The New Economics of Learning: Making Smarter Build vs Buy Decisions in an AI World

For years, one of the biggest considerations in deciding whether to build or buy learning solutions has been the time and cost involved. Building something ourselves takes time, requires specialist skills and can be expensive. Buying an existing solution is often quicker and easier. But what happens when AI fundamentally changes that equation? Learning teams can now develop outlines, draft content, create activities, adapt materials and produce digital learning experiences in a fraction of the time these tasks previously required.

While this doesn’t mean creating learning is free, or that AI-generated content is automatically good, it does mean that some of our traditional reasons for outsourcing or purchasing learning solutions deserve closer examination.

Every L&D team faces decisions about whether to develop learning internally, curate existing content, customise a solution or partner with an external provider.

But before we ask what we can afford to buy or whether we have time to build, we should be asking what capability the organisation needs, who needs it and what kind of learning experience will genuinely address that need.

When I’m faced with a build, buy or partner decision, I work through six key questions. In this article, I’ll share those questions and demonstrate how I apply them in practice.

Six questions to guide your build vs curate decisions

1. Strategic importance: How critical is this capability?

Not all learning needs carry the same strategic weight. Some capabilities are fundamental to an organisation’s competitive advantage, culture or ability to deliver its strategy. Others are more generic skills that employees need to perform effectively.

When a capability is strategically important, there may be a stronger case for investing in a solution that is tailored to the organisation.

That doesn’t automatically mean building it internally. But it does mean we should be particularly deliberate about how the capability is developed.

2. Scale and access: How many people need it?

Scale changes the economics of learning. A programme needed by ten employees may warrant a very different investment approach from one needed by thousands.

We also need to think about access. Is this something a small group of specialists requires, or a capability that should be available to everyone?

However, scale alone shouldn’t determine the decision. A large audience can still be well served by curated content.

3. Contextual relevance: How specific does the learning need to be?

This is one of the most important considerations for me.

How much does the learning need to reflect our organisation, industry, culture, region or audience?

4. Quality: Does the solution meet the standard we need?

The availability of content doesn’t necessarily mean it’s good content. We need to evaluate the quality of the learning experience, not simply the accuracy or appearance of the material.

Does it create opportunities for practice? Is it engaging and accessible? Will it help learners apply what they’ve learned?

This is particularly important in an AI-enabled world, where producing polished-looking content has become remarkably easy.

A professional-looking programme is not necessarily an effective learning intervention.

5. Longevity: How long will the content remain useful?

Some learning content has a long shelf life. Other content needs frequent updates as products, processes, regulations or technology change. This influences the economics of our decision.

If we expect to use a programme repeatedly over several years, investing in an internal asset may make sense.

If the content changes frequently, we need to consider how easily it can be maintained and who will be responsible for keeping it current.

6. Market availability: What’s already out there?

Finally, we need to understand what the market can offer.

Sometimes an excellent solution already exists, and building something similar would be an unnecessary use of our resources.

Sometimes existing content is good but needs customisation.

And sometimes the market doesn’t offer anything sufficiently relevant, making a bespoke solution the stronger option.

The question isn’t whether we can build something. It’s whether building it is the most sensible strategic choice.

From framework to decision

I deliberately describe these six considerations as a decision matrix rather than a formula. I don’t believe we should count how many boxes we’ve ticked and declare that building or buying has won. The factors need to be weighed together.

A highly specialised capability, with limited market availability and a long useful life, may justify an internal build.

A generic capability needed by thousands of employees may be better served through curated content.

Strong existing content that lacks cultural or organisational relevance may be best addressed through curation and customisation.

And sometimes the right decision is to partner with a specialist provider.

The objective isn’t to build everything ourselves. It’s to make deliberate decisions about where different approaches will create the most value.

Of course, we can’t ignore cost. Once we’ve identified the strategically suitable options, then we need to evaluate their economics.

But we shouldn’t allow the price tag we see to determine our learning strategy.

Applying the framework: Building AI literacy

Let me share an example of how I’ve applied these questions in practice.

