It is an uncomfortable question for an L&D function that has spent years building, buying and curating a library. But for most large enterprises the honest answer is yes; and the volume itself is not the real problem. The problem is that almost none of it is structured well enough for anyone, or anything, to use with confidence.

The symptoms are familiar

You probably recognise the pattern. Thousands of SCORM packages accumulated across years and acquisitions. Multiple content libraries from multiple vendors, with overlapping coverage. Internal knowledge spread across systems nobody fully maps. And underneath it all, a set of questions L&D genuinely cannot answer quickly: Which content is actually good? What is being used? What is duplicated three times over? What is quietly out of date and still being served to learners?

When GSK began rationalising its estate, it was looking at roughly 20,000 internal SCORM packages. When DLA Piper prepared for a platform migration, it had around 50,000 learning assets to assess and tag. These are not unusual numbers for a global enterprise. They are normal; and they are unmanageable with the tools most L&D teams currently have.

The hidden cost of content you cannot see

Content overload is expensive in ways that rarely show up on a single line of the budget. You pay to license it. You pay to host and maintain it. You pay, indirectly, every time an employee cannot find the right material and a manager fills the gap with something improvised. And you pay the largest cost of all in lost return: enterprises invest enormous sums in learning content, then have no reliable way to see which of it is delivering value.

This is what makes “content ROI is invisible” such a persistent phrase in enterprise L&D. It is not that the ROI is bad. It is that the estate is not structured well enough to measure it at all.

The traditional fixes do not scale. A manual content audit takes months of specialist time and is out of date the moment it finishes. External consultants are expensive and leave no permanent capability behind. And adding an LXP on top simply layers another interface over the same unstructured pile; it does not make the underlying content any more legible.

From more content to intelligent content

The shift that actually changes the economics is moving from managing the volume of content to understanding the quality of it.

That means three things working together. First, cracking content open to module level, so what it actually teaches is readable; not just its title in a catalogue. Second, mapping every asset to a skills framework, so content is organised around capability rather than format or vendor. Third, scoring every asset continuously for quality, relevance and duplication, so the estate effectively maintains a live view of itself.

Done well, the results are significant. AstraZeneca, having gained a single view of its content estate for the first time, identified substantial overlap and reduced content spend considerably. Across deployments, the pattern repeats: organisations discover that a large proportion of their library is redundant, outdated or unmapped; and that focusing investment on the content that is genuinely relevant delivers a far better return than continuing to grow the pile.

The better question

So “do you have too much content?” is the wrong question to end on. Every large enterprise has more content than it can manually govern. The question that matters is sharper: how much of your content can your people; and now your AI tools; actually find, trust and use?

If the honest answer is “we are not sure,” that is not a content problem. It is a structure problem. And it is solvable: not by buying more content, and not by buying another portal to put in front of it, but by making the estate you already own intelligent; scored, mapped, and finally legible to the systems and people that depend on it.

The enterprises pulling ahead in 2026 are not the ones with the biggest libraries. They are the ones who know exactly what is in theirs.