The LMS at 30: From Course Management to Learning Management (At Last)

Our coverage at On EdTech is shifting to look more at big picture impacts and less on day-to-day changes in markets, particularly for the LMS market. Today we have a guest post by Matt Pittinsky, co-founder of Blackboard and currently a board member at Instructure via the acquisition of Parchment. Matt is no longer an active EdTech executive, but he has a wealth of historical knowledge and insights. When he shared a draft post with me, I encouraged him to develop it further and invited him to publish at On EdTech to benefit the community. Of course, this article is his and represents his perspective. - PH

Introduction 

The LMS conference season is over in North America and it is clear from reports of their proceedings that the market is moving full-speed ahead in its embrace of AI. The LMS is evolving, changing, in both sustaining and disruptive ways.

Zooming out, a new technology is rapidly emerging that has the potential to impact educational access, quality, efficiency, and equity. While this new technology, AI, adds choices and enhances existing practice today, it is easy to imagine it replacing and changing practice over time.

The future feels uncertain. And yet we’ve been here before, at least sort-of. The closest analog to AI’s impact on education is the Internet’s rise in the mid-1990’s, which gave birth to the LMS among other concepts over the past three decades (e.g., MOOCs, OPMs, Flipped Classrooms).

Indeed, this year marks a milestone in the Internet’s impact on education: the 30th anniversary of the Learning Management System (LMS). My peg is 1995, the year the Instructional Management Systems Project (IMS) was founded, a pivotal moment in shaping the future of digital learning.

If the rise of the Internet is the closest analog to what we are experiencing with AI, then stepping back to reflect on the origins and evolution of the LMS is timely. Documenting the LMS history — my experience was on the supply/vendor-side — offers insights into what lies ahead. I was fortunate to be an early participant in IMS, then co-founder of Blackboard, an academic, and now board member at Instructure following its acquisition of Parchment – how’s that for a twist?!!

As a 30 year observer and participant, it seems to me that previous technology platform shifts like SaaS and mobile did not fundamentally change the LMS. AI is different. We’re standing at the precipice of LMS 2.0, where the branding change from Course Management System to Learning Management System will finally live up to its name. Unlike SaaS or mobile, AI represents a technology platform shift that will transform the way participants interact with learning systems – and with it, the nature of the LMS itself.

Given the transformational potential of AI, it is useful to set the context and think about how we got here, especially on this 30th anniversary of the LMS.

The IMS Project

The Instructional Management Systems project – later IMS Global and now 1EdTech – was founded to define the technical architecture of digital, networked learning.

At its core, IMS was a middleware strategy. It wasn’t attempting an end-user solution. Instead, it foresaw a world of discipline-specific, pedagogically rich learning applications that would need to interoperate across institutions. It also anticipated a hybrid academic and commercial software market, or ecosystem, that translated the IMS architecture and standards, if colleges and universities insisted, in a way that ensured seamless instructional management systems for instructors and learners. For networked learning to be open and robust it needed technical standards from the start.

Then, as now with 1EdTech, IMS was a consortium of schools, universities and companies.¹ While a short list of individuals can be credited as IMS’s “founding figures,”² I’ve always viewed Carol Twigg as the visionary who saw the potential for technology to transform instruction rather than just digitize it; to harness technology to improve learning outcomes and lower costs in the college and university classroom. Carol ran Educom’s National Learning Infrastructure Initiative, the sponsor and organizational home of IMS.³ In 1998, Educom and CAUSE merged to become today’s Educause.

The Original Vision of IMS

As noted, IMS wasn’t envisioned as a solution so much as an orchestration layer – an open, interoperable framework allowing instructors to integrate discipline-specific, pedagogical, and course administrative tools into a cohesive digital learning experience. Its premise was that faculty shouldn’t be EdTech “systems integrators” and institutions shouldn’t be locked into monolithic vendor ecosystems. The goal was an instructional environment where content, experience, and technology could be developed, shared, adopted, and aligned seamlessly, enabling innovation rather than constraining it. Today, this vision is most directly represented by 1EdTech’s Learning Tools Interoperability (LTI) standard through which countless university and vendor developed tools and courseware “speak” to each other, often through the LMS, to enable rostering, grade exchange, and more.

