What LMS to Choose After Blackboard?

This weekend I got an email from a leading university that is planning to leave Blackboard and wanted some advice as to how Sakai might fit into their future.

For the past few years, when I get these “What LMS to choose after Blackboard/ANGEL/WebCT…?” emails from folks, I usually send a short note that effectively says, “You have my permission to go to Canvas – everybody is doing it. At least Canvas will be better than Blackboard.” (i.e. Canvas is the new Black)

But this particular morning, with a cup of fresh coffee in hand, I wrote and sent the following note that evolved into an open letter to Blackboard customers who are getting a “little nervous”. The note was to a school outside the US – so you see my focus on privacy issues.

Subject: What LMS to Choose After Blackboard?

Thanks for the note and thanks to your colleague for the reference. I have spent almost 15 years in the LMS market now and know the products and systems very well. As you might well imagine I am very excited about Sakai and believe it to be the best solution in the marketplace for the present and the future.

To me the most important dimensions of an LMS are: (a) an LMS must continuously evolve to meet the needs of on-campus, distance, and education at scale (i.e. MOOCs) as experienced by a modern world-class university, (b) for a university outside the US having 100% control over all student data is a must, and (c) over the next decade while the LMS will continue to be the backbone resource for teaching, increasingly innovation must happen outside the LMS and be seamlessly integrated into the LMS – this allows the LMS to focus on stability, scalability, and consistency while freeing local instructors, instructional designers and IT staff to explore innovative pedagogies freely without compromising the stability of the LMS.

With this context, for a leading institution that is looking forward to innovating in their approach to teaching and learning I feel that Sakai is simply the best LMS in the market. Sakai is the only “100% open source” product and that matters. In a truly open source product, we have no shareholders and no product managers – our shareholders and product managers are our adopters and community – that means you. The schools, companies, and individuals that invest time and resources are the ones that set Sakai’s direction.

Sakai is part of the Apereo Foundation and works with other open source projects that support the academic mission. Perhaps you use CAS as your single signon on your campus. Of particular interest to me is the Tsugi (www.tsugi.org) project that is an open source framework that allows the agile development of a wide range of pedagogically innovative learning tools that are easily and seamlessly integrated into LMSs like Sakai and Canvas using the latest IMS Standards like Learning Tools Interoperability, Common Cartridge, and IMS Content Item.

This is my short assessment of the vendors in the LMS marketplace:

  • Sakai is by far the best LMS for a strong institution that wants to “move forward” and sees the learning technology space as needing further innovation over the next decade. The combination of Sakai and Tsugi is simply the best solution in the marketplace of the notion of a more ecosystem-oriented “Next Generation Digital Learning Environment” (NGDLE). If you want to be part of the community who will define the NGDLE, Apereo Open Source is the answer.
  • Canvas is my second choice as an LMS as it is a leader (like Sakai) in allowing external tools to seamlessly integrate into the product. If you are a school that is interested in being a “fast follower” or have a weak IT team, Canvas is an excellent choice. Canvas is a good choice for second-tier schools or community colleges that want to be innovative but don’t want to be a core influencer in the next generation architecture. Interestingly Tsugi also works very well with Canvas because Sakai and Canvas implement the same IMS standards in very similar ways. Canvas is not “100% open source”. They release portions of their source but the intention is not to have you run your own instance of the software. That of course means that you are putting your data in the cloud in the hands of a US company. I personally think it is important to control one’s own data.
  • Moodle is a fine LMS and it is certainly Open Source – but it is not participative open source like Sakai. End-user adopters have very little influence in the direction of the product. But if you are a small school with a weak IT department, you probably don’t have that much interest in changing the features of your LMS. Moodle generally trails the market in terms of implementing the latest interoperability standards but they get there eventually. Self-hosted Moodle at a small school takes very low levels of developer resources as long as you *never* customize the code base or upgrades quickly become almost impossible. Cloud hosting Moodle with a company in your country is probably sufficient to assuage your data concerns. I would not recommend cloud-hosting *any* LMS with a US-based company at this point.
  • Blackboard has been effectively frozen for the past four years because of their grand rewrite they call “Ultra” that has not gone well. The primary reason for the failure of Ultra is some very bad architectural decisions that make it challenging to make an application as complex as an LMS highly reliable. The result of over-investment in Ultra leaves Learn as a pretty clunky piece of antique software, lagging in implementing the latest interoperability specifications, and a future that will be a very painful upgrade as the gulf between Learn and Ultra widens. I am glad to hear that you are considering leaving Blackboard.
  • Desire2Learn/BrightSpace is a clever company and their customers seem to like the product very much. My experience is that they are the best product for a heavily multi-tenant architecture. Many of their most successful customers are state K12 systems in the US that want to purchase a single product to be used by 100+ small schools. D2L has historically been a laggard in implementing interoperability standards but in the past year or so – they have realized the importance of interoperability and are working to be on par with Sakai and Canvas although it will take some time.

In summary, if you are leaving Blackboard, if you are bold and want to be part of an effort to take on the next decade of challenges in teaching and learning you should choose Sakai. If you have a weak IT organization but want to appear to be “innovative” with less risk and can tolerate the privacy consequences of cloud-hosting with a US company, choose Canvas with my blessing. If you are a smaller school with a weak IT organization, and just want something solid that will work, choose Moodle – either self host it and *never change a single line of code*, or out source hosting to a non-US company.

I use all three of these LMS’s. Sakai is amazing and working with a thriving and robust community of like-minded innovators is even more amazing. Building Tsugi in this same community is the most fun thing I have ever done. Canvas is a good workhorse – it has fewer features but does a solid job on those features and has an industry-leading interoperability strategy. When I have to install and administer my own LMS in my spare time for small-scale applications, I use Moodle. But I have changed 50 lines in my Moodle installation and so I have not upgraded my Moodle for 3 years because I don’t have time to re-debug and re-implement my changes in the latest release.

