The Evolving LMS Market, Part I

As Casey Green said in my recent interview with him, the LMS space is a “market in transition.” In 2005, the year that Blackboard acquired WebCT, the two platforms had a combined total of 75.6% U.S. higher education market share, and the next closest competitor had barely cracked 2% market share. Today, the situation is substantially different and changing rapidly. But the narratives around exactly what’s happening tend to be off. Typically, I hear the frame as being a contest between Blackboard and “open source.” Has “open source” (by which we mean Moodle and Sakai, the two open source LMSs with significant market share in the United States) made inroads into the market?  If you read what the majority of sell-side financial analysts1 are writing, you may see the claim that “open source” is not putting a major dent in Blackboard. If you talk to Moodle or Sakai advocates, you might hear that they are crushing the company in sales. Neither account is really capturing what’s happening in the market, so I’m going to try to explain what we know about what’s really going on in a two-part series. In this post, I’ll talk about what the data are telling us so far about the recent shifts in the market, describe how colleges and universities come to decide that they need to go to market for an LMS, and assess the degree to which we may see an uptick in the number of schools that decide to look around and evaluate their options. In the second post, I’ll describe how the next four years of market transition may be different than the previous four and what signs we should be watching for to see which way the market is going to break.

What Has Happened So Far?

Let’s start with an analysis of where we’ve gotten to in terms of market share. Has “open source” (by which we mean Moodle and Sakai, the two open source LMSs with significant market share in the United States) made inroads into the market?

Here’s a graph of their growth, based on data from the Campus Computing Project’s annual survey:

 

In 2010, slightly more than one in five universities surveyed2 used one of the two most popular open source LMS platforms as its main, centrally supported, institutional LMS.

But that’s not the whole story. There is a third non-Blackboard LMS in the U.S. higher education market that has significant market share. Desire2Learn doesn’t happen to be open source, so it doesn’t fit the narrative. But that’s because the narrative is flawed. Here’s what happens when we add D2L to the picture:

 

We now see that the three largest non-Blackboard LMSs, as measured by U.S. market share, comprise over 30% of the market. That’s up from about 9% for the group in 2006, the year after the WebCT acquisition (which was announced in October 2005). But even this isn’t the whole picture, because Blackboard acquired ANGEL in 2009. What did the market look like when ANGEL was a non-Blackboard LMS? Here’s the picture, again drawn from data from the Campus Computing survey:

 

One of the interesting aspects of this graph is that the combined market share of the top non-Blackboard LMSs in 2010 exceeded their combined market share in 2009 even when ANGEL is dropped from the group in 2010. This is not to say that ANGEL’s market share contribution was insignificant or that Blackboard got nothing by acquiring the company. If ANGEL had remained independent in 2010, and if their market share had stayed the same as it was in 2009, the graph would have looked like this:

All in all, it’s pretty clear that non-Blackboard LMSs are growing very significantly in share, and that the growth appears to be accelerating for both open source and private source entrants. In 2009, before the ANGEL acquisition, the split between the open source non-Blackboard LMSs and the private source ones was close to 50/50.

But even that’s not the whole story, because higher education is not a uniform market. What happens if we look at the 2010 non-Blackboard market share by market segment? Let’s take a look, again drawing on the Campus Computing Project’s annual survey for the numbers:

These segments look very different from each other.3 For example, Desire2Learn has done extremely well in the community college segment, which so far has shown itself to be relatively open source-averse, but they don’t even show up in the private four-year college segment, where Moodle dominates. More generally, Desire2Learn does particularly well in those segments where open source does particularly poorly. It is possible that a significant number of schools are making decisions to either go with open source or avoid it early in their evaluation processes. If that’s true, then it may be that Desire2Learn isn’t even running into Moodle and Sakai in many competitive bid situations, since schools would have decided to filter out either one group or the other before making invitations to vendors. If, as colleges tend to do, they are copying their close peers on these decisions, then these patterns could easily propagate through market segments.4

The dominant narrative of “Blackboard versus The Open Source” captures none of this nuance. Nor does it capture the fundamental dynamic of how and why colleges and universities decide to go to market in the first place. It’s worth taking some time to examine that dynamic in a bit more detail.

