It's three weeks after go-live. A member walks up to the counter to add a beneficiary to her account. Your teller has done this a dozen times, just not on this system. The screens are different, the steps moved, and the guide on the shared drive still shows the old core. So she taps someone on the shoulder. That someone is already helping two other people.
I hear some version of this on almost every call with a bank or credit union heading into a conversion. For months after go-live, your help desk and your subject matter experts get buried. Everyone is learning a new system at the same time, and the few people who know the answers can't get their own work done.
Here's what I've learned from those conversations: the best time to fix your procedures isn't after the conversion. It's before.
A core conversion changes nearly every procedure at once, and most institutions rebuild their documentation after go-live instead of before it. Staff who knew how to do something suddenly don't, the old guides point to screens that no longer exist, and every question lands on the same handful of experts.
The stories people share with me are rough. One credit union told us they were still finding procedures to update 18 months after their conversion. Their team was drowning, and staff would rather yell across the lobby than search the intranet. A contact center manager described queues of 45 members waiting up to 30 minutes after their core transition. He figured about half of those calls could have been handled with better guidance. And a banker who'd been through five conversions put it more bluntly: "Bankers will say a core can kill you."
Our CEO, Greg DeVore, points out the part that catches people off guard. A conversion doesn't just challenge your new hires. It turns people who knew a procedure cold into people who are unsure, or wrong. Your 20-year teller is a new employee again.
Most conversion plans lean on classroom training: a few days of sessions, a binder of job aids, and a lot of hope. The problem is that you can only pour so much into someone's head. I think of it as mental RAM. You can memorize only so many processes before things start falling out.
Timing makes it worse. If you convert in March, your team won't remember much of that training by the time year-end processing comes around. Our implementation team hears the same thing after almost every conversion: people went through the training, but they "can't remember how to do half the stuff they told us."
And the hardest procedures are the ones nobody does often enough to build muscle memory. Deceased member accounts. Wires. IRAs. Year-end. Those are exactly the moments when a member is waiting and your teller is guessing, and those teller mistakes add up fast.
Greg's advice here is simple: plan for people to forget what you taught them. Use training for what they need to understand, like what's changing, why it matters and how the new system fits together. Then give them guides they can follow in the moment.
Before go-live, alongside the conversion work itself. You're going to rebuild years of training material and job aids anyway. If you wait until after the conversion, you're rebuilding them while your help desk is on fire. The credit unions that get the most out of it are the ones that did it in tandem.
Doing it in tandem doesn't mean waiting for perfect system access, either. Build what you can now. Leave placeholders where screenshots will go, then finish them once your test environment is ready.
Farm Credit of the Virginias went through this when they moved to Fiserv DNA for loan servicing. The change touched 10 departments and nearly 200 employees. Their procedures lived in multi-page PDFs and Word files on SharePoint. People leaned on memory, their managers and binders they'd made themselves.
So they didn't wait for go-live. Department managers picked the daily and high-impact tasks people would need to do in DNA, and those became the first procedures. The team brought in Fiserv's lengthy documentation and turned it into step-by-step guides. And they launched their procedure hub right alongside DNA, not months later.
They kept going after go-live, too. When people hit a snag, the team published workaround procedures so employees and members weren't stuck waiting. As their Chief Risk Officer, Pete Cypret, put it: "We needed to build systems to support this change and minimize the impact on our employees." They're now using the same approach for their next system rollouts.
Read the full Farm Credit of the Virginias story
When this works, the results show up fast. A credit union in Hawaii rebuilt its training around finding and following guides and took time to proficiency from 12 months to one week. Greg likes to measure three things: how long it takes someone to work independently, how often mistakes happen, and how many questions get escalated. All three should drop after a conversion, not spike.
As soon as your conversion date is set and you have access to a test environment. Starting early lets your experts capture procedures during testing instead of scrambling after go-live.
Not necessarily. But you do need one source of truth that staff can search with a plain question and trust is current. If guides for the old core sit next to guides for the new one, nobody knows which to follow, and they'll tap someone on the shoulder instead.
Give every procedure an owner and a date to review it. Conversions don't end at go-live. Settings change, workarounds get fixed, and someone needs to keep each guide accurate.
This is a great time to talk. We work with over 80 financial institutions, including more than 50 credit unions, and conversions come up constantly. My suggestion: send us your hairiest procedure and let's see if it holds up.
Not ready to talk yet? Start with our SOP playbook. It walks through organizing what you have, turning it into usable guides, and getting them in front of staff.