From IMS to Course Management

IMS’s middleware vision evolved – actually, some might say contracted. Rather than an orchestration layer that awaited things to orchestrate, the market voted with its feet and moved toward all-in-one Web course management platforms. Early applications like Web Course in a Box, WebCT and Blackboard defined what became known as the Course Management System (CMS), providing everything an instructor needed to create and manage a course Web site: content delivery, grading, discussions, and administration. The emphasis was on managing course activities and Web publishing rather than rethinking how digital learning environments could orchestrate learning. 

It was about providing the 20% of capabilities that 80% of faculty needed, not the platform for enabling the 80% of discipline-specific and pedagogically-specific capabilities that, for any given capability, only 20% of faculty may need.

It took far longer than many anticipated for IMS’s vision to translate into the widespread adoption of a standard like LTI. While institutions recognized the importance of standards, abstractly, what they urgently needed was a solution to the problem of enabling non-HTML-coding faculty to create and manage course Web sites. As well, you can’t solve the problem of tool interoperability before there is broad adoption of the platforms that enable tools in the first place. We put the proverbial cart before the horse. Platforms emerge from products, not the other way around. 

From CMS to LMS

Phill Hill (On EdTech) has done a nice job graphing LMS vendor entrants and exits back to 1997,  which was the year we started Blackboard. While today the market is dominated by a few providers – Instructure (Canvas), Blackboard, Desire2Learn (Brightspace), Moodle, Google (Classroom) and Powerschool (Schoology) – over the past 30 years names like Universal Learning Technologies/WebCT, IBM (Lotus LearningSpace), MadDuck Technologies (Web Course in a Box), Sakai, George Washington University (Prometheus), Angel Learning, Real Education, LoudCloud and others have come and largely gone.

While most of these applications were proprietary, a few were open-source, most notably Moodle and Sakai. The appeal of open-source for learning management was the model’s promise of the best of two worlds – a solution for web course tools that also enabled high degrees of customization, an ecosystem of interoperable tools, and a hedge against vendor lock-in. For the same pragmatic reasons that IMS turned into Web course management systems, these open source applications came to look more and more like commercial solutions, or faded, even as commercial solutions such as Canvas also came to market with an open source option.

By the early 2000s, “Course Management” had been relabeled “Learning Management,” aligning with the product category term used in corporate learning. The rebrand was aspirational at best, as the product was still much more about managing courses than actual facilitation of learning  –  more on that in a moment. We changed the name but the fundamental category paradigm remained unchanged.  

The last 30 years have seen a massive expansion in technology use in classrooms, and technology ownership by students and instructors. We’ve also seen major technology platform shifts, but I would argue not category redefining shifts. The emergence of mobile added new conveniences, but didn’t really challenge the core model of what an LMS does and how it does it. SaaS disrupted the market dynamics – favoring cloud-native providers like Canvas – but again, the LMS paradigm stayed largely intact.⁴   Today’s LMS may be more scalable, prettier and easier to provision, but presented side-by-side, an LMS course Web site today closely resembles those from 1999. Indeed, it is remarkable how much even the strategic framing of today’s LMS is the same as then.⁵  

Consider how similar this early version of the Blackboard UX is to what you typically experience in an LMS today.

And consider this excerpt from a 1998 Blackboard investor practice presentation and how similar it sounds to the strategic framing and aspirations of LMS vendors today (e.g., platform for content).

The LMS in Online, Professional, and Continuing Education 

Not all LMS use cases evolved the same way. A notable exception is online, professional, and continuing education (aka CareerEd or OPCE), which developed on a separate trajectory.

From the late 1990s, platforms like Real Education (which became eCollege) and Embanet (based on FirstClass’s Softarc, a Lotus Notes competitor) focused on fully online delivery rather than traditional course augmentation.⁶  These platforms prioritized structured, linear learning experiences and placed greater emphasis on student experience and instructional orchestration. Unlike traditional LMSs, this segment was more heavily impacted by mobile adoption, as the need for flexible, on-the-go learning drove design choices. Today, we still see this divide, as professional and continuing education providers are often at the forefront of technology-driven instructional innovation. The descendants of Real Education are Coursera and the OPMs, whose platforms are now being licensed to institutions, for example EdX and Western Governor’s University, because of their unique product-market fit for this segment of the market. Many of the ideas exemplified by Coursera, EdEx, and Udemy were anticipated years ago by AllLearn (University of Chicago) and Fathom (Columbia University), but the market wasn’t ready. Even when you accurately predict the future, timing matters!