What Drives Schools to Consider an LMS Migration?

Switching your institutional LMS is not like switching your web browser. It’s a huge and usually painful process involving moving over large quantities of content, developing new integrations with other campus systems, and retraining faculty, staff, and students. I don’t believe the LMS market is functionally differentiated enough that most colleges and universities are likely to feel compelled to consider migrating (or to dismiss the possibility of migrating, for that matter) based primarily on feature differences. Functional differentiators can be decisive in terms of which platform is ultimately chosen, but generally not in whether one is going to contemplate migrating in the first place. Service isn’t typically the main reason why schools move either. It is occasionally possible that a school can have such an unacceptably bad experience with product reliability or vendor customer service that it drives them to leave, but my observation is that, so far, relatively few schools find such situations to be primary drivers for deciding to consider a migration (although a bad customer service experience can be a strong aggravating condition.) That leaves two other factors: cost and forced migration.

Cost: Blackboard has never positioned itself as a low-cost provider. To the contrary, Blackboard’s executive management has indicated publicly in the past that they are more focused on average contract value than on number of LMS customers. It’s very natural for larger companies to increasingly focus on their larger, more profitable customers, since that is usually where most of their growth will come from. I argued in 2007 that this was exactly what was happening at Blackboard and, furthermore, the client loss the company was seeing showed no signs of affecting their financial growth prospects at that time, suggesting that the customers they were losing were indeed primarily the most price-sensitive Blackboard Basic and WebCT CE customers who were forced to move for budgetary reasons. But one of the risks associated with a strategy to move upmarket is that, when hard economic times hit, your customers may find cheaper solutions to be adequate. The question now is whether the current financial crisis leads more and different customers to move to lower-cost platforms.

At this point in any market share discussion that frames the issue as “Blackboard versus The Open Source,” the issue of total cost of ownership (TCO) typically comes up, with somebody asserting that open source LMSs have a higher TCO because of hidden costs. But there are several problems with this argument. To begin with, there are no studies I know of to support this assertion. In fact, what little we do have in the way of TCO analysis cuts the other way. The North Carolina Community College System found that moving to Moodle from Blackboard resulted in a 35% increase in TCO during the transition year followed by a 72% decrease in TCO in the post-migration years. Likewise, the University of North Carolina, Chapel Hill found that the total annual cost of Sakai would be $332,000, as compared with $620,000 with Blackboard. The increasing numbers of open source LMS customers who use hosted solutions and do not hire developers also provides a significant and growing body of evidence against the argument that an open source LMS must come with hidden costs. And finally, the argument completely neglects to mention the possibility of non-open source alternatives that are less expensive. Neither Blackboard nor Desire2Learn publish their prices, but anecdotal evidence suggests that Desire2Learn’s prices may be significantly lower than Blackboard in some cases. At the very least, D2L appears to be price-competitive.

All of this raises two questions. First, will more—and larger—schools be inclined to switch based on cost differences than have in the past due to new financial realities? Second, will Blackboard come under pressure to cut more aggressive deals in order to hold onto their customers? These are both pretty big unknowns at the moment.

Forced Migration: This is the big one. Blackboard is retiring WebCT in January of 2013 and ANGEL some time in 2014. Customers on those products will be migrating whether they want to or not. In that situation, the majority are going to look around to see what their alternatives are when they otherwise might have chosen not to do so. Blackboard can incrementally reduce the migration effort required to moved to the mainline Blackboard platform versus the competition, but overwhelming anecdotal evidence suggests that, at least so far, they have not been able to pass a threshold at which customers feel like moving to Blackboard won’t really be a migration. Since LMS evaluation and migration efforts typically take about two years, Blackboard probably has about twelve more months to change the equation for WebCT customers and a little more for ANGEL customers. Based on the math that Casey Green laid out in the earlier-referenced recent interview, it appears that there are a bit under nine hundred customers left on legacy Blackboard LMS products. About eight hundred seventy-five Blackboard customers will have to migrate to something in the next four years. That something may or may not be a Blackboard-owned product.

Are the Alternatives Viable?

