How GoldFinch Earns Trust in ERP Implementation: Facts, Feelings, Needs, Next Steps

More than a go-live

An ERP decision is a bet-the-business decision. When a manufacturer, distributor, or food bank chooses GoldFinch, they hand us their inventory, their orders, their books, and the daily routines of everyone who touches them. If we get it wrong, trucks don’t ship, and invoices don’t go out.

So our job doesn’t end at go-live. It continues through the first month-end close, the first physical count, and the first new hire who has to learn the system cold. Over that stretch, we wear a lot of hats. Some days we are solution architects. Some days we are trainers. And some days we are closer to counselors: helping a controller feel ready for their first close, or showing an operations lead how the new system will make their team stronger.

That last role never appears in a statement of work, and it’s often the one that decides whether a customer trusts us. This post is about how we try to do it well.

A framework we borrowed

Psychologist Marshall Rosenberg developed a method called Nonviolent Communication (NVC) for conversations where the stakes and emotions run high. Despite the name, it’s less about avoiding conflict than about being precise when things are tense. It was never meant for software, but it maps closely onto an ERP implementation. The four steps are:

  1. Observation: state what happened, without judgment.
  2. Feeling: name the emotion it creates.
  3. Need: identify the underlying need behind that feeling.
  4. Request: ask for a specific, doable action, one the other person is free to decline.

Rosenberg’s insight is that conflict escalates when people trade judgments and demands, and eases when they surface the needs underneath. In an ERP project, that difference shows up every week.

Step 1: State the facts

ERP projects generate a lot of judgment. “Your data is a mess.” “The team isn’t engaged.” “This vendor overpromised.” Every one of those statements might feel true, and every one of them puts the other person on the defensive.

Replace Judgments with Observations

Instead of a judgment
Try a verifiable fact

“Your item master is a disaster.”

“1,240 of your 6,800 items have no unit of measure, and 312 appear to be duplicates.”

“Your team keeps missing deadlines.”

“Three of the five test cases due Friday haven’t been started.”

“We’re behind.”

“We’re two weeks past the original due date, and here is what moved.”

Facts lower the temperature. They also hold us accountable: if the slip is on our side, the same rule applies, and we say so plainly. Customers can tell the difference between a partner who reports facts and one who manages impressions.

Step 2: Name the feelings

An ERP implementation is an emotional event. The people on the customer side are doing their day jobs and the project at the same time. Many of them have lived through a failed system change before. Some quietly worry that the new system will expose mistakes, or replace them.

That emotion rarely shows up labeled. It shows up as “why does this take six clicks when the old system took two?” or a sharp email at 9 p.m. Our instinct as consultants is to answer the literal question. The better move is to listen for what’s behind it first:

“It sounds like you’re worried month-end is going to land on your desk with a lot of open issues. Is that right?”

Naming the feeling doesn’t solve anything by itself. But it tells the customer we see them as people, not just as a project. We also try to be honest about our own side: “We’re concerned about the go-live date too” is a more trustworthy sentence than false confidence.

Step 3: Surface the needs, on both sides

Behind almost every frustration in an ERP project is a legitimate need. When we find it, the conversation shifts from who’s right to what both of us require to succeed.

What we hearWhat’s often underneathThe need
“The old system was faster.” Fear of falling behind during the transition Confidence and competence
“Why wasn’t I consulted?” Feeling sidelined in decisions about their own work Inclusion and respect
“Just tell us it will work.” Anxiety about the board, auditors, or owners Certainty and safety
“We can’t freeze operations for testing.” Pressure to keep shipping Continuity

This works in the other direction too. A partner who never states its own needs isn’t being humble; it’s setting up a surprise. We need clean data on time, a decision-maker who can say yes or no, and testers who are given hours in their week to test.

Saying that early, without blame, is part of earning trust. Our first phase, Establish the Operational Foundation, typically runs three to six months, and we write these mutual needs into the project plan before the build starts.