At a leading retailer, we identified AI literacy as an important capability for our employees. In my division alone, we had approximately 500 people who could benefit from developing these skills.

Rather than immediately selecting a learning solution, I considered the six questions.

Strategic importance: AI literacy was becoming increasingly important to how our people would work. This wasn’t a specialist capability for a select few.

Scale and access: With approximately 500 employees in my division, I wanted a solution that could be made available to everyone, rather than limiting access to a small group.

Contextual relevance: Much of the available AI learning content was generic. I wanted something that connected the concepts to our industry, business environment and employees’ roles.

Quality: The programme needed to provide a meaningful introduction to AI and help employees understand how they could apply what they learned in their work.

Longevity: AI was evolving rapidly, so I needed to consider how we could keep the learning relevant as the technology developed.

Market availability: There were already good programmes available. I didn’t believe we needed to create everything from scratch, but I did believe we needed something tailored to our audience.

With these considerations in mind, I evaluated three possible approaches.

Build? Not necessarily. AI literacy was a relatively broad capability, and suitable content already existed in the market. Developing everything from scratch would have meant investing time and resources in recreating material that was already available.

Buy? Yes, but not simply purchasing an off-the-shelf programme. The content needed to be relevant to our employees and accessible at scale. I needed a solution we could adapt to our context and make available through our own learning platform.

Partner? Bringing in an external provider to facilitate training sessions was another possibility. However, this would make reaching approximately 500 employees more challenging, particularly if we wanted everyone to have access to the learning.

The most appropriate approach was therefore to buy and customise.

I investigated the cost of sourcing an existing programme, adapting it to our needs and hosting it internally. The solution I selected involved a once-off investment, including customisation and perpetual usage rights.

Approximately 500 employees could access a programme relevant to their work.

The decision wasn’t simply about choosing the cheapest option. It was about selecting the approach that best addressed our capability need, and then ensuring that the economics made sense.

The investment L&D teams should be making

AI presents an incredible opportunity for L&D teams. I believe that every L&D team, regardless of size, should be developing the capability to create at least some of its own digital learning solutions.

I see three areas of investment.

Learning design strategy. The ability to diagnose capability needs, understand business priorities and select the most appropriate learning response.

Instructional design. The ability to create effective learning experiences that support practice, application and transfer, rather than simply producing attractive content.

Technical capability. The ability to use AI and digital learning tools effectively, critically and responsibly.

When these capabilities come together, L&D teams have more choices. They can respond more quickly, create contextually relevant learning experiences and reduce unnecessary dependence on external development.

This doesn’t eliminate the need for providers. In fact, I believe specialist providers will remain incredibly valuable, particularly when they bring expertise, experiences or outcomes that internal teams cannot reasonably replicate.

But the value proposition is changing. As internal teams become more capable, generic content development becomes harder to justify at a premium. The strongest providers will be those who offer distinctive expertise and genuine value beyond what organisations can create themselves.

Making the next decision differently

The next time a learning request lands on your desk, resist the temptation to immediately search for a course or open your authoring tool.

Start with the capability the organisation needs.

Consider its strategic importance, the scale of the need, contextual relevance, quality, longevity and what’s available in the market.

Then evaluate the economics of the most suitable approaches, including how AI might change the cost and effort of creating or adapting a solution.

And if your team doesn’t currently have the skills to execute the option you believe is best, consider whether that’s a capability worth developing.

Because when you do this, the investment isn’t only in the next programme. It’s in your team’s ability to respond to the many learning needs that will follow.

Not building should be a strategic choice, not a capability limitation.

AI hasn’t changed the need for sound learning strategy. It has changed what skilled L&D professionals can do with the time and resources available to them. And that gives us an opportunity to make better decisions about where we invest, what we create and when we partner.

About the author

I’m Sheree Comninellis. I work in corporate Learning & Development, with a focus on building enterprise capability. I write and speak about learning, leadership and change.

Have a perspective to share or want to continue the conversation? Connect with me on LinkedIn or get in touch through my website.