52 Bicycles

I’ve been asking myself “what should a subscription price look like relative to a traditional perpetual license?” Here’s the rough sketch I’ve worked out.

First, I will limit this analysis to revenue and the financial costs of a subscription model, so as to get numbers in real terms. The operational costs, and therefore profitability, can be subject of a future discussion.

Key Inputs:

1. Maintentance Revenue: What we expect to earn each year for support and upgrades
2. Churn Rate: What percentage of customers will stop purchasing updates (or renewing the subscription) each year
3. Cost of Capital: What it costs us to borrow/raise money, in terms of an annual interest rate (APR)

So now we need to delve into “lifetime value” of a customer. Staying completely focused on revenue, we could describe the value of a one-product customer as follows:

1. Initial License Revenue + Net Present Value of Maintenance Revenues; or
2. Initial License Revenue + ((Annual Maintenance Revenue) / (Discount Rate)); which is
3. Initial License Revenue + ((Annual Maintenance Revenue) / (Churn Rate + Cost of Capital))

So, if we assume that changing the pricing model does not affect the churn rate or cost of capital, and we want the same lifetime value from a customer, it stands to reason that an “annual” pricing model would be:

1. Lifetime Value = Net Present Value of Annual Price
2. Lifetime Value = (Annual Price) / (Discount Rate) which, as established above, is
3. Lifetime Value = (Annual Price) / (Churn Rate + Cost of Capital), so, arithmetically,
4. Annual Price = (Lifetime Value) * (Churn Rate + Cost of Capital)
5. Annual Price = (Initial License Revenue + ((Annual Maintenance Revenue) / (Churn Rate + Cost of Capital)))*(Churn Rate + Cost of Capital)
6. Annual Price = (Initial License Revenue) * (Churn Rate + Cost of Capital) + (Annual Maintenance Revenue)

Keep in mind, this is normative: it’s looking for a way to interpret the two in the absence of other intervening factors. So it helps to think about this stuff.

Let’s plug in some numbers. Salesforce.com has a base product for $995 / year. If we assume for a minute that they would charge a maintenance/service/upgrade fee equal to 15% of the original purchase price, a churn rate of 25%, and a cost of capital of prime, which is around 8%, this implies:

1. $995 = (Initial License Recenue)*(25% + 8%) + (ILR * 15%)
2. $995 = (Initial License Revenue)*(25% + 8% + 15%) - Turning it around,
3. Initial License Revenue = $995 / (25% + 8% + 15%) so
4. Initial License Revenue ~ $2,000.

Implications:

1. The financial risk doesn’t appear to be the key driver. Moving the cost of capital up or down a few percentage points doesn’t drive the math all that much.
2. In contrast, the risk associated with churn is huge. Even this math assumes an average customer stay of four years . Since SAAS actually cuts switching costs, it’s reasonable to think that four years is a long time to keep a customer when the environment inevitably turns competitive - the likelihood of losing a customer in the first few years to a competitor - either direct, substitute or DIY - should not be discounted lightly.
3. This implies to me that SAAS is really driven by the AAS - the ongoing service revenue is what will drive profitability.
4. Interestingly, there do not appear to be multiplicative effects - changing any of the variables on the right side of the equation has only a linear impact.

Need to think about the strategic implications of this math some more.

 Update: Cleaned up implications when I realized that they were repeating.

This entry was posted on Friday, February 10th, 2006 at 9:49 am and is filed under Technology Business. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed. Reddit This