Legacy Software Costs More Than the License

Ask a plant manager or a claims director what is slowing them down and you will hear about the ERP, or the policy admin platform, or whatever the oldest system in the building is. 

It has a name, a vendor, and a replacement cost somebody can put in a deck, which makes it the easiest thing in the room to point at.

It is also still doing its job, taking the order or adjudicating the claim the way it always has. 

The cost sits in everything your team does between those systems: the export that becomes the production schedule, the record re-typed because the first system cannot send it, the status call a portal should have answered. 

That work is the real price of legacy software, and replacing the platform is the most disruptive way to go after it. The cheaper fix is usually a layer built around the system you keep.

‍

Why do companies blame the legacy system first?

Companies blame the legacy system first because you can get a quote to replace it, and nobody will quote you for the twelve manual steps around it.

It is also still load-bearing. A manufacturer's ERP holds order history, production scheduling, inventory movement, quality records, and supplier data. 

A health plan's core platform holds claims history, member records, policy data, provider network information, and the appeals and audit trail that proves what happened and when. 

A system can be twenty years old and still be the most reliable thing a company owns, because it has absorbed two decades of business rules that exist nowhere else in writing.

‍

What does replacing a legacy system actually cost?

Replacing a legacy system puts daily operations at risk for as long as the cutover takes, and the cutover takes longer than the plan says.

Zimmer Biomet went live on SAP S/4HANA on July 4, 2024, and on September 4, 2025 it sued Deloitte for $172 million, alleging the implementation left the company unable to send products out the door, receive inventory, issue invoices, or generate sales reports. The complaint puts implementation fees at $94 million, which is 36% above what was quoted, with 51 change orders accounting for $23 million of that, and claims a further $72 million the manufacturer spent cleaning up after go-live.

The scope had already been cut back from North America and Latin America down to North America alone. 

This is a medical device manufacturer with a real budget and a top-tier integrator, running a narrowed project, and a year later the thing it was buying was still being argued about in court. Whoever signs a project like that owns the quarter it goes sideways.

There are real reasons to replace. 

The vendor is ending support. 

The data model cannot represent how the business works now. A compliance requirement cannot be met inside the current system. 

A company whose actual complaint is that the work takes too many steps does not have one of those. Replacement is sometimes the right call, and it is never the low-risk one.

‍

Where does operational drag come from in manufacturing and insurance?

Operational drag lives in the handoffs between systems, where a person has to carry the work across because nothing else will.

Integration complexity was the most common explanation for why technology investments had not delivered, named by 47% of the 610 operations and supply chain executives surveyed last spring, with data issues close behind at 44%. 

Both answers describe the same building, where every system works and no two of them agree on the same record.

On a plant floor that means sales enters an order production planning cannot act on until somebody builds the schedule in a spreadsheet, and quality flags a defect that customer service hears about from the customer first. 

The same gap runs through a health plan, where a prior authorization status moves without the provider portal showing a next step, and a provider's address changes in one database while the member-facing directory keeps the old one.

Prior authorization is the clearest version of it, because the industry has already priced the gap. Physicians handle 40 prior authorizations a week and spend 13 hours of physician and staff time on them, and 40% of practices now employ someone whose entire job is prior authorization. 

That job exists because two organizations cannot complete a transaction without a person standing between their systems.

‍

Why do spreadsheets end up between your systems?

A spreadsheet appears wherever two systems have to agree and nothing was built to make them agree.

The usual places:

• Production exports demand data because the plant schedule cannot be built inside the ERP
• Quality issues get tracked outside the system that recorded the defect
• Maintenance requests live in a shared sheet because the asset record has no queue attached to it
• Claims exceptions sit in Excel because the core platform has no status for them
• Prior authorization follow-up lists live outside the platform doing the authorizing
• Provider corrections move through a manual tracker before anyone updates the directory
• Month-end reporting gets reconciled by hand before leadership sees a number

None of that is a discipline problem. 

A spreadsheet that routes work, tracks exceptions, or reconciles two sources is the tool you are missing, and whoever built it wrote your requirements for free.

‍

What can you build instead of replacing the system?

The alternative is a connective layer: software that sits on top of the systems you keep and carries work between them.

That usually looks like one of these:

• A customer, member, provider, or vendor portal that reads from the system of record
• An internal workflow tool for the steps that currently happen in email
• Targeted integrations between two systems that already hold the same record
• An exception-handling queue for the cases the core platform has no status for
• Role-based dashboards built on an agreed source of truth
• A reporting layer that removes the manual cleanup before a number is trusted

For BeaconMedaes we designed and built MyMedGas from the ground up, a fully integrated platform layered on top of Beacon Medaes' existing infrastructure, purpose-built to power their new division and grow their footprint in the market. 

No off-the-shelf solution could integrate with the software stack they already had, and none of what they already had needed to be thrown away. 

On the insurance side, Bignition got new screens, workflows, and dashboards that give agents a structured view of their business development activity, all while maintaining compatibility with existing AMS360 agency management systems.

We believe that business is built on transparency and trust, and that good software is built the same way, which in a legacy environment means a team can point at where a number came from without opening four systems to find out.

‍

What should you look at before funding a modernization project?

Before anyone scopes a replacement, walk the work and answer seven questions about where it actually gets stuck.

1. Where does work leave one system and enter another? Those handoffs are where the hours go.
2. Where is a spreadsheet acting like software? Anything that routes, reconciles, or tracks exceptions is a missing tool with a prototype already built.
3. Where does someone enter the same information twice? Duplicate entry means the systems are not carrying context forward.
4. What do customers, members, providers, and vendors call to ask? Repeated status questions mean the outside view and the inside reality have come apart.
5. Which reports get cleaned up by hand before leadership sees them? That points back to data ownership and definitions, which no new platform fixes on its own.
6. Who is the one person who knows how the process really works? Tribal knowledge marks the steps the system never captured.
7. Where would one connection create immediate relief? Modernization can start at the worst handoff instead of the biggest platform.

AI belongs after that work. 

A model asked to summarize a claim or route an exception needs a source of truth, defined permissions, and rules the business has actually settled, and a workflow held together by exports cannot supply any of the three.

There is also a clock on this in health insurance. CMS finalized its Interoperability and Prior Authorization rule on January 17, 2024, with most API requirements landing on impacted payers by January 1, 2027. 

Plans that have spent those years connecting their workflows will be adapting. Plans that spent them on a platform migration will be starting.

Pick the handoff your team complains about most and follow one record across it, writing down every export, re-key, and phone call it takes to get from one side to the other. 

That list is your first project, and it will be smaller than the one your vendor quoted.