The Quality Data Management Imperative
By John Toepfer
(The following blog post is the third in a series on the need for the Investment Management Industry to embrace sound data management practices. You can access the first and second posts here.)
Earlier today I was on a conference call with one of the largest asset management firms in the world. The call was with marketing people representing operational groups in London, Hong Kong, Vancouver and New York and the purpose of the call was to discuss fact sheet automation and the need to unify their operations. The interesting thing was that we could hardly stay on topic about documents and automation tools because their real problem was source data.
During the call, each group expressed their internal challenges in collecting clean, verifiably accurate and finalized data for reporting use.
As a provider of automation solutions for marketing communication needs in this industry, we are on the receiving end of the data supply chain for a large number of investment management companies. In fact it’s one of our primary stocks in trade; the ability to collect source data in whatever format it’s available, turn it into a clean reporting data warehouse, and then use it flexibly to drive a wide variety of communication efforts.
What is the Status Quo?
When I started in this industry more than 20 years ago, I immediately learned it was unrealistic to expect asset management firms to have a clean data warehouse to feed marketing and reporting processes. Data in this business lived in a combination of spreadsheets created on people’s desktops and in frequently obtuse and overlapping dumps from various internal and external systems.
That’s just the way it was and, in fact, it’s still the state of the art at many firms today.
Things have improved somewhat in recent years, however. These days, the most common data story we hear from a new client is that they do have a data warehouse, but it’s not complete. And we can expect to get maybe 65% of the data we need for marketing and reporting purposes from that system. The rest will come by way of — you guessed it — Excel spreadsheets.
Regardless of the storage and processing techniques (databases, warehouses, Excel) the most shocking thing we find missing from the asset management industry’s data processes is not a type of data or a piece of technology, it’s a comprehensive quality data management initiative.
Related: Interview with Data Quality Expert, Maria C. Villar. What is Data Quality and Why is it Important for Marketers?
While terms like “data quality”, “data management”, and “data governance” seem self-defining and even interchangeable, in most companies they are treated as token terms to hang on just about any database project. It turns out however that there is actually a well refined and formal set of tools, standards, and practices in these areas that an enterprise can embrace if they truly seek to improve their data-related business risk-factors.
As an example; Data governance refers to the overall management of the availability, usability, integrity, and security of the data employed in an enterprise. A sound data governance program includes a governing body or council, a defined set of procedures, and a plan to execute those procedures.
If you’re part of the asset management or investment marketing world, my guess is that this term and concept is almost completely new to you.
Marketers as Data Aggregators
The company I was speaking with on that conference call was no different than most in the industry. The operating expectation is that the marketing operations group will collect data from multiple systems and vendors, consolidate and normalize the data as may be necessary, somehow verify its accuracy and then move those numbers into documents.
As with most firms, their concept of enterprise data quality and data governance is practically non-existent. Oh, I’m sure that deep in the enterprise IT and risk management groups they use the term to justify various expenditures, but on the street where the marketing department lives, they get their data in haphazard forms and formats. This leaves the onus on the marketing team to figure out what is the authoritative information and how to QA it.
The Trickle-Down Effect of Bad Data
In the last two decades, we’ve encountered very few internal IT organizations that provide truly quality-controlled data with a real data ownership and stewardship concept to support the firm’s vital marketing and sales communication functions.
They support quality data for statements; holdings, trades, shares and values, but support for other marketing and reporting functions is haphazard at best. Once the trades are closed and the accounts balanced, it can be extremely hard to find someone who will take actual ownership for the extended reporting data sets of holdings, composition, categorizations, statistics, expenses and the myriad ways that the data has to be sliced and diced and reported.
One fund company recently related to me that their internal data warehousing team was going to cease providing bespoke reports on data to them as a service. The marketing group would then have to go to a data warehouse web interface and create their own reports. Sure, this kind of self-serve reporting is technically feasible, but it’s completely lacking in auditable controls and ownership. Or, as the marketing director put it, “We have no idea what data we’re pulling or if it’s final, complete, or will be the same if we pull it again tomorrow!”
This problem gets magnified with any firm or product line that is selling investment products more complicated than basic funds and ETFs. Strategy products which blend and combine investments from many different sources and of many different types tend to expose the biggest challenges in the industry related to data aggregation, quality control, and reporting.
If you are a compliance officer or regulator, yes, you’re probably starting to break out in a cold sweat as you read this! I think everyone will agree that this must change.
The formula for quality data management
So, what is the formula for creating good quality data management practices within an asset management firm, and in support of the marketing operations?
There are really four principle factors:
- Ownership – someone has to own every reportable data point and understand its origins, purpose, and validity. And if the source, measure, form, format or meaning of the data changes they need to actively manage the introduction of the change to the firm’s data management, compliance and communications teams.
- Controls – a system has to be put in place that creates auditability of the data. Where did it come from, how was it then calculated, transformed or combined. How do you know that process was both accurate and consistent?
- Availability – one of the recurring theme’s we’ve encountered is that access to the data in any officially sanctioned storage location is often a secret. People don’t know where it is, how to get it, or what services the data group can provide…or when. They solve this problem by going back to some spreadsheets that circumvent the process.
- Active Warehouse Maintenance – This is a very common trap. IT groups building corporate data resources like to fire and forget. They build the database and the import/export utilities but then the team “rolls off” onto other projects as if the state of the data sources and uses is invariable…which of course it isn’t.
I’m continually shocked during the data on-boarding stage of our projects how frequently some version of the following happens:
There’s a data miss-match of some sort and, after re-checking our loaders and datamaps and digging around the source data files, the client finally says; “Oh, you’ll need this additional lookup table.” Or, “oh, you can’t use that data from the data mart, here’s a spreadsheet that’s more up to date.”
This is an illustration of a failure in all four of the principal factors listed above. There was clearly no ownership of the data point, maintenance of the data mart, or controls to make sure that accurate information is made available in a clean form to the ultimate user of the data.
It’s not the crime, it’s the cover up
Forgive this overly dramatic heading. I’m not trying to fear-monger. What I am trying to do is very clearly point out that, as a publisher of investment data, your liability for publishing incorrect data is minor compared to your liability for having irresponsible data management practices.
Fortunately there are vendors, tools, and educational resources that companies can avail themselves of to jump ahead with a strategy and facilities supporting data quality management. If you’re interested in changing the status quo at your firm, I’d be happy to give you a free and honest assessment of your situation. Just drop me a line or connect with me on LinkedIn. I’m always happy to help. – John
Did you like this post? Please share! You can also subscribe to our blog.Here are some related resources that might interest you:
From the Blog:The 3 Ways to Approach Content Automation
From the Blog:Why Data Management Should Drive your Document Automation Vendor Decision
From the Blog:What is Data Quality and why should marketers care?