Original Article | GreenProfit
The question of “should we implement a data management system?” is already answered. Yes. Yes. And yes. So what’s holding up progress? Oh right, all the unknown costs. What does the system cost? Are there hidden charges? What about ongoing expenses?
We’re going to answer all of these cost-related questions about data management systems in this article.
Whether you’re a large $5B+ credit union, a solidly mid-size $400M one, or part of the majority of <$100M institutions, money is always a factor. Namely, how much is available for what purpose. You understand that spending in one area might mean having less for another.
In this case, budgeting really can be a zero-sum game. How can we both maximize the impact of every dollar spent and open opportunities to bring more revenue into the institution?
Yes, data management offers benefits that make it a cost worth tackling.
Implemented correctly, a data management system can become the “best dollars you ever spent”. More so, it can help your institution celebrate all the dollars you never spent. Because inefficiency costs, too. It’s a core aspect of digital transformation efforts.
Here’s what this article will cover:
- Specific and conceptual costs of implementing a data management system at your financial institution
- Differences between cloud and on-premises setups
- Ballpark pricing figures for what you can expect to spend on different types of systems
- How costs can vary based on size as well as factors you can control
I’m ready to chew on some values. Are you?
“Levels” of Data Management System Providers
From our research, there seems to be three “styles” of data management system provider complexity. None are “better” or “worse”, just different approaches. They were explained to us in a car analogy (yeah, we’re car people):
Hold Your Hand the Entire Way: Order a Car Online, Have it Delivered
Car buying doesn’t need to be hard. You can just order one online, complete the financing and other legalese digitally, then have it delivered to your driveway. It’s not that you do nothing, it’s that you get more from your time and energy.
This level helps with implementation and offers managed service in-house. You have a direct data concierge and team to help ensure it all goes smoothly. Then, they are available to keep everything running as you expect. It also tends to be the most expensive.
There to Help: Build Your Kit Car
Imagine getting all the pieces for your car delivered. Everything is there, and you receive a full set of instructions. It’s just up to you to do the build. For some people, that’s perfect. And you’ll definitely save a ton of money on your completed vehicle, ahem, data management system.
The provider will help you build a data model and outsource implementation. Hosting will be on your own servers (cloud, hybrid, or on-premises).
Custom Development: Make Your Own Car
Are the cars available to purchase just not the perfect fit for your institution and members? Do you want to work directly with the experts to build a custom vehicle that’s just what you had in mind? The professional services and development approach of this level is for you.
Example providers that offer services across these 3 ranges: AvantEdge Digital, Trellance, and Arkatechture.
Stuff Your Mind, Not Your Inbox. Get the info you need a couple times a month.
Subscribe Now!
What does a Data Management System Cost? (With caveats)
Let’s first make it clear we cannot tell you how much your institution will pay. However, we can give you ranges based on the level of complexity and service, as well as your asset size and data needs.
In other words, it helps to know whether you’re looking at budgeting a thousand, ten thousand, or a million dollars, right?
We’ll align some of the price ranges with the kinds of services and degree of automation chosen. That way, you can use these numbers to help decide what approach makes the most sense to pursue.
Note that these price ranges do not include any staffing costs which may be necessary to set up and maintain such a system.
Let’s ballpark away!
(For all of these ranges, we are assuming you’ve chosen to use cloud or hybrid setups.)
Low End Pricing
This section assumes one of the following:
- You’ve chosen to work with a data management system provider, but really want the most minimal of services. You will do most of the data organizing, cleaning, linking, and rule-setting in-house. You’ll get limited dashboards and other reports.
- You are a smaller institution with lower data management needs, both up-front and ongoing.
To get started, your institution is looking at around $100,000 for implementation and monthly costs around $5,000. Thus, your first year cost would be about $160,000, with ongoing expenses of about $60,000 annually.
High End Pricing
This section assumes one of the following:
- You want a data management system partner in every sense. From assistance in preparing your data to governance advice to a swath of customized dashboards for your departments, this is the premium experience.