The rapid growth of online enrollments and the unique needs of the OPCE market have led to meaningful evolutions in what an LMS does and how it does it – I’d argue the exception that proves the rule.

The LMS in K-12 

The LMS category originally evolved in higher education and then expanded downward (by age-grade) into K-12. It wasn’t long after Cornell adopted Blackboard CourseInfo that the Ithaca School District did the same. That said, K-12 adopted the LMS more slowly, with many districts today still lacking a true district-wide LMS platform. When they do adopt, school districts have chosen both higher education LMS vendors (e.g., Blackboard, Canvas) and K-12 specific vendors as well (e.g., Google Classroom,⁷ Schoology). Several historical dynamics explain why.

First, in the early days, few school districts had the resources to operate an on-premise “enterprise” application like the LMS with, for example, the peak loads that come from every student in the class completing the same assessment at the same time. The shift to cloud-based SaaS removed this obstacle, but from a later starting point than higher education, which had more capacity to run on-premise enterprise applications earlier.

Beyond provisioning there is also the challenge that K-12 is really three segments: elementary, secondary, and professional development. The needs of elementary schools are different from middle and high schools, and the needs of school district professional development are different from regular school-aged instruction (this is true of community colleges and small liberal arts colleges, of course, but not at the level of the learner to the same degree.) Higher education solutions met the needs of high schools and professional development, but not the full district. At the same time, K-12 LMS solutions lacked the robustness and familiarity of higher education LMS solutions.

Market size also matters. Annual Contract Values (ACVs) and the addressable market of higher education dwarfed K-12, partly because it matured faster, making K-12 an attractive expansion or secondary market for higher education LMS providers, but a difficult market for a K-12 pure play.

Most importantly, to state the obvious, the instructional model of a standards-based, standardized curriculum classroom, centered around formative and summative assessments, also make the needs of K12 quite different from higher education. More on that in a moment.

For these reasons K-12 has been under-served as the higher education LMS set the standard. With the rise of AI and an increased emphasis on instructional orchestration in LMS 2.0, I think we are likely to see influence flow in the opposite direction – where K-12 informs higher education. As well, the K-12 addressable market will expand, as the LMS 2.0 taps spend on course materials, tutoring, course management, and even labor.  

As noted, the LMS in K12 has historically had a stronger focus on instructional orchestration due to the structured nature of assessments, rubrics, clear state standards, learning objectives, and proficiency-based measures. Moreover, the incentive structures in K-12, particularly funding models tied to learning outcomes, have reinforced the need for precise tracking and alignment of content with state standards. Yes, levels of academic autonomy are quite different in K-12 than higher education, but some subsegments (e.g., community colleges) and course delivery models (e.g., professional instructors), provide space for innovation to develop from K-12 into higher education and spread from the margins. The growing embrace of competency-based education is a prime example.

K-12 is already ahead in both availability and adoption of AI tutors / virtual TAs (e.g., Flint).

LMS and Subscription Software 

No discussion of the history of the LMS – at least a history that centers on the LMS supply-side – would be complete without considering how business models evolved. Unless and until proven otherwise, I think the LMS was the first enterprise software category that relied on subscription licenses from the start. In other words, what we take for granted as a core characteristic of SaaS – recurring revenues from subscription licenses – was the “native” licensing model for the LMS ab origine.⁸  Beyond bragging rights and novelty, this matters to how the market evolved.

Prior to the LMS, the main enterprise applications adopted by universities were administrative in nature – SIS, HR, Finance, etc. As in other sectors, these applications were licensed through a large one-time “purchase” price, and a much smaller recurring maintenance fee, usually a percentage of the one-time price. As the LMS expanded from faculty and departmental adoption, to full university licensing, vendors like Blackboard framed the category as analogous to the SIS, but for the “front office” of teaching and learning. By that logic, LMS’s should also command a large one-time fee and a smaller recurring maintenance fee.

