This article is based on an interview with Storyblok CEO and Co-Founder Dominik Angerer and Greg Kihlström for The Agile Brand Guide’s Expert Mode series.
Most marketing teams think their old content is handled. The 2019 pages came out of the navigation two reorgs ago. The dead campaign URLs got pulled from the sitemap. Whatever’s left lives in a spreadsheet tab labeled “archive,” and nobody has opened it since.
Dominik Angerer, CEO and Co-Founder of Storyblok, thinks “handled” is doing work it hasn’t earned. In written responses for Expert Mode, he argues that the way content goes bad has almost nothing to do with anyone’s decision, and that the controls marketing teams rely on to bury old pages were only ever hiding them from one audience. Storyblok’s research with FT Longitude, part of the Financial Times, puts the enterprise cost at $4.63 trillion globally, an average of $663.4 million per company. Read that number with the vendor’s interest in plain view — Storyblok sells the platform the study concludes companies need, and the same research is packaged as a calculator and a “recovery plan” download. The mechanism he describes is still worth understanding, because the mechanism doesn’t depend on the number being right.
Debt Nobody Chose to Borrow
Technical debt usually describes a trade somebody made on purpose. Ship it now, clean it up later, and everyone in the room knows the bill is coming. Content debt doesn’t work that way, and Angerer doesn’t pretend it does. He describes something closer to erosion, where the original decisions were all defensible and the accumulation happened anyway.
No one sits down and decides to create thousands of outdated pages. Content just accumulates. Product changes, a new page gets published, an old one stays live, information gets duplicated, someone leaves, and eventually a company has a huge amount of content that’s difficult to keep accurate and causes more business impact than executives realize.
That distinction matters for how a marketing leader assigns the problem. Debt with a borrower has someone to send the bill to. Debt that accrues through turnover and reorgs has no such person, which is exactly why it survives every budget cycle — nobody’s name is on it, so nobody defends the line item to fix it. In the same research, 78% of executives said their organization carries more digital content than it can realistically keep accurate. Angerer keeps the word “debt” anyway, and his reason is practical rather than rhetorical: it comes with a sequence. Inventory what you have, decide what needs attention, rank it, pay it down. He also notes that AI is what made the balance visible. Information that sat unread on an old page can now surface in a generated answer or a product recommendation.
De-Indexed Is Not Deleted
The specific failure worth walking through: a customer asks a chatbot about your return policy and gets an answer from a page you retired years ago. Asked what has to be true for that to happen, Angerer went to the plumbing rather than the model. The page has to still be reachable, and taking a URL out of the sitemap doesn’t make it unreachable.
The problem is that hiding a page from Google doesn’t necessarily make the information disappear. A page can be removed from search results and navigation while the URL still works and the content still exists somewhere where an LLM can find it. So the page may look gone to the marketing team while the underlying information is still available somewhere.
An old backlink still resolves. So does an archived copy, a third-party republication, or a PDF someone on the sales team downloaded in 2023. We’ve spent a decade treating de-indexing as removal because search was the only distribution channel that mattered, and for search it worked well enough. Angerer’s Monday-morning prescription has two parts, and neither needs a budget approval. First, audit your own domains for live pages that appear in no sitemap, no navigation, and no content inventory. Second, take the questions customers actually ask — returns, pricing, warranties, product specs — and run them through the major AI tools yourself. If an answer comes back sourced from a page you’d forgotten, you’ve found the gap. From there the choice is update, redirect, or actually remove, which is not the same as hide. Anyone doing serious generative engine optimization work should recognize the first audit as table stakes, and most teams still haven’t run it.
The CMS Is the Last Move, Not the First
The most commercially useful number in the study for Storyblok is that 69% of executives call content strategy more of a technical challenge than a creative one. It points directly at the tech stack, and Storyblok is a tech stack vendor. Asked to argue the opposite case — where replacing the CMS is the wrong move — Angerer took the question rather than deflecting it.
The wrong move is putting a new platform in place before answering some basic questions: Who owns this content? Who can retire it? What gets reviewed, and how often? What happens when a product or policy changes? … If those basic questions don’t have clear answers, a better CMS can actually make the problem worse. You just have a faster, cleaner way to create more content.
He extends it to incentives without being prompted. If teams get rewarded for publishing volume rather than for maintaining or retiring what’s already live, the CMS isn’t what’s broken. And if the organization can’t measure whether its content works at all, new software won’t supply that either. For a CMO weighing a migration, this is the useful test, and it’s one you can run without the vendor in the room: name the person who can unilaterally retire a page, and name the review cadence that would have caught the 2019 return policy. If those two answers don’t exist, a platform decision made now buys a faster way to make the same mess. Angerer’s own framing is that technology helps only after the organization has decided where the debt lives and who’s responsible for it.
What Actually Comes Off the Budget
The spending picture is where the pitch stops being abstract. The research puts average remediation spend at $4.8 million a year, which works out to 34% of total content budget, alongside 105.4 hours a week maintaining existing content. A third of the budget goes to cleanup, and the cleanup isn’t finishing. Angerer characterizes most of those hours as reactive work — someone finds a bad URL, a customer gets wrong information, a product changes, and a team scrambles.
I’d stop treating content audits as one-off cleanup projects. An audit should give you a clear picture of what you have, what’s outdated, what’s duplicated and what needs to be removed. But if you do that once a year and then go back to the same process that created the backlog, you’ll be doing the same cleanup again next year, costing the same amount of business impact.
So the line to cut isn’t the audit. It’s the annual audit as a standalone project, followed by a return to the process that produced the backlog. What replaces it is continuous inventory plus the governance to keep it accurate. For measurement, Angerer names the remediation share itself: if the work is paying off, the 34% should come down year over year. He pairs it with a coverage metric that’s easier to move in the first quarter — how much of your live content has a named owner and a review date. The research also reports that organizations with the highest content confidence are less likely to say their systems limit responsiveness (38% versus 59%) and more likely to exceed financial targets (64% versus 46%). Correlation, not proof, and it’s a vendor’s survey. But the first of those two gaps is large enough to be worth a look at your own numbers.
Back to the archive tab nobody opens. The claim isn’t that it’s costing you $663 million, and no CMO should walk into a budget meeting with a vendor’s modeled figure as the argument. The claim is narrower and harder to dismiss: pages you believe are gone are still reachable, still being read by systems your customers now ask first, and still speaking in your brand’s voice about a policy you changed three years ago. Two moves this quarter, neither requiring a migration. Crawl your own domains and inventory everything still live, starting with whatever carries compliance or customer risk. Then give what survives an owner and an expiration date. The number to watch is what share of content budget still goes to fixing old content — and whether, twelve months from now, it’s smaller.