- You are a high asset-sized financial institution with complex data needs. There are a lot of disparate systems to connect, existing data silos, and a large staff to train in proper usage.
Implementation can be around $250,000 with monthly costs of around $25,000. In our conversations to acquire this information, we were advised that most credit unions budget based on their first 6 months after implementation.
With those figures, a large institution with high-end data needs would expect to pay $400,000 for the first 6 months ($250,000 + ($25,000 * 6)). In year two, and onward, costs would be around $300,000 annually.
We get it. That’s not cheap. And the costs of not improving your data process are less “expense” and more “lost opportunity”. So it’s tough to directly compare, especially when implementation could take some time.
Since we brought it up, what’s involved in that implementation cost?
What’s in a One-Time Setup Cost?
In the previous sections, you saw how the upfront cost can be massively higher than the ongoing expenses. Why? Is it just to ensure the time investment of the data management vendor isn’t wasted? Partly, but most of it goes towards the real work involved.
When you recruit a data management system provider, you’re asking for them to design and build a platform suited for your institution. It will have to accommodate all your data, from a range of sources.
The first part of that implementation cost is the professional service time required by their team of people. We’ll discuss the timeframe below, but for now, assume it can take many hours. Their expertise doesn’t come free.
The second part of the up-front cost is to cover the integration of your data sources. For example, you’ll want your Loan Origination System, core, and online/mobile banking sharing data seamlessly. Whether in technical effort or purchase of APIs, that’s not cheap.
Sidenote: One of our partners connects their GAP platform to nearly all LOS providers. To get that integration, they have to pay to “get connected”, even if the technical challenge is low. It can be in the tens of thousands of dollars. You are in a similar situation. Don’t like the costs? Ask your providers to reduce the rates to “connect” to their technology.
A typical institution has a “core + 3” data requirement. Those extras are your LOS, online banking, and credit card processor. If you have more, the one-time implementation cost might be higher. It could also take longer to set it all up. Which brings us to…
How Long to Get Set Up?
Unfortunately, there’s no “set it and forget it” or “turnkey” data management system for your credit union or bank. If you are told by a vendor that their platform is, ask for references. Then, find out how long each took for implementation and how well it’s still working.
We heard a story about one credit union which spent 2 years with a provider before giving up. Once they brought in another vendor, their system was ready for use in six months.
The key is to ensure your institution and chosen provider are on the same page from the start. Communicate your expectations clearly, and request the same from them. Challenges will emerge, but if they aren’t transparent about timeframes or costs, maybe it’s not the right fit.
Options in System Setup and Hosting
All of the previous content assumed a cloud or hybrid setup. You may be wondering, “what’s that even mean?” Great question. To keep us all on the same page, let’s define:
Cloud-Based Setup
The servers holding your data, doing any number-crunching, and the systems managed by your provider are all “somewhere else”. Most often, you’ll be hosted on Microsoft Azure Cloud, Google Cloud, or Amazon’s AWS platform.
Using this approach means you have no costs nor concerns of maintaining physical infrastructure. And it’s highly unlikely to have issues from a hardware failure, since your data is distributed across servers (you may be able to choose server locations).
Hybrid Setup
This takes what you loved most about the on-premises system, then removes all the downsides. You know, like the, “we need a place to securely put our servers” or “it sure costs a lot to keep these systems cool, powered, and physically maintained”…downsides.
Instead of “servers in your basement”, you’d be connected to a local data center where you could go visit your data to say hi. I mean, if that’s your thing. Essentially, you’re preserving physical access to the hardware while reaping the benefits of a cloud operation.
What’s in the Monthly Costs?
Now that we have a basic understanding of the costs for a data management system, let’s learn where that money goes. Besides obviously paying for the experts at the provider, there are costs they incur for using software and services not developed in-house.
This is a good thing. The more “standard” software used, the more likely issues will be addressed quickly and updates will continue for features, stability, and security. We’ve struggled with custom builds in the past, and it’s just not worth the effort most of the time.
