make a living. All he needs is the same number of clients per month. And if you have a constant population (a fixed market), these are easy to figure out.
LL’s business model is quite different. They sell server space mostly. These days, their income comes mostly from selling new regions, not tier fees. This means that they need to attract new clients every month to buy new regions. Since SL’s landmass grows — and so does the total number of residents — LL has more running costs every month. All these residents — all 15 millions of them — don’t pay anything. Just a hundred thousand pay something. So what LL needs is to keep selling more and more regions every month, to keep up with the millions and millions that continue to grow — half a million per month — but which don’t pay anything.
This is, as history shows, the problem with the business model. It’s unimportant if SL is growing more or less over time (it’s actually growing slightly more, with some exceptions during the summer months). The question is if it expands fast enough to provide LL with new orders for private islands. And every month there have to be more orders. Basically this means that LL’s business model relies on increased sales every month to keep itself afloat. So far, they managed to survive, for far longer than most dot-com businesses in the past. And they even managed to lower prices!
You’ll also see a peculiarity of this model, which will make you suddenly realise that although we have such a high churn rate of residents (ie. a high number of people dropping out of SL), LL is not really so worried about it. LL gets a constant income from the 30,000-region-grid — from tier. But if someone goes away and leaves their land to be claimed back by Governor Linden, LL makes a profit. The best kind of user population is one where people are constantly leaving land for LL to claim back and resell as new. For LL, it’s far better to have 100 new clients every month ordering islands from them, and going away next month, and getting 100 new clients, than having 100 faithful clients paying tier every day. It’s obvious: LL makes far more from new sales than from tier. Once they have the hardware in place, setup fees are pure profit, while tier just has a slight margin due to the recurring costs. Thus, for LL, a model where they get half a million new users, but half a million old users leave SL every month, is not important. What counts is how many new islands they sell per month — and making sure that there are not more people leaving than joining SL (taking a look at the statistics, this is not likely to happen).
And here is when things start to get confusing — since everything seems to be holistically connected to everything else, and it’s hard to get the whole picture.