MARKET RESEARCH
The Market Need for Repository Health
What current research says about AI and unready document content, and where the AITS Content Tracking System (CTS) fits.
of AI projects unsupported by AI-ready data are expected to be abandoned through 20261
of organizations say their unstructured content is somewhat or fully ready for AI2
say less than half of their unstructured data is discoverable and usable for AI3
ON THE AITS WEBSITE
Solution Brief: Is Your Repository Ready for AI?: The Repository Health Check in two pages.
Content Tracking System (CTS): The platform behind the Repository Health score.
SUMMARY
AI is being funded faster than content is being measured
Organizations are investing in AI faster than they are measuring the content it will read. Structured data has had years of attention. Documents, scans, emails and records have not, and they are what most AI assistants and search tools actually work from.
This page gathers what independent research published in 2025 and 2026 says about that gap, how analysts size the market around it, and what kinds of product already address it. It then sets out where the AITS Content Tracking System (CTS) fits.
The research supports three points
- 1The need is measured, not assumed. Independent 2025 and 2026 research agrees that document content is the least AI-ready data organizations hold.
- 2The market is real and growing. Analysts size unstructured data management and governance in the billions of dollars, growing 15% to 24% a year.
- 3One position is still open. A health score that an executive defines, that covers every repository, and that keeps running after the assessment.
Research current to October 2026. Every figure is linked to its source at the end of the page.
01 / THE NEED
What the research says
AI projects fail on data
Gartner predicted in February 2025 that through 2026, organizations will abandon 60% of AI projects that are not supported by AI-ready data. Its survey of data management leaders found that 63% either lack the right data management practices for AI or are unsure whether they have them.1
Document content is the weak spot
Three separate surveys published in 2026 reach the same conclusion from different directions.
| Finding | Research |
|---|---|
| 65% say structured data is somewhat or fully ready for AI. Only 39% say the same of unstructured content such as emails, PDFs and images. | Hyland survey of AI decision makers, May 20262 |
| 70% say less than half of their unstructured data is discoverable and usable for analytics or AI. | BARC study of 225 data, analytics and AI leaders, 20263 |
| 94% say they struggle to manage unstructured data effectively. | Nasuni survey of 1,000 enterprise purchasing decision makers, 20264 |
Buyers already ask for what a health score measures
Asked how they judge whether unstructured data is ready for AI, the leaders in the BARC study named the measures below.3 These are properties of the content itself, and they are the kind of KPI a Repository Health score is built from. The same study found that human validation still ranks as a top readiness measure, which is why CTS keeps people checking the AI’s work.
| Readiness measure | Share naming it |
|---|---|
| Accuracy | 54% |
| Consistency | 48% |
| Business relevance | 44% |
| Completeness | 39% |
| Timeliness | 31% |
Rollouts are stalling in practice
A Gartner survey of IT leaders found that overshared data caused 40% of organizations to delay the rollout of a workplace AI assistant by three months or more.5 In a separate survey, 86% of organizations had done some content cleanup to prepare for AI, but only 51% had done it across the whole organization.6
Cleanup is happening in pieces, and without a baseline to measure it against. An organization that cannot say how healthy its content was before the cleanup cannot show what the cleanup achieved.
02 / THE MARKET
How analysts size it
No analyst sizes content health scoring as a market of its own. It sits inside unstructured data management and governance, which analysts put between roughly 4 billion and 9 billion US dollars in 2025, growing 15% to 24% a year.
| Market | 2025 size (USD) | Forecast (USD) | Annual growth |
|---|---|---|---|
| Unstructured data management7 | 9.2 billion | 28.5 billion by 2033 | 15.3% |
| Unstructured data governance8 | 3.79 billion | 10.99 billion by 2030 | 23.8% |
| Unstructured data solutions, the broadest definition, including storage and analytics9 | Not stated in the public summary | Adds 39.68 billion between 2025 and 2030 | 15.2% |
Figures are from the public summaries of each firm’s report. The firms draw the market boundary differently, so the rows are not additive.
These figures show direction more than precision. What they agree on is the rate of growth, and what is driving it. The governance segment is the closest match to repository health and the fastest growing, and the report that sizes it names unstructured data quality and AI readiness as its leading trend.8
03 / WHAT IS ON OFFER
Four kinds of answer, each with a limit
Scoring content for AI readiness is no longer a new idea. The offerings available today fall into four types, and each is tied to one platform, one audience or one moment.
| Type of offering | What it does | What it leaves open |
|---|---|---|
| Platform-native readiness checks | Built into a single collaboration platform. They report oversharing and readiness for that platform's own AI assistant, and some give one readiness score. | One platform only. The vendor sets the weights, not the client. |
| Discovery and cleanup tools | Find redundant, obsolete and sensitive content across file stores, usually ahead of a migration or a cleanup. | Built around a cleanup project, not an executive's ongoing measure of health. |
| Data quality platforms adding documents | Monitor continuously and score individual documents before they feed an AI pipeline. | Built for data teams and data lakes, not for records repositories. |
| Consulting-led assessments | A one-time AI readiness review, delivered as a report. | A point in time. Nothing re-scores the content as it changes. |
A single score is therefore not what sets one product apart from another. What matters is whose score it is, how much of the organization’s content it covers, and whether it is still true next month.
