There’s resource based pricing

Bruce Boston (aka Mr January) has just blogged There’s move from it’s current in-world pricing scheme to what Bruce is calling ‘resource based’ pricing - he’s even included spreadsheets for those that want to tinker with the model.

Forming a link between virtual items and physical resources is an interesting move as it provides a kind of ‘iron standard’ for pricing, though more often than not pricing depends much more on perceived value rather than actual cost.

It going to be fascinating to see what impact this has on the overall economy – though I wonder with worlds like There and SL just how broke the economy has to be for the world to really suffer.

Comments

[Cory Ondrejka]

1.

So, who knows how representative the basket of goods are in Bruce's spreadsheet, but this change is a whopping 30% average drop in wholesale price for the basket of goods and 2% drop in submission price for the basket.

The average wholesale price is now US$1.21 and average submission price is US$6.70. "Holy planned economy, Batman!" (or, perhaps, "wholly planned economy, Batman!"). A custom hoverbike built today (submission price/wholesale of US$8.95/$4.09) is significantly less valuable than last year (US$11.19/$7.55).

I'm also confused by the argument that Bruce makes about the purpose of the There.com economy being to "information/communication and liquidity of talent." Haven't free markets proven quite superior to planned economies?

[Cory Ondrejka]

2.

Sorry, early in the morning here. Last line should read: "Haven't free markets proven quite superior to planned economies in those areas?"

[Edward Castronova]

3.

This makes a lot of sense to me. If I am selling a good, I should pay attention to the cost of providing. A company that 'sells' virtual items is actually rendering a service: they agree to maintain a database record that assigns certain properties and rights to certain human beings - my account has an avatar named 'Mabyn' who is assigned a certain sword, which gives me, through her, certain rights with respect to the code (I can slash through a certain gnoll, for example, and acquire thereby certain other database assignments).

Well, providing that service is costly. For an open world (one that freely pegs its currency to the dollar), the prices of in-world items should ideally be set at the marginal service cost. Marginal cost pricing doesn't work for high fixed-cost products, but even there, pricing should pay attention to some kind of cost (average cost, for example).

It's hard to tell exactly how close Bruce's prices are to a marginal or average cost scheme, item by item, but it's a very sensible idea.

[Hellinar]

4.

I’m not a Netizen of There, so I may have this wrong. The prices Bruce is “fixing” are wholesale prices for Pixels, object storage etc. There.com is the monopoly supplier for items, so the free market doesn’t apply. Surely the free market kicks in on the retail pricing? At that point the market price mechanism does its magic to determine those unknown utility values people place on the virtual shirt etc.

When it comes to pricing a monopoly service such a electric power, there is a long tradition of price fixing of this kind. And that is without absolute measures of the resource costs of production. Seems to me it would be absurd to apply market based pricing to the production cost of virtual shirt. The server knows precisely how many vertices, pixels, data storage bytes etc. go into the production of the shirt. So use it. The market can provide some estimate of the resource cost of object. But when you have a exact resource cost from a server, why not use it?

As I see it, a “planned economy” has much wider applicability in a VW than the familiar world. Many factors that can only be estimated with market mechanisms in the familiar world are precisely known in a VW. Applying markets to estimate demands that are already precisely known doesn’t seem like a rational exercise to me. It used to happen quite a lot in discussions of the Egyptian economy in ATITD. I’m happy to see someone using the server data to good effect.

[Cory Ondrejka]

5.

Hellinar> Surely the free market kicks in on the retail pricing? At that point the market price mechanism does its magic to determine those unknown utility values people place on the virtual shirt etc.

Sure, but objects aren't going to sell for less than their wholesale price, so you're placing an artificial floor. It would be like the government saying "automobiles can't cost less than $10,000." This means that no matter how much parts are commoditized, how innovative the producers are, etc etc, they still can't use price to compete. Yes, they can change their profit margin, but the can't have a sale.

Ted> For an open world (one that freely pegs its currency to the dollar), the prices of in-world items should ideally be set at the marginal service cost.

