I’ve spoken with numerous credit unions about their interest in leveraging hardware to have automated tellers, and they’ve all had questions about how that might work alongside their core processing system. We’ll get to that in a minute, but first let’s talk about the hardware solution models as the springboard to that discussion.

First, there are kiosk solutions. These are ATMs on steroids, so to speak. They’re ATMs, but they allow members to perform transactions through core integration so they can do much more. Members can pay loans, cash checks, take an advance on a line of credit, and transfer money, all without assistance from a teller. Self-service is the name of the game here. These transactions are accomplished via direct core integration versus riding ATM network rails.

Then you have interactive teller machines (ITMs). These are ATMs where members can push a button and chat with a live remote teller if the member finds themselves requiring functionality beyond the capability of an ATM. Essentially, you’re putting a teller at your ATM. Some ITMs can also leverage the previously mentioned kiosk integration to allow for a more robust transaction set when tellers aren’t staffing the video teller capability.

Different technology, different strategy

Now, let’s look at the strategy behind these two models. First, with kiosks, credit unions are looking to let members self-serve to lessen resource requirements, to serve members with routine tasks. This is a reflection of an evolving teller role. Some credit unions don’t necessarily want to reduce teller head count, but repurpose it for cross sales to members or concierge-type services.

The ITM group, on the other hand, is typically wanting to remotely put tellers in underserved areas, drive-thrus, or even lobbies, all with the idea in mind to maximize teller resources while retaining a human presence and interaction with the member. Furthermore, this model works to remove the limitations on the transaction set supported by kiosks through human intervention directly with the core system.

However, some ITM users do not want dual entry between the ITM and the core processing system, which creates the risk of errors. Remember, with non-integrated ITMs a teller must key the transaction into the ITM and then turn to the core system and key it in there. It works, but is not ideal and could lead to errors. Both models, kiosk and ITM alike, want to enjoy an expanded transaction set that only integration with the core can provide.

What to know before you start

Regardless of which strategy a credit union adopts, most (though not all), seem to share one trait in common: they misjudge the effort and the participation that will be required from them to make their vision a reality. Hence, I’ve created a list below of things you should be ready for and realistic about before you start writing checks. Consider it a helpful guide to get you thinking. If taken in earnest, these points will help ensure a successful project and avoid disappointment and frustration down the line. Consider these very early on, and keep them in mind as you move down the line and get closer to realizing your vision.

1. Know your strategy

What is it you’re trying to accomplish with this technology? Is your goal to reduce teller head count, provide new services, or perhaps create new business models? Are you trying to serve markets you otherwise couldn’t because you can’t build physical branches there? What’s your end game, and does it make sense both strategically and, of course, financially? Define it, write it down, and refer to it as you make decisions. This is your North Star.

2. Define what types of integration you may or may not need between your solution and your core processing system

Don’t cop out here and just say, “fully integrated.” That’s not helpful to you or the programmers you’re asking to develop the integration. Be specific; you don’t want to overbuy for features you won’t use. What types of transactions do you truly need to support? How will check holds, limits, etc., play into the mix? Think it through. Think about staffed hours vs. off hours; how does the transaction set change? This is important to have clearly defined as it will not only affect your costs and timing, but also your vision and strategy.

3. Be realistic on timing

I had one credit union become very frustrated because they believed that since we were already working on an integration for one credit union, they could expand upon it and have it in 3-6 months. However, the reality was that we were still 9 months away from beta for the first credit union, and that wasn’t even considering the expanded features the second credit union wanted. These integration projects can easily take upwards of a year if they are a first go at an integration. Don’t make assumptions. Get commitments before you set expectations.

4. Understand your role

I had one credit union say, “talk to my vendor and I’ll jump in later.” This is not how to start a project off right. Understand this is your project. Your vendors and partners wish to help you, but it’s just that, helping you. You core processor and ITM vendor don’t have a formal relationship, you are the bond between them. They are both working for you on your vision.

You need to get them to the table and get them talking. Have your vision and expectations for deliverables and timing on the table and agreed to. Be ready to make the calls on modifications should one vendor be unable to accommodate your vision. If you don’t hold them accountable and manage the project, you will be the one frustrated. This is not a hands-off endeavor.

5. Costs

Understand what the costs for integration are from each party. Make sure you’re clear on one-time costs and ongoing costs. Integration usually has costs from both the hardware manufacturer, their reseller, and your core processor. Costs are typically upfront and ongoing for support, but can also have special costs if recertification or additional support are required down the line as environments change, etc. You buy it again and again, not just once. Be prepared for that.

6. Be ready for an NDA battle up front

I hate NDAs, but they seem to be required before anyone will start collaborating. Getting lawyers past these can take months and can really hinder operations. Let me give an example. Right now, I have an NDA with a hardware manufacturer, but they did not execute one with the credit union for which we’re doing the project. So the manufacturer and I can talk all day, but I just can’t say anything protected to the credit union who’s paying both of us to execute on their vision. It’s inconvenient for everyone, and it doesn’t make much sense. Crazy, right? Welcome to legalities!

7. Think about your rollout strategy

How will you test the solution once delivered? Plan for a slow rollout. Don’t set yourself up for a specific day. If you just flip a switch on a production system, you may find yourself very frustrated, almost as frustrated as your members trying to use the equipment. Rather, think about testing it off hours with staff then moving it to a controlled environment where staff are close at hand to assist members as needed. Have the experts on standby. Finally, push to your members in production.

8. Don’t over engineer

I have seen credit unions invest in elaborate hardware and software solutions in a strategy I’ll call, “technology for the sake of technology.” While at the same time, others have just gotten started with simple, effective (though not always efficient) solutions. I have seen the simpler groups get to market and benefiting while the more elaborate solution groups linger, trying to get it all right. If you can get a simple solution off the ground and working wonderfully, then if you do invest to go to the next level, you have a great justification for it. Prove it to improve it, so to speak.

9. Do your members want it?

Are you considering your membership in this move? Do you want it for them or do they want it? Some members love technology and the convenience, others not so much. Think about if they will like it or be put off by it and, most importantly, don’t just consider the members of today, but also the members of tomorrow. What type of members are you trying to attract? Consider them.

10. Keep your eye on the prize

After reading the above, you may be thinking, “why bother?” So much work, so many pitfalls. What if you get it all wrong after spending lots of money? Look, this is all doable and the rewards can be there, but you must be realistic about it all and have the proper expectations.

Go into it with the right expectations and you’ll come out of it satisfied. Don’t get discouraged, leverage the experts, keep pushing, and keep reminding yourself of why you’re doing the project and the benefits it will bring.