This argument failed miserably. Despite my confidence in the earlier investor video, colleges were not paying $150k for an enterprise LMS in 1997! Few institutions saw the LMS as more than a course Web site editor – a tool – while fewer still had the capital budgets to make a large one-time software purchase for academic computing. Quite by accident, the fall-back of early players like Blackboard was to license the LMS as an annual subscription at the approximate price of the smaller recurring maintenance fees charged for administrative systems. The LMS took hold through the operating budget of the university, not the capital budget. And the LMS category became the accidental early adopter of today’s dominant SaaS pricing model.⁹

More precisely, the LMS vendors tended to charge tiered fees, with low entry pricing for departmental-level adoptions up to larger fees for the full campus, but never approximating the price points of analogous “run-the-business” type solutions in education (e.g., SIS) or other industries.¹⁰  These fees were also tiered based on size of institution, as measured by Full Time Enrollments (FTEs), a standard way of counting degree-seeking students. The consequences of this accidental and bottoms-up licensing model have been positive, negative, and mixed.

On the positive side, the LMS recurring license model means that in any given year, the financial performance of the company is driven much more by retaining/servicing existing customers than winning new ones. On the negative side, orienting product roadmaps towards existing customers can dull innovation and miss disruptive technologies. Also on the negative side, the relatively low cost of the LMS meant vendors needed to quickly prioritize new markets and new products to expand TAM and grow – being a higher education LMS isn’t “enough.” On the mixed side, defining institutional size by FTE missed the growth of non-FTE-denominated learners and tied vendors to a metric, traditional enrollments, that by most forecasts will be declining for demographic and other reasons.

Equity Financing and Consolidation

This discussion of pricing raises yet another key topic in the development of the LMS, one that is highly relevant as we read about money chasing AI companies in every sector. My rough estimate is that approximately $1 billion of equity financing has fueled the evolution of the LMS.¹¹ This does not include the large checks written to buy companies, just the funding that went to the balance sheet of LMS providers to innovate and grow.

Simply put, I do not believe we would have seen the rapid maturation and adoption of the LMS had the category not attracted funding that enabled providers to grow “ahead” of the market. This is true of sales and marketing, as much as product development and R&D.

That said, if all this capital helped accelerate innovation and adoption, why has the category remained fundamentally unchanged as we head into the AI era? We’ve discussed part of the answer already. The U.S. higher education market is finite. There are 4,500 institutions or so, with 1,100 accounting for most of the software value given their size. Once all that investment achieved near full category adoption, the need to grow pushed LMS providers towards product, segment, and geographic expansion. True, a company like Instructure grew by offering a better LMS product and partnership model, replacing vendors who didn’t keep up. But that is the exception that proves the rule. Most growth came and comes from going beyond the LMS, as growth capital demands… growth.

The other part of the answer is consolidation. The elephant in the room when writing a history of the LMS is (old) Blackboard’s voracious tendency to acquire competitors. As someone who played a role in pursuing this strategy I can explain the motivation, and take responsibility for the unintended consequences.

Our motivation was simply the belief that the LMS is a “winner-takes-most” category. The benefits of institutions standardizing on a common LMS include simplification and acceleration of third party tools and courseware, enablement of cross-institutional program models (e.g., course sharing), and more leverage from faculty and student LMS literacy as they move to and from institutions, to name a few.

In other words, our motivations were, in a sense, the original IMS and platform vision. Absent very robust standards that are broadly adopted, or an open source architecture that is broadly adopted, if you believe that the LMS is the “operating system” of instruction then the market evolving around one high market share provider makes sense. We executed a strategy to be that provider.

The biggest cost of this strategy was that a lot of product development ended up focused on consolidating LMS providers as opposed to innovating and evolving the LMS. Attention spent on technical debt, migration cycles, and acquired company retention was attention not spent on platform shifts (i.e., the cloud and SaaS), and reinvention. Paradoxically, the strategy that led to a single broadly adopted LMS platform with the scale to pursue instructional orchestration, was the very same strategy that frustrated the LMS’s ability to evolve from a Web course system to that operating system for learning.