That is such an interesting assertion. After all, this scheme still doesn't really solve problems related to load ("Everyone with hoverboards shows up at the tiki hut!"); the load characteristics of different client hardware radically changes the underlying assumptions (cost of additional objects versus object complexity versus textures etc); the actual load is a balance between storage, simulation and display that varies greatly based on how often the product is actually in-world as opposed to in the catalog; and, the more markup There.com applies the more artificial the economy becomes. After all, at the current pricing, the submission price of a hoverboard is more than their monthly charge. If that is actually their marginal cost for a hoverboard (hell, if that is the marginal cost within an order of magnitude) I would be surprised.

[bruce boston]

6.

First, let me say thanks to Ren for picking this up, and thank you to those who have posted your thoughts here. We can always use feedback from a number of perspectives.

So let me start here.

Cory> After all, this scheme still doesn't really solve problems related to load ("Everyone with hoverboards shows up at the tiki hut!")

I would agree that this resource based pricing model doesn’t solve the problem that you asserted here. I would also say that it doesn’t try to solve this problem. Why? A) Because we have never had that problem, our residents are smart enough to know that if they all tried to hoverboard in a tiki hut at the same time that they would have a pretty bad experience. B) Even if we did have this problem, I would consider it either an engineering problem (i.e. the system should not allow more hoverboards in an area than the server can support), or a social problem (i.e. why are members trying to crash our servers?). In either case, I don’t see this example as an economic problem, and I certainly wouldn’t set up the economics to try and solve fringe issues. It makes more sense to set economic rules in relation to actual aggregate activity.

Cory> The average wholesale price is now US$1.21 and average submission price is US$6.70. "Holy planned economy, Batman!" (or, perhaps, "wholly planned economy, Batman!"). A custom hoverbike built today (submission price/wholesale of US$8.95/$4.09) is significantly less valuable than last year (US$11.19/$7.55).

So, as Hellinar points out you may be confusing resource costs with market prices. The example you give points out that it has becomes cheaper to produce hoverbikes, not that the market value of that production has decreased.

Ted> For an open world (one that freely pegs its currency to the dollar), the prices of in-world items should ideally be set at the marginal service cost.

While I can’t say much here, I will say that (obviously) we do know the total cost of running our service, and this model really just seeks to explain how those costs are spread out across the community. We could have chosen any number of factors, but ended up choosing a group that we felt best represented the key factors (like art resources, engineering resources, rendering resources, bandwidth, data warehouse, etc.)

Cory> Sure, but objects aren't going to sell for less than their wholesale price, so you're placing an artificial floor. It would be like the government saying "automobiles can't cost less than $10,000." This means that no matter how much parts are commoditized, how innovative the producers are, etc etc, they still can't use price to compete. Yes, they can change their profit margin, but the can't have a sale.

I don’t think I would agree with the basic assertion here. An easy example is that we found that hoverpacks take fewer resources than buggies do. And, if you look at the pricing model this is reflected in the wholesale prices. If, in the future, we (there.com or our members) develop a vehicle that takes fewer resources than a vehicle we currently have, I would also expect that to be reflected by the calculations made in the pricing model.

Cory> Haven't free markets proven quite superior to planned economies in those areas?

I would agree that free markets on a national scale have proven to be more efficient as a general rule. However, on the local levels economic planning at the community level has also proven to be highly valuable (ex. City of Irvine).

Even so, I don’t think either of these examples apply to the current pricing model that we are incorporating. The basic principle behind this pricing model is that a resource should cost the same no matter how it is used. A pixel should cost the same whether it is used to color a hoverbike or a bowling shirt. To date, this was not the case, as we based the wholesale pricing for bowling shirts on the demand for bowling shirts and not the resources that it consumed. Going forward, it should be interesting to see how the community uses resources like pixels to maximize the benefits in the virtual world.

Additionally, as noted above, this now gives us a great tool to communicate the relative costs of running the service and to explain how one change or another in the real world affects the availability of resources in the virtual world.

-bruce

[Hellinar]

7.

Cory> Sure, but objects aren't going to sell for less than their wholesale price, so you're placing an artificial floor. It would be like the government saying "automobiles can't cost less than $10,000." This means that no matter how much parts are commoditized, how innovative the producers are, etc etc, they still can't use price to compete. Yes, they can change their profit margin, but the can't have a sale. <

As far as rendering costs for objects are concerned, I don’t see that it is any more artificial than saying an electric automobile that uses x kilowatt hours per mile will cost y cents per mile to run electricity is z cents per kilowatt hour. If you are renting out electric cars, that gives a floor price for your cents per mile. I don’t see how that is “artificial”.