04 / WHERE CTS FITS
The executive’s own score, for every repository, kept current
CTS gives an executive a single Repository Health score: an at-a-glance answer to a question that used to take a sit-down with BI charts and reports. A high score is grounds for confidence. A low one means there is work to do, and the breakdown behind the score says what that work is.
Four things set it apart
- 1The client defines and weights the KPIs. Other scores use the vendor's formula. A composite built from the measures an executive already uses to judge health is easier to trust and to defend.
- 2It works with any repository. Most organizations hold content in several systems, and more than half of the databases holding unstructured data are on-premises or hybrid. A check tied to one platform cannot see the rest.3
- 3It is continuous, and it learns. When staff confirm, correct or dismiss a finding, that decision sharpens the model. A one-time assessment stops at the report.
- 4It suits records of legal weight. Scanned, historical and registry records are harder than office documents. Even a national land registry reports that handwritten and poorly scanned records remain difficult for automated systems.14
The official repository is never touched
CTS runs on a replica of the organization’s environment in a secure cloud environment hosted in-country. Nothing it does is destructive, and any result can be compared against the original. That answers the residency and risk questions public-sector buyers raise first.
People stay in charge
CTS recommends action on each issue it finds. Once it has learned how the organization resolves a type of issue, it can be given permission to handle future issues of the same kind, with staff checking its work.
05 / REGISTRIES AND REGULATED RECORDS
Where the need is sharpest
The pressure is greatest where records carry legal weight and reach back decades. Modernization money is being spent in these areas now, mostly on systems and channels.
Land titles
Registries are moving from paper to online systems of record, with electronic submission and new legislation to support it. Those programs fund systems and channels. The measured health of the records behind them is usually not in scope, and a new system that fills forms from existing title data makes the quality of that data visible.
Vital events
Demand on historical records is rising. One Canadian province put more than one million historical vital statistics records online in 2026 and reported a surge in requests from people needing documents to support citizenship applications. Many of those records began on paper and film.11
Human resources
The driver is exposure. An AI assistant surfaces whatever a user can technically reach, and HR records and pay details are among the first things to appear when permissions were never cleaned up.13
Proof that AI works on registry records, with people checking it
HM Land Registry reported in March 2026 that its AI tool processed one local authority’s land charges records in four weeks with four staff, against an estimated three months with 20. Every batch passed its data quality checks the first time, and a quality assurance team still verifies the output.1014
Policy is pointing the same way
Canada’s AI Strategy for the Federal Public Service 2025-2027 sets up an AI Centre of Expertise whose stated role includes supporting data readiness.12 Readiness of the underlying information is now part of how government expects AI to be adopted.
AITS WHITE PAPERS
Land Titles Registry | Vital Events Registry | Human Resources
Find out your repository health
A Repository Health Check gives an executive a scored baseline, the detail behind it, and a prioritized list of what to do next. CTS monitoring keeps that baseline true as content changes.
Contact AITS to book a Repository Health Check info@appliedits.comON THE AITS WEBSITE
Solution Brief: Is Your Repository Ready for AI?: The Repository Health Check in two pages.
Content Tracking System (CTS): The platform behind the Repository Health score.
ABOUT AITS
Applied IT Solutions (AITS) is a Toronto-based consultancy founded in 2008 and led by its principal consultant. It advises government and Fortune 500 organizations on artificial intelligence, enterprise data strategy, document and records management, and business process improvement, with one aim: deriving tangible business value from the application of information technology. Most engagements are delivered end to end by the principal, from requirements gathering through solution design and delivery to train-the-trainer and hand-off to the client's own team. For larger projects, AITS brings in experienced consultants it works with regularly. AITS has run projects worldwide and works in any time zone. Out of this work, AITS developed the Content Tracking System (CTS), an AI-powered monitor that scores the health of a document repository against the organization's own KPIs.
SOURCES
- Gartner, "Lack of AI-Ready Data Puts AI Projects at Risk," Feb. 2025; survey of data management leaders, July 2024. View source
- IT Brief, "Enterprise AI ambitions outpace readiness, survey finds," May 2026; reporting Hyland research. View source
- IDM, "Most Unstructured Data Not Ready for AI, Study Finds," 2026; reporting the BARC study "Harnessing Unstructured Data for AI Innovation." View source
- Unite.AI, coverage of Nasuni, "The State of Enterprise File Data Annual Report 2026." View source
- IDM, "Copilot Gets Sensitive Data Guardrails in Purview," 2026; reporting a Gartner survey of IT leaders. View source
- ShareGate, survey of IT leaders on AI governance and content cleanup, 2026. View source
- DataHorizzon Research, "Unstructured Data Management Market," forecast 2026 to 2033. View source
- The Business Research Company, "Unstructured Data Governance Market Report 2026." View source
- Technavio, "Unstructured Data Solution Market Analysis, Size, and Forecast 2026-2030." View source
- HM Land Registry, "Transforming local land charges migrations with AI," March 2026. View source
- Government of New Brunswick, "Provincial Archives putting history within easy reach," 2026. View source
- Canadian Government Executive, summary of the AI Strategy for the Federal Public Service 2025-2027. View source
- CoreView, "Best Practice for a Microsoft Copilot Readiness Assessment," Aug. 2026. View source
- Results Sense, "HM Land Registry AI tool cuts land charges migration from months to weeks," March 2026. View source
Market figures are in US dollars. Survey figures are as reported by each source.