Here’s a question many finance teams are wrestling with: when does your tech stack actually need an upgrade, and when are you just chasing shiny objects? Hear from Lonnie Bloom, CPA (Withum), Walter Merkas (Withum), and Jason Berwanger (HubiFi) on when teams should customize their ERP and when it’s time to upgrade it entire. 

‍

The Real Problem Behind Slow Closes

‍

It’s a constant refrain:  "Our month-end close is taking forever. Should we buy a new ERP?" But that's asking the wrong question. A slow close is a symptom, not a diagnosis. The actual disease varies wildly depending on your business model.

‍

For B2B companies with complex contracts but fewer customers, an ERP can genuinely solve problems. You're managing orders, entitlements, billing, and cash application in one place—that makes sense.

‍

For B2C or high-volume businesses, the story changes completely. Your ERP becomes just another hub where accounting manually consolidates data from CRMs, billing systems, payment processors like Stripe, and order management platforms. Adding NetSuite in this scenario doesn't fix the problem—it just gives you a fancier general ledger while your team still drowns in reconciliations.

‍

As Jason put it: "We got a better general ledger, but we didn't actually solve the problem."

‍

Where Month-End Actually Breaks Down

‍

A poll of 50+ finance and accounting leaders showed that what slows them down most is clustered around a few areas:

‍

• Revenue recognition and AR (especially for high-volume businesses dealing with ASC 606 complexity)
• Poor integration between billing platforms (fragmented data across multiple systems)
• Multi-entity consolidations (particularly with foreign subsidiaries on different accounting systems)
• Manual journal entries and adjustments (the human-in-the-loop tax)

‍

The pattern? Data fragmentation. One horror story involved a company with QuickBooks in the US, Xero in Mexico, and a different system in India—all reconciled manually in Excel by one controller. When that person left, institutional knowledge walked out the door.

‍

Build vs. Buy: Why Customization Usually Loses

‍

When does it make sense to customize? Wally's advice is blunt: "Most implementers need to develop a backbone and say no."

‍

The math rarely works in favor of customizing your ERP:

‍

• Custom scripts cost around $45,000 just to reverse-engineer when the engineer who built them leaves
• Maintaining customizations creates ongoing technical debt
• You're essentially becoming a software company for your back office

‍

A better approach? Look for third-party solutions that actually solve the end-to-end problem. Not just another SaaS tool that becomes "another source of truth that isn't a source of truth”, but technology that takes a meaningful chunk of your workflow and genuinely automates it.

The AI Reality Check

‍

Everyone's pitching "AI-native" solutions right now. But here's what accountants need to understand: AI is probabilistic. You might get different answers to the same question. That's fine for some use cases—terrible for financial reporting where you're personally liable for the output.

‍

The hierarchy that actually works:

‍

1. Clean, consolidated data (solve fragmentation first)
2. Deterministic automation (end-to-end workflows with audit trails and reconciliations)
3. AI on top (for analysis, anomaly detection, research)
4. ‍

Trying to slap AI onto messy data is like using a sophisticated search algorithm to find a truck when the inventory system doesn't track bed length. The problem isn't the AI—it's the inputs.

‍

What You Can Do Today

‍

Before evaluating any new software:

‍

1. Map your data flows - Where do customer transactions actually happen? Where's your revenue generated, invoiced, disputed, collected?
2. Document your business model and accounting policies - The intersection of these two reveals where you actually have problems
3. Start small - Walk through 10 transactions from contract to cash with your team
4. Question transaction volume pricing - Ask vendors when your rates spike. Many companies get burned here.

‍

The Bottom Line

‍

A fast close isn't necessarily better than a thorough one. As Wally noted: "The goal is a close you can explain, defend, and trust."

‍

Don't let FOMO around AI or ERP upgrades drive decisions. Instead, identify the actual constraint in your process. Sometimes that's technology. Sometimes it's people. Sometimes it's process documentation that walked out the door with your last controller.

‍

Constraint breeds innovation—but only if you're solving the right problem.

‍

Watch the full webinar recording here or contact us to discuss how HubiFi handles high-volume revenue recognition for SaaS and subscription businesses.

‍