If you do choose to move from Blackboard to Sakai, there are companies that will help you through the conversion. There are local Sakai hosting companies that we can put you in touch with. There are also Sakai commercial companies that will maintain your Sakai installation on your campus using your hardware so you have 100% control of your data but leave someone else to deal with patches and upgrades. True “100% open source” leads to a diverse, international, rich and innovative commercial provider community.

I am happy to talk more and I expect to be in in your area next month and might be able to swing by your campus to give a presentation on Tsugi, Sakai, and Apereo. Even if you choose Canvas or Moodle – you should join Apereo so you can work with us on Tsugi.


  1. Jerry Lewis says:

    I’m interested in why you say never host your data in the cloud in the US. What are the privacy consequences?

  2. I am not a lawyer and I am simplifying things a bunch. I am talking about a European, Canadian, etc school storing their learner data in the cloud facility of a hosted US LMS vendor. The US Government (with the proper paperwork) can request access to any data owned by a US company and they must comply. The school whose data has been requested likely does not even know their learner data has been handed over to the government. Most US Universities look at this and thing “meh – we would have had do give them the data anyways and all we would know is when our data is being taken”. But laws outside the US are clearer as to the duty of a school to protect the privacy of their learner data.

    Having the data from a non-US University handed to the US government is firmly against the law in most countries and the school has a responsibility to see that this does not happen. The sure way to keep your non-US school’s student data from ending up in the US Government’s hands without you knowing it, a non-US school should host in a non-US facility from a non-US company. Your data may be turned over to the local government without your knowledge (but at least it is not against the law). A non-US school that wants the maximum compliance with strict student privacy laws should host their LMS servers on their own campus. They still must comply with legitimate requests from the their government but at least the school is aware of what is happening with their learner data.

    Of course this is a question of principle versus convenience when it comes to these universities and their student data. I am kind of pure about this stuff – but most campus CIOs and US-hosting companies are quick to find some grey area to rationalize the most convenient choice. So Canadian, UK, and European schools are putting their learner data in hosted facilities owned by US companies. As campus IT organizations become less and less about actual IT and more and more about managing vendor relationships having a 100% commitment to learner data privacy is simply too much work. So lots of non-US schools “follow the leader” and move their data into clouds owned by US-companies and accept the risks. At least to this point, the US government has not abused its ability to peer into non-US learner data hosted by US companies in their cloud so up to this point, while I think the decision is wrong on the right/wrong scale, the risk to the school of getting in trouble for leaking student data to the US Government by outsourcing their LMS hosting to a US company has traditionally been low.

    Sakai schools outside the US that self-host know *exactly* where their learner data is and who is looking at their learner data. I like that. Pure and simple – no grey area.

  3. I received an email asking why I did not include “OpenEDX” in the list of “Open Source LMSs”. IMHO OpenEDX is a MOOC platform not an LMS so it is not a good replacement for a campus Blackboard installation. OpenEDX interoperability strategy is very haphazard – end users hack around the edges but interoperability is not a core goal of edX itself. OpenEDX is fine for a school that wants to self-host a few MOOCs as an experiment but not sufficient for an enterprise-wide deployment.

  4. Matt Clare says:

    On the cloud and a compelled company:

    Good cloud hosting should encrypt data so there platform/infrastructure, etc sub’ed company, or other areas of the larger company, can’t mess around with student data. That should protect the data wherever the server sits.

    But of course it doesn’t, a duly compelled company can be compelled to assist in providing keys and means, etc.

    My observation is, doesn’t that argue against involving a company from another jurisdiction in any on-campus activities too? It prevent support contracts, and make all software patches suspicious (if you can review them or not). This window gets narrower and narrower (and maybe the 3 letter agency changes?), but it doesn’t close or become simple.

    So that’s my pitch for a little state-level actor risk and a contract with lots of penalties and warrant canaries etc. balanced against all the other concerns in these huge decisions.

  5. Matt, well stated. There is a difference between (1) a university using “hardware in the cloud like an AWS EC2 Instance” where the university is the steward of the data and can encrypt it, control backup policies, discard old copies of data, format the “hard drive”, etc. If the the provider of the “cloud hardware” is presented with the subpoena, they can only give out the encrypted data. It is different when the University uses (2) a “cloud service” like Canvas, the university is in no way the “steward/owner” of the data and is completely uninvolved in the backups or retention policies of the data. In scenario (2), the cloud provider can hand over the raw/unencrypted data to the government given the correct paperwork with no effort at all. And when the extra form is provided by the government, the university may never know when/if the data was handed over. I have no problem with scenario (1) – even without owning the physical hardware, the university remains the “owner/steward” of the data. Scenario (2) concerns me when the company is the owner/steward of the data and the university only has access to the data defined by a contract. I think that scenario (2) is a real concern for those outside the US when using with US cloud hosting providers. For those of us in the US while scenario (2) feels “less than ideal” to me – it is more of a distinction without a difference because the US Government can request unencrypted data from on campus servers as easily as they can ask for data from cloud provider servers.

    And lots of non-US schools come up with some mumbo-jumbo rationale that justifies their choice. It is really all about an IT organization more interested in its own convenience than the risks they are exposing their faculty and students to. This is why IT departments at US schools don’t ask faculty if the faculty think outsourcing email is a good thing any more – because faculty would see their privacy more importance than the IT department’s convenience and the IT department (not surprisingly) sees it differently :)

Leave a Reply