Database Licensing
As your data management system provider helps facilitate better access to your data, you need a way to…manage it. Likely, they will be licensing Redshift (Amazon), Snowflake, or another database platform.
Data Visualization (Business Intelligence) Licensing
Beautifully organized and managed data is great. Accessing it to generate the insights you need is another. Here’s where data visualization comes into play. You may also know it as business intelligence. There are licensing costs here as well.
It’s a minor part, but more licenses (access for staff) adds to that monthly cost.
Common systems include:
Dashboards & Analytical Models
How many dashboards and models are included in your system? For example, with Arkalytics, their baseline plan includes 20 dashboards. If you need more specialized visualization, the next tier up offers another 10.
Discuss with your provider early on what dashboards will be necessary to achieve your goals across all departments. You can always add more later, but getting everything set up with what you want is ideal.
Support
You’re going to need some help along the way. Agree upon the number of hours of support included at no extra charge. Consider if the provider has an hourly rate for additional assistance.
Data Input
How much data do you process? This is really a “how big is your institution” question. More account holders means more loans, transactions, and overall…more data. Additional data storage and transfer may impact costs (discuss volume ahead of time to avoid surprises).
Are there costs beyond what the provider charges?
This is a question best suited to your team and provider. However, remember that your data journey is something you take together. Which means your staff must be trained in its best operation. In reality, the biggest extra cost for your institution is time.
Data Lineage and Governance
Across all departments, build a team of subject matter experts. For example, Judy in Lending is responsible for addressing data quality defects originating from the Loan Origination System (and has a written process to train a potential replacement if she moves up).
Have leaders throughout your institution with similar responsibilities to ensure the data going in is always clean and formatted as the system expects.
What would a data management system cost to do ourselves?
Given all these insights, you may be wondering, “why not just do it ourselves?” You’re not the first. Does it make financial sense? Let’s look at what the self-service approach might involve.
Your largest cost is going to be the people. In discussing with data management partners, they suggested the following cost structure for a typical credit union:
- 5 data analytics experts @ $70-$100K each ($350K-$500K per year)
- This would be a “bare-bones” team that (hopefully) gets you the equivalent of our “low-end pricing” from above. And…
- Licensing costs for chosen database (most likely Microsoft SQL Server), preferred Business Intelligence tool, and servers (if not done hybrid or cloud-based)
- These would be more expensive, since a data management provider can negotiate lower pricing through economies of scale.
Already, we can see that going it alone is almost definitely far more expensive, just to come close to their promised service level. And that’s without discussing licensing or even buying physical servers (for on-premises).
You will also need to maintain a secure space for your servers (both from physical and digital attacks) with power backup, network redundancy, and fire suppression. Ideally, you’d opt for colocation (another space) to provide a mirror for data backup and uptime security.
Additionally, you’ll eventually need to replace the hardware, while regularly maintaining the batteries/generators, and fancy safety/security systems of your server room.
In far fewer words, this section is the why for having a cloud computing industry at all.
The Most Important Cost of a Data Management System
It may sound cliché, but the greatest cost of a data management system is not having one. Can you glean instant insights from your borrower activity? Are your marketing efforts based on real-time data of aggregated debit card usage?
In other words, are you making decisions on what seems right or what your data says is happening?
Get More Insights on Data Management…and More
With an entire category devoted to data management, the Learning Library is your first stop on the journey. Looking for the why of good data practices? Or perhaps a few of the providers in the industry?
We’ll help you learn the Pros and Cons, then tackle four steps to setting up a data management system. Of course, here, you got a better understanding of the costs involved, both up front and ongoing.
The Learning Library offers more than just data on data. Across over a dozen categories, you are inside a due diligence resource like no other for our industry.
Subscribe to get regular (though not too regular) insights in your email. And bookmark this page so you can always see the latest article.
Blogger. Speaker. Part-time Jedi.
Focused on helping your bank or credit union grow in the face of emerging challenges.