How to Choose an LMS – Questions to Ask Before You Sign
The demos all looked great. Every vendor had a clean dashboard, a friendly learner home page and a slide that said "AI-powered." Now there's a contract on the table, a renewal term measured in years, and a nagging feeling that you've seen what each product does best but not what it does on an ordinary Tuesday. Figuring out how to choose an LMS isn't about finding the flashiest platform. It's about finding the one that fits your learners, your admins and your data. So what should you ask, and test, before you sign?
To choose an LMS well, define your requirements before you watch demos, then make every vendor prove those requirements with your own content, your own users and your own reports. Start with who your learners are, what training you deliver, which systems must connect, and who will run the platform day to day. Ask specific questions about standards support, integration, reporting, data export and pricing, and get the answers in writing. Evaluate AI features the same way: ask to see them working on your content, and confirm how your data is used. Finish with a hands-on trial scored against a checklist you wrote before the trial started.
What Should You Decide Before Talking to Vendors?
Decide what problem the LMS is solving and who it's solving it for. Without that, every demo looks good, because each vendor steers toward its strengths.
Write a one-page brief that covers:
- Audiences: employees, contractors, customers, partners, the public? Each may need different logins, branding or pricing.
- Training types: compliance, onboarding, product, leadership, technical certification, instructor-led classes, blended programs.
- Must-have integrations: HR system, single sign-on, video platform, content libraries, ecommerce, reporting tools.
- Standards: SCORM 1.2 and 2004, xAPI, cmi5, Learning Tools Interoperability (LTI) for external tools.
- Admin team: how many people, how technical, how much time they really have.
- Budget and timeline, including implementation, content migration and internal staff time.
Then split requirements into must-have, nice-to-have and don't-care. Be ruthless. A long list of must-haves usually means nobody prioritized.
Which Questions Should You Ask Before You Sign?
Ask questions that require specific answers and, ideally, a demonstration. Here are the ones that separate good fits from expensive regrets.
For learners
- How many clicks from login to launching an assigned course?
- How does it work on a phone? On a shared device? With a screen reader?
- Can you share a current Accessibility Conformance Report, built on the Voluntary Product Accessibility Template (VPAT), for both the learner and admin sides?
- How does search work, and can learners find content by plain-language questions?
For administrators
- Show me how to assign a course to everyone in one role at three locations, with a due date and reminders.
- Show me how to handle a learner who changes departments mid-program.
- Which tasks need vendor support tickets instead of admin settings?
For content and standards
- Which SCORM versions and editions are supported? Is there a built-in Learning Record Store (LRS) for xAPI and cmi5?
- What happens to learner progress when we replace a course package?
For integration
- Which single sign-on standards do you support, and who configures the connection?
- How do user accounts get created, updated and deactivated from our HR system?
- Can completions flow back to HR or a reporting tool automatically? Through an API, a scheduled export, or both?
For data and exit
- Can we export every learner record, transcript and certificate, in a usable format, at any time?
- What does the contract say about data return and deletion when we leave?
- Are API access, extra admins, storage or extra learners priced separately?
Don't let the exit question slide to the end of the list. Your next LMS change will be far easier if you negotiate data portability now, while the vendor still wants your signature.
Who Should Be on the LMS Selection Team?
Put the people who will live with the decision on the team, not just the people who sign it. A small group of five or six works better than a committee of twenty.
- A training or L&D lead who owns the requirements and the final recommendation
- One or two LMS administrators, because they'll use the admin side every single day
- An instructional designer or developer who can test content publishing and tracking
- Someone from IT for single sign-on, security and integration questions
- Someone from HR or HRIS who understands where user data comes from
- A procurement or legal partner for contract terms, especially data ownership
Add a handful of real learners during the trial. They'll notice things the team can't, like a confusing home page or a course player that misbehaves on older phones.
What Does an LMS Really Cost?
The license fee is only part of the cost of an LMS. When you compare vendors, add up the full picture over the whole contract term, not just year one.
- License model: per registered user, per active user, or a flat tier. Active-user pricing can be cheaper for seasonal or occasional learners.
- Implementation and setup fees, including SSO and HR integration work.
- Content migration: moving courses, records and certificates from your current system.
- Add-ons: extra modules for ecommerce, advanced reporting, content libraries or AI features.
- Support tiers: what's included, and what costs extra.
- Internal time: admin hours, IT hours and training for your team.
- Renewal terms: caps on price increases at renewal are worth negotiating up front.