Given they all use the same electric power, electric car designers compete on the kilowatt hours per mile. Similarly, the number of vertexes an object uses is under the control of the designer, so object designers can compete on that. A good case in point is the current competition between WoW and EQ II. It seems very clear that the Blizzard designers have won the vertex price war there. A lot of people won’t even run EQ II because the vertex budget for its object is beyond their “pocketbook”.

It seems to me though that what you are actually selling to the Player is not a vertex budget per frame, but a frames per second experience. In my own VW I’m hoping to gather enough client statistics to get a good Vertex budget to fps translation. Then players can “spend” their vertex budget on objects they create with some sense of its final impact on frame rates. Though for my gardening world, I hope to cloak such techie measures in more world appropriate terms like “nutrition requirements” and “food supply”. I do see things like vertex budgets as the “real” resource costs of VWs, and not “artificial”. In VWs that rely a lot on Player creation exposing them to buyers and sellers seems to me a step forward.

[Cory Ondrejka]

8.

Bruce> I don’t think I would agree with the basic assertion here.

So, people can sell for less than the wholesale price? Otherwise, which basic assertion are you disagreeing with?

Correct me if I'm wrong, but my understanding is that the "wholesale price" is the amount of the sale that goes to There.com as a way to drain Therebux from the system and to encourage more people to buy Therebux from There.com. So, unless people can sell below wholesale, you have placed a floor on the price on items. In the case of the hoverbike, $4.09 to There.com when someone buys the bike. Yes, someone can charge more in order to get some of those Therebux themselves, but that by definition increases the price.

Your plan is that $4.09 is the marginal cost (aka resource cost) of adding one additional hoverbike to There.com (when someone buys it from the catalog.) That value is clearly way above the marginal cost, which seems like a planned economy to me.

[bruce boston]

9.

Hi Cory,

Cory> "Your plan is that $4.09 is the marginal cost (aka resource cost) of adding one additional hoverbike to There.com (when someone buys it from the catalog.)"

There is a big difference between marginal cost pricing and resource cost pricing. Marginal costs only look at the additional cost of producing the next Nth unit, while resource based pricing seeks to proportionally divide the total costs of all resources across the total production volume. As I have stated a few times, we looked at a wide variety of resource costs and have proportionally divided them across the developer program.

Cory> So, people can sell for less than the wholesale price?

In the auto industry, auto part manufactures sell car parts to auto manufactures at 'wholesale'. The auto manufacture can resell the parts for less than wholesale, but then they would lose money.

In the case of There, our developers are buying the virtual parts and doing some amazing things with them. I guess they could sell them for less than they buy them for, and some people do through donations, but for the most part people have a tendency to sell them above what they pay for them.

Re: Planned Economies

I think there may also be some confusion in what planned economies are and what they are not. You might check here for reference.
http://en.wikipedia.org/wiki/Planned_economy

Planned economies are inefficient at national levels because they seek to plan too many things including; market price, production levels, resource use and allocation.

In There.com, clearly, members are free to buy as many of any product as they want. Developers are also free to produce at any level that they want. Resources cost the same no mater how they are used. Allocation rules are left to the developers; i.e. they can make an item limited or not.

Again, I think where most free market economists get upset at central planning is when central planners try to claim that there are formulas to accurately predict future supply and demand, and then seek to control prices and production levels based on those calculations. They also have a tendency to get upset when governments try to encourage one group or another through the use of economics, for example, by making a resources less expensive for one group or another based on how that group intends to make use of the resource.

On the other hand, I don't know anyone who claims that there is no planning in a free market economy. Again, I will be the first to admit that we are doing some ‘planning’ here (see City of Irvine example). That said, I just don’t think this is the type of planning that most ‘free market’ economists would get upset about. As stated above, the basic principle is that a resource is a resource and it should cost the same no matter how it is used.

-bruce

[Cory Ondrejka]

10.

Bruce> resource based pricing seeks to proportionally divide the total costs of all resources across the total production volume

Sure, this is why the upload cost is more than the wholesale cost, but this doesn't change the fact that since *every* additional hoverbike that is purchased generates $4.09 that means it is a marginal cost, right? Sell one bike, get charged $4.09.