Closing Thoughts

I subscribe to Bill Gate’s maxim that we tend to overestimate the amount of technology-driven change in the short term and underestimate it in the long term. I thought I covered my bases when I used 10 to 20 year timeframes the last time I made predictions about the development of digital learning technologies.

Forecasting the time-scale of change in education is tricky, as tricky as declaring “2.0” for a category that has developed over thirty years!  Indeed, what is particularly mind bending about AI’s potential speed and scale of impact is the way it “stacks,” or builds, on all the prior technology shifts, whether they changed the LMS or not.

I don’t have conviction when it comes to the specific pace with which the LMS will change in the AI era. That said, I have complete conviction, based on its history, that it will change and roughly how.  

First, the LMS will embrace AI as a sustaining technology, which is to say a 100x better way of doing course publishing and management. Second, AI tutors / agentic TAs will pioneer a new dimension of the “instructional management system” – learner-centered support of teaching and learning. Put simply, the name change from course management to learning management will be true in fact, not just positioning. Third, these capabilities will combine with and transform the LMS into a new, expanded, more valuable, core platform for learning adopted by schools and universities around the world.

During this 30th anniversary year, it is important to recognize the Instructional Management Systems project and the remarkable evolution and adoption of learning management systems across education segments, worldwide. The foresight and hard work of people like Carol should be recognized. Reflecting on the history of networked learning – albeit one participant’s experience – provides the opportunity to discuss and debate lessons learned, mistakes to avoid, and opportunities to grasp as we enter the AI era. If my characterization of LMS 2.0 is correct, it is back to the future. The early vision of a core platform for instruction that orchestrates the teaching and learning process, customized to the context of the course and personalized to the needs of the learner, may soon be upon us. At last. 

The main On EdTech newsletter is free to share in part or in whole. All we ask is attribution.

Thanks for being a subscriber.

1  Universities included University of Michigan, University of North Carolina, California State University and companies included Thompson Publishing, KPMG Consulting, IBM.

2  Bill Graves at UNC, Carl Berger at Michigan, and the project co-leads Mark Resmer (Cal State Sonoma) and Steve Griffin (UNC).

3  Carol went on to run the Center for Academic Transformation, which suspended operation in 2018 but whose ideas remain highly relevant today.

4  For example, compare Blackboard’s early UX to LMSs today.

5  I think this has the virtue of being true, but refer the reader to Blackboard’s original business plan, as well as this first practice investor pitch I presented in 1998.  Also consider this discussion of online learning at Maricopa Community Colleges 25 years ago between Murray Goldberg of WebCT and me.

6  Embanet was short for eMBA Network.  It was developed by students in an MBA program.

7  Google Classroom is worth additional consideration for a few reasons.  First, while over the years several large tech companies declared an interest in serving the LMS market (e.g., IBM, Oracle), Google is the only one that actually delivered and stayed the course (pun intended).  For big tech companies the juice often wasn’t worth the squeeze.  Second, the tendency of districts to use Google Classroom in elementary schools, and Canvas or Schoology in middle and high schools, is further proof of the different needs within K-12.  Finally, Google is uniquely positioned to pioneer LMS 2.0 given its Classroom and AI capabilities… if it chooses.

8  Salesforce, founded in 1999, is widely considered the pioneer of SaaS, combining both cloud and subscription licensing.  While not cloud-based, Blackboard, founded in 1997, was also based on a subscription license model from the start.  Blackboard went public on June 18, 1999, five days before Salesforce.  Blackboard’s IPO road-show, having preceded Salesforce’s by a week or so, was among the first to introduce public investors to many of the concepts and metrics that are characteristic of SaaS today: deferred revenue, annualized revenue, churn, etc.

9  Believe it or not, there was a moment when the LMS could have adopted an advertising-based model.  At the headiest moments of the dot com bubble, when companies like Campus Pipeline and Mascot Networks were attracting attention with free campus portal solutions, companies like WebCT and Blackboard were under pressure to focus on scaling “eyeballs” over sound unit economics.  Thankfully, neither did.

10  Again, fully online education was different, with vendors such as RealEducation taking a fee per enrollment model similar to today’s OPM arrangements.

11  This estimate anchors on the private fundraising rounds and initial public offering raises of Blackboard, WebCT, eCollege (RealEducation), Instructure, Schoology, and D2L.