Ask each vendor to price the same scenario, with your learner counts and modules, so the numbers are actually comparable.
LMS vs LXP: Which Do You Need?
An LMS (learning management system) manages assigned, tracked training. An LXP (learning experience platform) focuses on discovery: personalized recommendations, curated content from many sources, user-generated content and social features. If you mostly assign required training and need audit-ready records, you need an LMS. If you want people to explore and build skills on their own, an LXP adds value.
| Need | LMS | LXP |
|---|---|---|
| Assign required training with due dates | Core strength | Limited or none |
| Compliance records and audits | Core strength | Usually relies on an LMS |
| Personalized recommendations | Growing, varies by product | Core strength |
| Curating outside content (articles, videos, courses) | Some | Core strength |
| Social and user-generated learning | Some | Common |
The line keeps blurring. Many LMS platforms have added discovery features, and some LXPs now handle assignments. Don't buy based on the label. Buy based on which of the rows above matter most to you, and test those rows in the trial. The Wikipedia overview of learning management systems gives useful background on how the category developed.
How Should You Evaluate AI Features in an LMS?
Evaluate AI features by the job they do for your learners and admins, with your content, during the trial. AI in learning platforms has become truly useful in a few areas, and it deserves a serious look. Common examples include natural-language search across your catalog, content recommendations, automatic tagging of courses by skill, help drafting course outlines or quiz questions, translation, and admin assistants that answer "who's overdue in the Denver office?" without building a report by hand.
For each AI feature a vendor shows, ask:
- Can I see it working on our content, not your demo library?
- Where does the AI run, and is our data used to train models? What do the contract and data processing terms say?
- Can admins review, edit or turn off what the AI produces, such as tags or generated questions?
- Does it work in the languages our learners use?
- Is it included in the price or an add-on?
Good AI features save admins real time and help learners find the right thing faster. The best implementations keep a person in the loop for anything learners will see, which is simply good content practice. If you want a closer look at what's working, my article on AI LMS features for search and admin walks through the most useful ones.
How Can AI Help You Run the LMS Selection Itself?
General AI assistants such as ChatGPT, Claude or Microsoft Copilot are excellent helpers for the paperwork side of choosing an LMS. They won't pick the platform for you, but they can save hours of drafting and comparison.
- Turn messy stakeholder interview notes into a clean requirements list grouped by must-have and nice-to-have.
- Draft a request for proposal (RFP) and the questions above, tailored to your industry and audience.
- Summarize long vendor responses into a side-by-side comparison table.
- Write step-by-step trial scripts so every vendor is tested the same way.
- Spot gaps: "Which of my requirements did this vendor's response not address?"
Treat the output as a strong first draft. Check every claim in a comparison against the vendor's actual documents, because a summary can blur details, and follow your organization's rules about what can be pasted into which tools. Used that way, an assistant lets you spend your time on judgment calls instead of formatting.
How Do You Run a Fair LMS Trial?
Run the same scripted tasks on every shortlisted platform, with real users, and score them against criteria you wrote before the trial began.
- Pick two or three finalists. More than that and nobody tests thoroughly.
- Load real content: a SCORM course, a video, an xAPI or cmi5 course if you use them, and one instructor-led session.
- Create test users that match your real roles and locations.
- Have admins run the scripted tasks and time them.
- Have a few real learners take assigned training on their own devices and tell you where they got stuck.
- Build the three reports your managers ask for most often.
- Score each platform against your must-haves and record evidence for every score.
If a vendor won't provide a trial environment with your content, that tells you something about how support will feel after you sign.
What Red Flags Should Make You Pause?
Some warning signs show up early if you look for them:
- Vague answers about data export or contract exit terms
- "That's on the roadmap" for a must-have requirement
- Reporting that only works through a separate paid tool
- Pricing that jumps sharply with learner count, storage or API use
- Reference customers who don't resemble your organization at all
None of these automatically rules a vendor out. Each one deserves a direct question and a written answer.
How to Choose an LMS You'll Still Like at Renewal Time
Knowing how to choose an LMS comes down to doing the homework in the right order. Define your audiences, training types, integrations and admin capacity first. Ask specific questions and get written answers, especially about standards, data export and pricing. Judge AI features on your own content, and let AI assistants speed up the drafting and comparison work. Then run a scripted trial with real users.
Your first move is the brief: audiences, training types, integrations, admin capacity and budget, written down before the next demo. Every later step gets easier once you know exactly what you're shopping for.