Data quality: why "the data arrived" is not the same as "the data is right"

That question, is this data right, is the one most operations teams answer manually, every day, without ever being asked to. This article looks at the layer beneath everything else we have covered so far: data quality itself. 

FE fundinfo's 2026 Asset Managers Report shows why this matters beyond one firm's operations desk. 65%t of respondents say fragmented fund data is preventing their organisation from improving operational efficiency. Fragmentation is not only about where data sits, but also whether the version in front of you can be trusted without being checked against another source. 

The problem: quality gaps are invisible until someone finds them 

Bad fund data rarely announces itself. A share class identifier that has not been updated, a NAV that lags a corporate action, a static field copied forward from a previous quarter;, none of these break a data feed. They just make it wrong, silently, until someone downstream notices a number that does not reconcile. 

By then, the cost has already moved. What started as a data quality gap becomes a client query, a reconciliation break, or a compliance question about why a disclosure did not match the underlying holding. The error is rarely caught at its source. It is caught several steps away from it, by whoever happens to be looking closely enough at the right moment. 

This gets harder to manage as coverage grows. A distributor pulling data on tens of thousands of funds across dozens of jurisdictions is not validating one dataset. It is validating as many data quality standards as it has source relationships, each with its own update cadence, its own conventions, and its own blind spots. 

The consequences: manual checking becomes the real cost centre 

The direct response to unreliable data is more validation. Teams build reconciliation spreadsheets, add manual review steps, and hold data back until someone has eyeballed it. Every one of those steps is a cost, and none of them is the reason the team exists. 

69% of respondents to FE fundinfo's 2026 Asset Managers Report said the speed and accuracy of fund data are becoming more important differentiators in winning and retaining distribution partners. A distributor that is still manually checking data before it can act on it is losing on both counts at once. The data arrives more slowly because of the validation checks, and any error that survives the check still arrives inaccurate. 

There is a second cost that shows up further downstream. 64% of asset managers believe AI will only deliver meaningful value once firms first improve the quality and structure of their underlying data. Every model, every automated report, and every client-facing tool built on top of fund data inherits whatever quality problems sit underneath it. Poor data quality becomes the ceiling on what the rest of the business can safely automate. 

Regulatory exposure follows the same pattern. Data governance under MiFID II, PRIIPs and SFDR is judged on the accuracy of what a firm discloses, not on how many manual checks happened before publication. An error that reaches a client-facing disclosure is a governance failure regardless of how many spreadsheets tried to catch it first. 

The solution: quality controls built into the source, not bolted on after it 

The alternative to checking data after it arrives is sourcing data that has already been validated. That means quality controls applied at the point of collection, not as a separate downstream task assigned to whoever receives the feed. 

Through FE fundinfo's Data Feeds, fund data is collected directly from investment managers and validated through automated workflows before it reaches your systems, rather than being checked for the first time once it lands. Standardisation against industry formats such as openfunds and FinDatEx means every field follows the same convention regardless of which investment manager it originated from, so a distributor is not reconciling formats as well as figures. 

In practice, this looks like static data and regulatory documents, including EMT, EPT and EET templates, maintained centrally and refreshed as changes happen, rather than carried forward until someone notices they are out of date. It looks like one governed, quality-controlled source feeding your CRM, fee engine or distribution platform, instead of separate teams each holding their own version of the same fund's data. 

This matters most at scale. With coverage of more than 100,000 active funds and 300,000 share classes across 75 or more jurisdictions, FE fundinfo's direct relationships with investment managers are what make consistent quality control possible in the first place. Validating data at source, once, is the only way to keep quality standards intact as coverage keeps expanding. 

Why this is a distribution problem, not only an operations one 

Distributors compete on how quickly and confidently they can put fund data in front of clients and platforms. Every manual check inserted between source and client is time that competitor is not spending. Every error that gets through is a piece of trust that does not come back easily. 

Data quality is not a background operations concern sitting behind the more visible parts of a distribution strategy. It is the precondition for all of them. Coverage only helps if the data behind it is right. Timing only helps if what arrives on time is also accurate. Embargo controls only protect an asset manager's information if the information itself is correct. 

Where this leaves you 

Data quality is not something you can bolt onto a feed after the fact. It has to be built in at the point the data is collected, or every downstream user is left checking someone else's work. 

We go into this in full in our latest whitepaper, The Distributor's Guide to Fund Data Excellence, which sets out what distributors should expect from a quality-controlled fund data source and how to assess whether their current setup measures up.