And, yes, parts suppliers charge "wholesale" but they can choose to compete on price or innovate to reduce their prices. Please correct me if I'm misunderstanding, but the price for the hoverbike in spreadsheet is adjustable by you, not by the user. Also, a quick perusal of web sites selling hoverbikes seems to indicate that $4.09 is a fairly large percentage of the final sale price.

I'm sure Ted or someone else knows, what is a representative cost breakdown in the auto industry of parts versus assembly versus design versus margin?

[bruce boston]

11.

Cory> Sure, this is why the upload cost is more than the wholesale cost, but this doesn't change the fact that since *every* additional hoverbike that is purchased generates $4.09 that means it is a marginal cost, right? Sell one bike, get charged $4.09.

The calculation is probably closer to the following example.

If it costs $100k/mo. to keep the world of There.com up and running, including salaries, overhead, servers, bandwidth, etc, etc and there were 20,000 widgets being produced every month, then you could find the average cost per widget by simple division.

Average real world cost per virtual widget produced: $100,000 / 20,000 widgets = $5

Obviously, we have an array of products so the calculation is a little more difficult (see spreadsheet), but that’s the basic concept behind it. Again, it's closer to ‘average cost pricing’ or what I call resource based pricing than marginal cost pricing.

Cory> Also, a quick perusal of web sites selling hoverbikes seems to indicate that $4.09 is a fairly large percentage of the final sale price.

I think you would find better comparables by looking at CafePress, than you would by looking at the Auto Industry. An IP owner can take a design and put it on a hoverbike in There.com or they can put it on a mug at CafePress. A hoverbike will cost you $5 per copy, a mug will cost you $11 per copy. I bet if you did a comparable margin analysis between CafePress developers and There.com developers you would find similar margins.

-bruce

[Espen]

12.

For some reason ;) this seems to remind me of the frequent Second Life vs There discussions that occur in There forums (and possibly Second Life ones too), where apples and oranges are compared as if they are the same thing. For There, the items sold ARE the source of revenue (unless you count the token $5 annual fee) whereas for places like Second Life and most other MMORPGs they are basically fringe benefits you get with your paid membership. As a result, There has to recover more than just the resource cost on the items they sell, but with the trade off that, as a member of the community, you only pay for what you use.

To me, THAT is a free market, unlike the mandatory monthly taxation in so many other worlds. Sure, I might pay more for the things I buy, but I am in complete control of my monthly spending. As a financial analyst in real life, I have always been impressed with There's commitment to free market economics, so it is certainly unusual to see it referred to as a planned economy.

One thing that hasn't really been mentioned much above is that the cost of submission is set to recover the cost of approving the item itself, which is still primarily a manual process to ensure product functionality and TOS/Copyright compliance. This provides the consumers of There with considerable confidence in the durability of the items they buy, compared to worlds where items may disappear later if they violate any of these rules.

As time goes on, and There hopefully expands and starts to see some economies of scale, I would assume that the lower fixed costs would allow the wholesale prices to drop, thus benefiting all the consumers in There.

[magicback]

13.

Bruce,

How often will There revise the prices? Monthly, quarterly, etc.

At my firm, we subscribe globally to a lot of information databases at a fixed yearly cost. As there are many P&L entities, we apportioned the P&L charges based on the apportioned usage for a given period. As the "refresh" rate is low, often there is a "sticker" shock when the usage charges was much greater than the last period.

Having near real-time update on "wholesale" prices would be interesting to track.

[bruce boston]

14.

Hi Magicback,

The current thought is that we won’t be repricing for at least another 4-6 months. Also, we typically like to give 30-90 days notice when any prices are changing so that people are anticipating the change well before they actually think about spending any money.

In the future though, I could see a system of dynamic pricing that allowed micro adjustments to prices as needed. For example, I think it could be fun to have a sale on pixels one month. That said, dynamic pricing is probably more fun for me than our average member, and so it's not very high on the current list of things to develop.

Also, on the tracking side, retail prices have always been much more interesting for me, as they are much more seasonal and measure the changes in the actual market value of all the many products in the world.

-bruce