You’ve spent $400,000 on your current ERP. Eighteen months of implementation. Countless hours of staff training. And it still doesn’t do what you need it to do.
So what’s the logical next step? Most leaders double down. They hire more consultants, buy more add-ons, and schedule more training sessions. Not because the math supports it, but because walking away feels like admitting failure.
This is the sunk cost trap in action, and it’s quietly bleeding mid-market companies dry. According to Gartner, more than 70% of ERP initiatives fail to fully meet their original business goals. Meanwhile, most ERP projects exceed their initial budgets by three to four times, according to RubinBrown’s analysis of industry benchmarks. That’s a staggering amount of capital tied up in systems that aren’t delivering.
The question isn’t whether your system has problems. It’s whether you’re making future decisions based on future value, or based on money you already can’t get back.
Why Smart Leaders Make This Mistake
The sunk cost fallacy isn’t a sign of poor judgment. It’s a deeply wired cognitive bias that affects even the most experienced decision-makers.
Behavioral economists Daniel Kahneman and Amos Tversky documented this through their prospect theory research: humans experience the pain of a loss roughly twice as intensely as the pleasure of an equivalent gain. Abandoning a system you’ve invested heavily in triggers that loss response hard. It feels like flushing money down the drain, even when the system is actively costing you more every month.
Barry Staw’s research at UC Berkeley added another layer. He found that managers who were personally responsible for an initial investment decision were significantly more likely to commit additional resources to a failing project than managers who inherited that same decision. That’s not logic. That’s ego protection dressed up as strategy.
Here’s what this looks like in practice. A 1985 study by Hal Arkes and Catherine Blumer asked participants to imagine they’d accidentally booked two ski trips for the same weekend: a $100 trip to Michigan and a $50 trip to Wisconsin. Researchers told participants they’d enjoy the Wisconsin trip more. The majority still chose Michigan. The price tag, not the expected value, drove the decision.
The same pattern plays out in boardrooms every quarter. Teams stick with underperforming business systems because switching feels like waste, not because staying is actually cheaper. Intel recognized this exact dynamic in 1985 when CEO Andy Grove asked a now-famous question: if a new CEO walked in, what would they do? The answer was obvious: exit memory chips and pivot to microprocessors. That willingness to override sunk cost thinking turned Intel into the dominant chipmaker of the PC era.
Your business system probably isn’t a memory chip business. But the principle holds. The right question isn’t “how much have we spent?” It’s “what would a new leader do with this situation today?”
Five Warning Signs Your System Is Costing More Than It’s Worth
Recognizing the sunk cost trap is one thing. Knowing when you’ve actually crossed the line from “fixable problems” to “time to move on” is another.
Before you can make that call, you need an honest assessment of where things stand. For many organizations, this means bringing in outside perspective. Working with dynamics 365 consulting services or similar advisory partners can surface operational gaps that internal teams have normalized over time. Fresh eyes catch what familiarity hides.
That said, there are clear signals that show up consistently across struggling implementations. Watch for these five:
- Your workarounds have workarounds. Every system needs a few creative fixes. But when your team maintains spreadsheets to supplement the CRM, manually reconciles data between modules, and has built an entire shadow process outside the platform, you’re no longer using the system. You’re working around it. That’s a maintenance burden disguised as normal operations.
- Integration costs keep climbing with no end in sight. ServiceNow research found that 44% of technology professionals rate difficulty integrating with newer technologies as one of the highest costs of legacy systems. If connecting your current platform to anything modern requires custom middleware, expensive consultants, or months of development time, the architecture itself is the bottleneck.
- Your best people are leaving (or threatening to). This one’s easy to miss because it shows up in HR metrics, not IT budgets. A survey by Stripe found that the typical developer spends roughly a third of their work week dealing with technical debt. When talented employees spend their days patching outdated systems instead of doing meaningful work, they start looking elsewhere.
- You can’t get a straight answer from your own data. Different departments pull different numbers from the same system. Reports take days to compile. Nobody trusts the dashboards. When your business system can’t produce reliable, timely data, every decision built on that data carries hidden risk.
- The vendor has moved on, even if you haven’t. When your platform version is no longer actively supported, when security patches arrive months late (or not at all), when the vendor’s roadmap has shifted away from your use case, you’re maintaining a system that’s becoming more expensive and more vulnerable with each passing quarter.
If three or more of these describe your situation, you’re not dealing with a system that needs tuning. You’re dealing with a system that’s past its useful life.
The Real Price of Standing Still
Most leaders underestimate the cost of keeping a bad system because the biggest expenses are invisible.
The direct costs are easy enough to track: license fees, hosting, support contracts, the occasional consultant to fix something broken. According to Gartner, Forrester, and Deloitte benchmarks, legacy maintenance typically consumes 60 to 80% of total IT spend across enterprise portfolios. A 2025 GAO report on U.S. federal IT found that 79% of a $105 billion annual IT budget went to operations and maintenance, leaving a fraction for actual innovation.
But the indirect costs are where the real damage happens. ITIC’s 2024 survey found that over 90% of mid-size and large enterprises report a single hour of downtime exceeds $300,000 in costs, and 41% report costs above $1 million per hour. Those figures don’t include litigation, regulatory penalties, or the customer relationships that quietly erode when service suffers.