Step 4: Agree on the next step

Vague requests produce vague results. “We need more engagement from your team” gives no one anything to do on Monday morning. A good request names who, what, and when, and is small enough to actually happen.

Vague Requests Get Vague Results

“Please clean up your data.”

“Could Maria review the 312 possible duplicate SKUs by Thursday and mark keep or merge in the shared sheet?”

“We need you to commit to testing.”

“Can we block two hours on Tuesday and Wednesday for your warehouse lead to run the receiving scripts with us?”

“Trust us, it’ll be fine.”

“We’ll give you a comprehensive testing plan, and your team signs off on the results before go-live.”

A good request names Who What When and leaves room for no.

Rosenberg drew a sharp line between a request and a demand: a request leaves room for no. When a customer says “we can’t do Thursday,” that’s not a failure. It’s information, and the next question is “what would work?” Negotiating the request together is how both sides end up owning the plan.

Try this before your next hard meeting

Before any difficult customer meeting, our project lead writes four lines:

  • The facts: what happened, in terms anyone could verify.
  • The likely feelings: on their side and ours.
  • The needs: what each side requires to succeed.
  • The one request: who, what, and by when.

It takes five minutes, and it often changes the conversation.

What this looks like at a food bank

Food bank implementations put all four steps to the test. A typical food bank syncs inventory from its warehouses into the portal that hundreds of partner pantries order from, often across more than 10,000 SKUs. Pantry managers need to trust that portal completely, because families are counting on what it shows.

From “Messy Data” to a Short List

10of 10,000 SKUs

didn’t sync from the warehouse to the partner pantry portal. Quantifying it turned a vague worry into a list we could investigate.

1The Facts

10 records failed, often for a specific reason like a validation rule or a misfired trigger.

2The Feelings

“You’re worried a pantry will order food that isn’t on the shelf, and a family will leave without it.”

3The Needs

The pantry needs confidence the portal matches the warehouse. We need someone to confirm what’s actually on hand.

4The Request

“Could your inventory lead confirm on-hand counts for these 10 items by 2 p.m., before tomorrow’s ordering window?”

After go-live, the relationship begins

Go-live is where most consultants hand off. For our customers, it’s where the real test starts: the first close, the first audit, the first peak season. That’s why ongoing support at GoldFinch is a relationship, not just a ticket queue.

The same four steps shape how we handle support. A support case starts with facts: what happened, which record, what was expected. We acknowledge the pressure behind an urgent ticket rather than just its priority code. We ask what the customer actually needs, which is sometimes a fix and sometimes a five-minute walkthrough so they never hit the issue again. And we close with a clear next step and a date, not “we’ll look into it.”

Over months and years, that consistency is what trust is made of. Customers stop wondering whether we’ll show up, because we keep showing up.

The same language builds our own team

A consulting team that can’t talk honestly with itself won’t talk honestly with customers. The pressure customers feel lands on our consultants too: late nights before cutover, scope that keeps growing, a customer who is upset with someone personally.

We use the same four steps internally:

  • In tough feedback: “In Tuesday’s call, the customer asked about lot tracking twice and we didn’t answer. I felt uneasy because I need them to trust our expertise. Could you send them a written answer today?”
  • In workload: a consultant can say “I’m stretched across three go-lives this month, and I need focus time. Can we move the new demo to next week?” without it reading as a complaint.

When a team practices observations over judgments and requests over demands, problems surface earlier and people stay longer. That stability is something customers feel, even if they never see it.

Trust is earned one conversation at a time

A great software demo gets a customer to sign. Trust is what keeps them with us through the hard weeks and the years after. And trust isn’t built by a methodology slide. It’s built in hundreds of small conversations where we told the truth, saw the person in front of us, named what both sides needed, and asked for something they could actually do.

At GoldFinch, that’s what we mean when we say we partner with our customers.

Weighing an ERP decision? Talk with one of our implementation leads about what your first six months would look like.

Scroll to Top