Update: Welcome, Chronicle readers. Since it wasn’t made clear in the Chronicle’s reference, I’d like to point out that the paper I’m quoting was written by Jim Farmer, who is the Coordinator of Georgetown University’s new Interoperability Center, formerly the Sakai SEPP Community Liaison and project administrator for the uPortal project. Jim has pretty impressive credentials across the board, including having served as the CIO for the Cal State University system. You can find his full bio here.
Sorry I’ve been incommunicado for a while. There’s been lots of stuff going on, both professionally and personally (including a new grandbaby).
Anyway, I’m going to make it up to you with a special treat. Jim Farmer has a fantastic analysis of how much it costs Blackboard to make a sale, what that means to their overall business model, and how that compares to the cost-per-sale of a commercially marketed open source software product and the cost for community-building the Open Source uPortal project. Some of it may be slow going if you’re not familiar with financial analysis, so I’m going to unpack a few of the good parts and speculate a bit on the implications.
Jim begins with the punch line:
Interested in the LMS market? Sign up to receive more information about our LMS Market Analysis service, including a free sample newsletter!
Using industry-average data, the cost of selling a Blackboard enterprise learning system is estimated to be $259,000 per sale….This cost compares to an estimated cost of $78,000 per sale for commercially-marketed open source software and $450 to $1000 for community-building for the uPortal product.
I take “cost of selling” to mean the cost of acquiring a new customer. Blackboard spends a quarter of a million dollars for each new school that signs on the dotted line. That money goes toward the labor-intensive dog-and-pony shows and proposal writing involved with each RFP bidding opportunity. Blackboard deploys a small army for each bid–not all of which result in a sale. Contrast this with Open Source. If you want to evaluate Moodle, you…well…you download Moodle. On top of the direct costs of sale, Blackboard has to sponsor conferences, support user groups, pay for fancy conference exhibit booths, and so on, because their customers expect them to do so.
OK. So. Blackboard pays a quarter of a million for a new customer. That’s adds up to be more than 30% of the total amount of money that they bring in. How, exactly, do they stay in business? How do they grow? $250K is a pretty high hurdle to jump, particularly when you consider that Blackboard is also constantly losing some customers (as any company does) even as they pick up others. According to Jim’s numbers, Blackboard needs to acquire 185 new customers every year to hold steady, 371 to get 5% growth, and so on, up to 742 new customers needed to grow at 15%. How many new customers a year has Blackboard been getting? 152. Less than they need to break even.
So Blackboard has to employ some other strategies in order to stay afloat. First, they try to sell to entire university and school systems rather than individual colleges and schools. For example, they’d prefer to spread that $250K cost across the entire 64-campus SUNY system by making one sale to the central offices than incur it 64 times to sell to each campus individually. Second, once they have a new customer, they upsell and cross-sell like crazy. Once you’ve signed up for the Blackboard LMS, they’ll try to sell you the Blackboard content management system, the Blackboard portal, the Blackboard toaster oven, and the Blackboard time share in Boca. The typical customer licensing the works will pay $160,000–per year. (Remember, you don’t buy Blackboard. You rent it.)
This sounds a bit like the business strategy of another well-known company: Oracle. One major reason that Oracle acquired Peoplesoft and Seibel is that it’s much cheaper to sell more stuff to the same customers than to find new customers, and in order to do that, they have to have more stuff to sell. This is typical with enterprise software; they all have high costs of sales. So at some point in the not-too-distance future, expect Blackboard to be the acquiree rather than the acquirer (just as Banner has been bought and sold a few times now), and Oracle is the most likely purchaser. And if you think that an acquisition of Blackboard by a larger company is necessarily good for the customers, think again.
There’s other interesting stuff here as well. Jim points out that Blackboard spends 11.9% on research and development, compared to 22.3% for Seibel. He notes, “This suggests the Blackboard products are more mature than Seibel’s–requiring less development as compared to ‘maintenance’ of current features.” First, I’m not sure how one distinguishes between “mature” and “arrested development”. Clearly, Blackboard has been under less competitive pressure to evolve its product. That may or may not last. In addition to more vigorous competition, a second reason why a company like Seibel might need to pay double in R&D costs compared to Blackboard is that CRM systems are currently much more likely to be heavily integrated into a bunch of other enterprise systems, while LMS’s are still mostly stand-alone. Maintaining interoperability against a range of products and standards is expensive, and so far LMS manufacturers have mainly only had to worry about integration with student registration systems. That will change.
An interesting anomaly that Jim points out is that Blackboard’s service business, while a significantly smaller portion of their revenues than Seibel’s, is hugely more profitable. They make a gross profit of 44.8% compared with Seibel’s loss of 13.7% I’m not sure why that would be. Part of it may be due to the integration issues again; it’s much easier for large, complex integration projects to spiral out of control and lose money than it is for other sorts of customization, development, etc. But that probably wouldn’t account for such a huge difference by itself.
Anyway, there’s much more in the report. I highly recommend reading the whole thing.