Even if schools decide they would like to look at alternatives, they have to feel that they have someplace to go. Just a couple of years ago, there was a sense in the market that viable alternatives were in short supply. In 2006, no non-Blackboard-owned LMS had significant market share in the United States, open source was still considered to be untested by many college and university CIOs, and Blackboard was suing Desire2Learn for patent infringement, creating significant uncertainty in the market about Desire2Learn’s long-term viability as a going concern.

Since then, a lot has changed. Here’s what’s true in 2010:

• Every single market segment in the U.S. has at least one non-Blackboard LMS that has at least nine percent market share, and all but one have at least one entrant in the double digits. This includes but is not limited to open source offerings.
• Blackboard has dropped the lawsuit against Desire2Learn, and as the market segmentation graph above shows, there is strong evidence that D2L is having a significant win rate for colleges and universities that are considering a migration from Blackboard but aren’t comfortable with open source.
• As externally hosted LMS solutions are becoming more popular, the advent of cloud computing has greatly diminished the advantages that big companies like Blackboard have in terms of the relative reliability and security of the data centers. For example Moodlerooms’ hosted offering runs on Dell’s data centers.
• SunGard Higher Education, an organization that has roughly the same financial heft as Blackboard and somewhere around 40% market share in the United States for mission-critical ERP/Student Information System software, has entered direct competition with Blackboard by offering hosting and support for Sakai in partnership with rSmart, making open source more palatable to risk-averse customers.
• Datatel has formed a close partnership with Moodlerooms somewhat similar to the SunGard/rSmart partnership.

There is some historical precedent for these last two bullet points and the impact that they might have on the market. Around 2000, SCT (the then-owner of the Banner SIS) signed a partnership and reseller agreement with WebCT. According to Karen Gage, who was the Vice President of Marketing for WebCT at the time, that partnership accounted for somewhere between 10% and 25% of WebCT’s LMS sales during the first few year or two of the relationship. Sales dropped off considerably from there and the partnership was ultimately ended—and anyway, the LMS market is quite different today than it was in 2000—so we shouldn’t read too much into this data point. But it certainly is enough of an indicator to give us reason to believe that these partnerships will have some impact on whether schools at least decide to go to market and evaluate their LMS options.

There’s an interesting question about the degree to which other entrants can break into this market. I’ll just mention two possibilities. First, Pearson recently bought Fronter, a European LMS (one that impressed me with their demo, by the way) and has created a new brand for their combined eCollege/Fronter product called LearningStudio. They also signed a deal with Arizona State University to support a small number of distance learning courses for the college that might be a leading indicator of the company getting more aggressive about going after the traditional higher education market. (Right now they have very high penetration among for-profit universities, but only about 1.3% share in the traditional not-for-profit U.S. higher education market.) Second, startup Instructure just recently signed a deal for their Canvas LMS with the Utah Education Network (UEN), a multi-institutional consortium that reaches about 109,000 college students and 40,000 K12 students. That’s a pretty big deal for an LMS startup, and it may presage a thirst for new approaches in the LMS space (in terms of both product design and support/pricing models) leading a percentage of the market to be more risk-tolerant than they have been in the past.

Where Does It Go From Here?

The bottom line is that there has been a dramatic increase in the perceived viability of the alternatives at a time when both forced migration and economic crisis are putting pressure on schools to minimize costs across the board, including LMS costs. We have to weigh the impact of these factors against current market trends. As a baseline, I tried to extrapolate the market share movement in the previous graphs out to 2014. What I did was I averaged the last four years of market share change and, since I wanted to capture the fact that changes in the market have accelerated over the last four years, I double-weighted the last year’s change. (Primitive, I know, but I’m not a statistician.)

Here’s what the graph came out looking like:

This projection has Moodle at around 31%, Desire2Learn at about 20%, and Sakai at about 12% in that year. By extension, that would put Blackboard with market share somewhere in the low- to mid-thirties by 2014, with a few percentage points left over for other options.

There’s a pretty substantial margin of error in this exercise. My completely unscientific guess is that we’re at something like plus or minus four percent (absent major unforeseen developments). There’s a lot that could happen within those lines that could have a significant impact on the various platforms and the vendors that support them. In my next post, I’ll get further into the weeds to see what signs we should be looking for in order to divine what’s actually going to happen over the next four years of LMS market share in more detail.