The Content Graveyard: Why You Keep Making Assets Nobody Uses, and What the AI Glut Does Next

It is 10:40 on a Tuesday night. A rep has a first meeting with a VP of Operations at 9 the next morning, the kind of logo that would make the quarter. She needs one thing: a short, credible case study from a company that looks like this prospect. She knows it exists. She remembers someone presenting it at a QBR in the spring.

So she starts looking. The shared drive. Then the old shared drive nobody deleted. Then Highspot. Then the "Q2 Assets FINAL v3" folder. Then Slack search. Then a message to two colleagues who are asleep. Forty minutes later she gives up and rebuilds a rough version herself in slides, gets it wrong on one number, and walks into the meeting a little less sharp than she wanted to be.

That forty minutes is not a one-off. It is the whole B2B content operation in miniature, and it is about to get much worse.

For CMOs, Sales Enablement Leaders, RevOps, and B2B Founders who keep approving content budgets.

The Number Nobody Wants Framed on the Wall

Somewhere between 60 and 70 percent of the content a B2B organization produces is never used by the sales team it was built for. Forrester put the figure at 60 to 70 percent more than a decade ago. SiriusDecisions, working from its Command Center data, landed near 70 percent. The years pass, the tools change, and the ratio barely moves.

Sit with the money for a second. If two-thirds of what you produce goes unused, then two-thirds of the salaries, the agency retainers, the design hours, and the review cycles behind that content produced nothing a seller ever put in front of a buyer. You did not lose that budget in one bad campaign. You lost it in a thousand assets that quietly went to a folder and stayed there.

The reason is not usually quality, though quality is a factor. The reason is that sellers cannot find the thing when they need it. SiriusDecisions found that 62 percent of organizations have reps hunting through six or more separate places to locate content. Six. By the time you have checked the sixth place, you have already decided it is faster to make your own, which is exactly what the rep at 10:40 did. So the library grows, findability drops, and the next person makes their own too.

Where the Selling Time Actually Goes

Here is the part that connects the content problem to a number your CFO already watches.

Most studies put the share of a rep's week that goes to actual selling at around 30 to 35 percent. The rest disappears into admin, CRM updates, internal meetings, and the search we just described. One frequently cited estimate has reps burning roughly 440 hours a year looking for the right materials to send. That is more than ten working weeks, per rep, spent not selling.

You hired sellers to be in front of buyers. Instead a large slice of them is doing unpaid librarian work, badly, because the library has no catalog. Multiply 440 hours by a fully loaded rep cost and a team of thirty, and the "we don't have enough content" conversation starts to look like the wrong conversation entirely.

More Is Not the Fix. It Never Was.

The reflex, when sellers say they lack the right asset, is to commission more assets. It feels responsive. It is also how graveyards get bigger.

There is a case that keeps getting passed around enablement circles because it is almost rude in its simplicity. One team cut its content library by 76 percent. Not reorganized it. Cut it. Engagement with the surviving content climbed from 4 percent to 55 percent. Removing the clutter did more for usage than years of adding to it, because sellers could finally see what was worth using.

That result is not magic. It is what happens when you stop treating volume as the goal. A library where 4 percent of assets get touched is not a resource, it is noise with a search bar. When you delete the noise, the signal shows up.

None of this argues against investing in enablement. The opposite. Companies with a formal, resourced enablement function report win rates about 49 percent higher and roughly a 4-to-1 return on the program. The winners are not the teams with the most content. They are the teams whose sellers can find the right content in one place, fast, and trust that it is current.

Now Add a Firehose

Everything above was already true in a world where content was made slowly, by people, one asset at a time. That world is over.

More than 70 percent of B2B marketing teams now use generative AI for at least one production workflow, with blog drafts, email sequences, and enablement copy leading the way. By 2026, something like 94 percent of marketers say they plan to use AI for content creation. And the machines do not produce one asset where a person made one. Generative campaign systems are turning out ten to fifty times the creative variant volume of a traditional campaign.

Read that against the 60-to-70-percent problem again. We had a findability crisis when content was expensive to make. We are now making it almost free, and ten to fifty times faster.

The graveyard is not going to grow. It is going to flood.

