BLOG – December 2025
Why you can’t afford to delay that data remediation project any longer
AUTHOR: NICK THACKER
Companies worldwide will spend over US$640 billion on artificial intelligence (AI) by 2028, notes IT leader and technology strategist Aparna Kumar. Yet Gartner warns 85% of projects fail to deliver expected value. The culprit, says Kumar, isn’t the AI. It’s the garbage we’re feeding them.
AI can only be as good as the underlying data it works off. And for almost every financial institution, that data foundation is broken. As Kumar points out, 97% of enterprise data is unstructured (in the form of emails, contracts, PDFs, call transcripts, etc.), with over 80% of it duplicated, outdated or irrelevant.
“In highly regulated industries,” she warns, “one hallucinated answer can cost millions.”
Is your data ready?
This data quality problem isn’t limited to institutions’ AI initiatives. It extends across and undermines every business function.
At DCI, we see it time and again. Wealth management firms invest in new tools, automation, analytics, AI … but if the underlying data is inconsistent or inaccurate, everything built on top – from investment decisions and regulatory reporting to client communications and governance – gets compromised.
The priority instead must be on:
- Clean data
- Curated content
- Strong governance
- Continuous monitoring
Fix the data foundation and the whole enterprise becomes stronger and better prepared for growth.
Which is where DCI comes in.
We help firms build the clean, consistent, trusted data underpinnings that the rest of their business depends on.
Our platform continuously identifies incorrect, missing and inconsistent data across systems – and ensures firms fix the root causes, not just the symptoms, so they have real and sustained data integrity. Having a community of users also means that whenever we create rules and enhance the checks on one client’s data, we offer them to any other user that could benefit. These pooled fixes enable everyone to piggyback off each other, gaining solutions for issues they may not have even considered or realised they had.
Don’t put off your data remediation
Despite this data integrity imperative, many industry participants continue to put off much-needed data remediation projects.
“We’re too busy with x.”
X can be a thousand different things.
- We’ve been snowed under with audits.
- Our new CRM system is going live in a few weeks, which is taking up all our attention at present.
- We’re changing custodian.
- We’re conducting a strategic review of systems.
- We are looking to hire a business analyst who will be able to support this process, so will have to wait until they are in place to proceed.
We get it. Teams are busy – often firefighting today’s emergency.
Data observability provider Monte Carlo estimates data teams spend 30%-40% of their time handling checking, remediation and other data quality issues. With time and resources so squeezed, what seem like non-urgent, discretionary costs get put on the back burner. But in the interim (often unrecognised) trouble continues to brew, creating a doom loop of constant emerging problems and reactive fixes. Address the root data quality difficulties though and teams would be freed to work on more productive, revenue-generating activities instead.
Other common reasons we hear firms give for battling on with the status quo include:
“Our data is already in good order”
Unfortunately, that is almost certainly not the case. Practically all wealth managers have data quality problems. Often they run into the thousands, sometimes millions of errors.
Operations chiefs have a choice: either to bury their heads in the sand, or embrace the challenge and start doing something about it before regulatory censure, poor performance or disaffected clients force their hand.
“Our existing vendor systems take care of it”
Unlikely. Some systems may include limited checks – such as some mandatory fields, look-up tables behind field entries and data format checks, for instance on email addresses and National Insurance numbers. But painful experience has shown that any data validation capabilities your existing systems have built in won’t be anywhere near comprehensive enough to catch all your bad data issues. Whenever wealth management firms repurpose fields or create custom ones for their own use (which is often), the system’s validation won’t then cover those either.
“We have an in-house data remediation capability”
Some wealth managers have opted to build their own solutions using in-house extracts written in Microsoft Power BI, Excel spreadsheets and the like.
An effective data quality capability though requires deep domain knowledge and ongoing financial commitment to integrate all the disparate data sources, and identify and tackle each data quality issue that emerges on a daily basis. Relying on an in-house expert or team with the requisite business and system knowledge to build and support a solution, and remedy day-to-day problems, also creates key person risk. Plus, by operating on a standalone basis, firms miss out on the ‘hive mind’ strength that comes from being part of a user community.
All too often staff don’t read the extracts that get emailed around in any case, creating a false sense of security, while flagged problems sit in inboxes and remain unresolved.
Effective data remediation can no longer wait
A clean, consistent, trusted data layer has long been recognised as industry best practice. With AI use cases accelerating rapidly, it has now become a must-have. There is no longer time to delay. Because, ultimately, if your data isn’t right, nothing else will be either.
As Kumar stresses in her article: “The organisations winning this race aren’t waiting. They’re cleaning up their content now. They are building data governance that treats information as their most strategic asset. They are creating cultures where everyone, not just data scientists, is a data steward.”
Your future competitive edge depends on doing the same.