Then there’s the opportunity cost. Every dollar and every hour your team spends patching legacy problems is a dollar and an hour not spent on growth. According to a 2024 IDC report, enterprises maintaining legacy systems spend up to 42% more on operational overhead compared to those running modern, supported platforms.
The math gets worse over time, not better. Legacy hardware maintenance costs increase 10 to 15% annually after warranty expiration. Premium support for end-of-life systems can run 50 to 200% more than standard support. Talent scarcity compounds the problem: specialized contractors for legacy platforms command $180 to $250 per hour in 2026, up from $120 just four years ago.
Standing still feels safe. Financially, it’s anything but.
And there’s one more cost that rarely makes it into the spreadsheet: security exposure. IBM’s Cost of a Data Breach Report 2025 found that the global average cost of a data breach reached $4.44 million. Organizations running outdated systems with limited logging, older encryption standards, and delayed patching are structurally more vulnerable to attacks. Cyber insurers have caught on too, with several major carriers now applying surcharges or payout caps for breaches involving unsupported legacy systems.
How to Make the Switch Decision Without Emotion
The hardest part of abandoning a system isn’t the technology. It’s overcoming the psychological resistance to doing it. Here’s a framework that strips the emotion out:
Run a zero-based evaluation. Pretend your current system doesn’t exist. If you were starting from scratch today, with your current business needs, team size, and growth plans, would you choose this platform? If the honest answer is no, the sunk costs are irrelevant.
Calculate total cost of ownership over three years, not one. Include everything: direct maintenance, staff time spent on workarounds, integration costs, opportunity costs of delayed projects, and risk exposure from security gaps. Most leaders who do this exercise discover that migration pays for itself within 18 to 24 months.
Use Intel’s “new CEO” test. Andy Grove’s question works because it separates the decision from the decision-maker’s ego. Ask your leadership team: if someone new walked in tomorrow with no history with this system, what would they recommend? The answer usually comes fast.
Set a decision deadline with clear criteria. Sunk cost thinking thrives on ambiguity. Define specific, measurable conditions that would trigger a switch, and set a date to evaluate them. Something like:
- If integration costs exceed $X in the next two quarters, we migrate
- If data accuracy issues cause more than Y reporting errors per month, we migrate
- If employee satisfaction scores in IT drop below Z, we migrate
When the criteria are set in advance, the decision becomes mechanical, not emotional.
Separate the “who decided” from the “what should we do.” Staw’s research showed that personal responsibility for a past investment makes people cling to it harder. If the person who championed the original system is also the person evaluating whether to keep it, you’ve got a structural bias built into the process. Bring in a neutral party, whether that’s an advisory board member, an outside consultant, or a peer executive from another division.
What a Healthy Transition Actually Looks Like
The fear behind the sunk cost trap isn’t just about wasted money. It’s about the pain of switching. Migration horror stories are real, and they give leaders a convenient reason to stay put.
But the data tells a different story when the transition is handled properly. Panorama Consulting reports that 97% of organizations see improvements after a successful implementation. RubinBrown’s research found that organizations engaging experienced consultants report an 85% success rate, compared to the industry-wide average where roughly half of implementations fail on the first attempt.
The difference between a painful transition and a smooth one typically comes down to a few factors:
- Phased rollout over big-bang launches. Over 50% of companies now prefer phased implementation, deploying module by module rather than switching everything overnight. This reduces disruption and gives teams time to adapt.
- Realistic timelines. For mid-market companies, expect 3 to 9 months for implementation. Enterprise-level projects can run up to 18 months. If someone promises faster, ask hard questions.
- Change management from day one. The technical migration is only half the job. The human side (training, communication, workflow redesign) determines whether your team actually uses the new system or quietly reverts to spreadsheets.
- Clean data migration. One manufacturing company discovered during migration that products listed as available didn’t exist, and discontinued items were actually best sellers. The new system worked perfectly with garbage data. Invest the time to audit and clean your data before it moves anywhere.
None of this is simple. But the track record is clear: organizations that plan carefully and invest in expert guidance come out ahead. The companies that stumble are typically the ones that treat migration as a purely technical project and ignore the organizational side entirely. Or worse, the ones that rush the timeline to “get it over with” and end up right back where they started, with another system nobody trusts.
One pattern worth noting: the organizations with the smoothest transitions tend to be the ones that involved end users early in the process, not just during training. When the people who’ll actually use the system every day have input on requirements and workflow design, adoption rates increase dramatically and resistance drops.
Making the Call
Here’s the uncomfortable truth. That $400,000 you already spent? It’s gone whether you switch or stay. The only money you can still influence is what you spend from this point forward.
If your current system is generating more friction than value, if your team is spending more time on workarounds than on real work, if your data can’t be trusted and your vendor has moved on, then continuing to invest in that platform isn’t loyalty. It’s the sunk cost trap doing exactly what it does best: making an irrational decision feel responsible.
The leaders who build resilient, competitive organizations aren’t the ones who never make bad technology bets. They’re the ones who recognize a losing hand early and have the discipline to fold.
Run the numbers. Set your criteria. Make the call based on where you’re going, not where you’ve been.