You can already see the strain. Roughly 68 percent of teams running generative AI at scale hit a quality-assurance bottleneck within the first two quarters, because humans cannot review at the speed the models generate. So the choice becomes ship it half-checked or let it pile up unreviewed. Both feed the same pile.

And buyers are noticing the sameness. LinkedIn's 2025 benchmark work found that organic reach on AI-generated thought pieces fell 31 percent year over year, while engagement with branded original research rose 18 percent. The market is quietly repricing generic content down and distinctive content up, at the exact moment most teams are pointing their new AI capacity at producing more generic content.

What To Actually Do

The instinct in an era of cheap production is to produce. The move that will separate teams in 2026 is the opposite: get ruthless about curation, findability, and distinctiveness. Here is a working sequence.

Audit before you create anything new

Pull usage data on every asset in your library. Most enablement platforms can tell you what has been opened, shared with a buyer, and tied to a closed deal in the last twelve months. Anything with no seller activity in a year is a candidate for the archive. You are not deleting institutional memory, you are moving the dead weight out of the sellers' line of sight so the living content is visible.

Expect the audit to sting. It is normal to find that a small fraction of assets does almost all of the real work. That fraction is your actual content strategy. The rest was motion.

Fix the "six places" problem first

No amount of great content survives being scattered across six locations. Before you brief a single new piece, decide on one source of truth for sales-facing content and make it the only place. If a rep can find the right asset in under thirty seconds from one search, you have solved more of the usage gap than a quarter of new production would.

Put a real cost on production, even now that it is cheap

When AI makes an asset nearly free to generate, the discipline has to come from somewhere else. Add a simple gate: nothing enters the sales-facing library without a named owner, a stated use case ("send after discovery to a mid-market ops buyer"), a review date, and a metric for what "used" means. Cheap to make is not the same as free to keep. Every asset in the library has a carrying cost in findability, and that cost is now the scarce resource, not production capacity.

Aim your AI at curation, not just creation

The same models flooding the library can help drain it. Use AI to tag and summarize existing assets, to match a live deal's stage and buyer role to the two or three assets most likely to move it, and to flag near-duplicates for merging. Pointing your AI budget at helping sellers find and personalize the best existing content will beat pointing all of it at generating the hundredth variant nobody asked for.

Change what you measure

Most content teams still report on output: pieces published, campaigns shipped, coverage of each funnel stage. Those numbers reward the exact behavior that built the graveyard. If you want sellers to use content, measure usage. Track the percentage of the library touched by a seller in the last quarter, the share of closed deals with an asset attached, and the time it takes a rep to find what they need. Put those three numbers in front of the same leaders who currently see the publishing count. The behavior follows whatever gets reviewed in the meeting, so change what shows up in the meeting.

Protect a lane for the distinctive stuff

If generic content is getting cheaper and less effective at the same time, the logical response is to spend the human hours you just freed up on the things AI cannot fake: original research, real customer stories with real numbers, a point of view your competitors would not dare publish. This is where the 18 percent engagement lift lives. Let the machines handle the variants and the drafts. Put your best people on the few pieces that actually make a buyer stop and think.

The Uncomfortable Reframe

For twenty years, the B2B content conversation has been a volume conversation. How many assets, how many blogs, how many nurture emails, how much coverage of the funnel. Volume was hard, so volume felt like the achievement.

Volume is now trivial. A model can produce a quarter's worth of assets before lunch. Which means volume is no longer a strategy, it is a liability you have to manage. The scarce resources going forward are attention, findability, and trust: your sellers' attention, your buyers' attention, and the trust that the asset in front of them is current, relevant, and not the same thing everyone else just published.

The teams that win the next few years will produce less and use more. They will treat their content library like a working tool with a maintenance schedule, not a warehouse that only grows. And they will spend the hours AI gives back on the handful of assets that are genuinely worth a buyer's time.

The rep hunting at 10:40 on a Tuesday does not need you to make her a hundredth case study. She needs to find the one that already exists. Solve that, and you will get more out of your content than any amount of new production ever gave you.

Start with the audit. Look at what your sellers actually opened last quarter. The strategy is already in the data. Most teams have just never looked.