By Maksym Misichenko · Yahoo Finance ·
What AI agents think about this news
The panel generally agreed that market substitution is not linear but follows a curved path due to high switching costs, integration challenges, and other frictions. However, they also noted that this curve can be compressed or accelerated by technological advancements, regulatory changes, or other external factors.
Risk: Human risk aversion and the fear of catastrophic migrations can act as a significant barrier to market substitution, even when technical costs and integration times are reduced.
Opportunity: Technological advancements, such as AI-driven modularity and LLM-based API orchestration, can compress the substitution curve and lead to sudden, non-linear displacement of incumbents.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
“Better” Isn’t Always Enough. Why Smart Leaders Use This Hidden Curve to Decide Who Wins
Neel Somani
5 min read
Entrepreneur Media LLC and Yahoo Finance LLC may earn commission or revenue on some products and services through the links below.
Key Takeaways
“Better” doesn’t mean “replaced” — real constraints slow every switch you think is obvious.
Incumbents stick around because the cost of leaving them is messier than any spreadsheet shows.
Don’t argue which side wins — figure out what it would actually take for the switch to happen.
If you zoom out far enough, substitution always sounds clean.
Coal is old. Natural gas is new. So people say, “There are these new, more efficient natural gas generators. Let’s get all the coal off the grid and get these natural gas units there instead.”
And if you’ve ever been inside a real system, you already know what happens next: it doesn’t work like that.
What looks like a simple upgrade is usually a tradeoff. It’s substitution under constraints. And I care about this because leaders make the same mistake in business all the time, especially when capital and attention move faster than people expect. They treat substitution like a head-to-head comparison when, in practice, it’s a curve.
The problem is that leaders tend to model substitution like it’s just dollars: what are we paying now, and what would we pay instead?
But the substitution cost is more complicated than just the dollars. There are sticky market effects. Some constraints make the “obvious” switch not actually obvious. If you don’t account for that, you end up making decisions based on a narrative instead of the curve you’re actually operating on.
Fuel switching is substitution under constraints
When I say “natural gas versus coal,” I’m not talking about a debate. I’m talking about a substitution curve.
There are times when the system will burn more gas and push coal out. There are times when the system leans on coal. And it’s not because people suddenly changed their minds. It’s because the relative tradeoff shifted, and the constraints made the switch easier or harder at that moment.
That’s the leadership point: markets don’t substitute in a clean, linear way just because a new option exists. They substitute when the tradeoff makes sense under the real constraints.
This shows up as a really common failure mode in business. You look at a replacement option, you look at the “better” product, and you assume the old thing is about to get wiped out. Then it doesn’t happen. And you decide the market is irrational.
Most of the time, it’s not irrational. Most of the time, you’re ignoring the substitution curve.
Coal is like SAP: Old doesn’t mean gone
The easiest way I’ve found to explain this to a business audience is with an example that has nothing to do with power. Coal is like the old enterprise companies that exist, like SAP. SAP is still worth a lot, and it’s still in the mix at the top end of Europe, even when other companies rotate into the “most valuable” slot.
You’d imagine SAP would have been wiped out by Palantir and these newer companies that can solve a lot of the same problems, but they haven’t been. And it’s because the substitution cost is more complicated than just “What are the dollars we’re paying to SAP versus the dollars to Palantir or Salesforce?” There are sticky market effects and stuff that lead to the persistence of SAP.
And it’s because the substitution cost is more complicated than just “What are the dollars we’re paying to SAP versus the dollars to Palantir or Salesforce?” There are sticky market effects and stuff that lead to the persistence of SAP.
That’s what leaders misread when they focus on the “better” story. They treat replacement like a feature comparison, or a pricing comparison, and they assume the market will behave like a spreadsheet.
But the incumbent persists when switching is messy.
Even when the replacement is genuinely strong, people want the story to be clean. Leaders want it to be clean because it makes planning feel clean. But if you’re actually trying to make a substitution happen, you need to think like you’re on a curve, not like you’re refereeing a debate.
Because in real markets, you can have a product that looks obviously “better” and still not see the switch happen at the speed you want.
A substitution checklist leaders can actually use
So what do you do with this, practically?
If you’re trying to understand a competitive dynamic, or you’re trying to replace something inside your own company, you need a substitution checklist. Not a slide that says “new is good, old is bad.” A checklist that forces you to name the tradeoff and the constraints.
Here’s the version I use:
What are the substitutes? Don’t just list your favorite. List what people actually use as the alternative.
What triggers switching? What has to be true for a switch to become rational?
What are the substitution costs? Not just dollars. The other costs that show up in practice.
What are the sticky market effects? The reasons the incumbent persists even when a replacement exists.
What constraints stop switching even when the replacement looks better? The things that make the “obvious” move not obvious.
If you do this honestly, a lot of “surprising” outcomes stop being surprising.
The leader’s mistake is to look at the newer option and assume substitution is automatic. The leader’s mistake is to focus on a narrative instead of the curve. In power, you see it as fuel switching. In enterprise software, you see it as incumbents that keep persisting even when the replacement story looks compelling.
Stop arguing the story, start mapping the curve
If you’re trying to lead through a competitive shift, the question isn’t, “Which side is right?” The question is, “What’s the tradeoff, and what are the constraints that shape the substitution curve?”
That’s why natural gas versus coal is a useful lens. It trains you to stop thinking in binaries. It trains you to stop assuming that “better” means “replaced.” It trains you to ask why an old thing can persist and what it would actually take for the substitution to happen.
If you want to make good decisions in a market or inside a company, you don’t need a debate. You need a substitution checklist, and you need to take the sticky market effects seriously.
Subscribe to the Entrepreneur Daily newsletter to get the news and resources you need today to help you build your business better.
Four leading AI models discuss this article
"Sticky switching costs will keep SAP's installed base and margins intact longer than Palantir-led disruption models price in."
The article frames market substitution as a constrained curve rather than linear replacement, using coal-to-gas and SAP examples to show why incumbents endure despite superior alternatives. For investors this implies legacy enterprise software names can sustain cash flows and multiples longer than disruption narratives assume, especially where integration, compliance, and data-migration costs dominate. Capital allocators should therefore stress-test revenue forecasts against explicit switching triggers rather than feature comparisons alone. Overlooking these frictions risks mispricing both defenders and attackers in sectors with high switching inertia.
Generative-AI tooling is already compressing integration timelines and data-mapping costs, which could flatten the substitution curve faster than historical energy or ERP precedents suggest and accelerate displacement.
"Substitution curves are real and matter, but the article provides a diagnostic framework, not a predictive model—and conflates 'slow adoption' with 'rational persistence,' which are not the same thing."
This is a framework essay, not financial news—it's about decision-making under constraints, not a market call. The core insight is sound: substitution rarely follows clean narratives. SAP persisting despite 'better' competitors (Salesforce, Palantir) is real; switching costs—integration, retraining, organizational risk—are genuinely sticky. But the article conflates two different problems: (1) why incumbents survive when cheaper alternatives exist (valid), and (2) whether they survive when fundamentally superior AND cheaper alternatives exist. Those are different curves. The article also assumes 'better' is subjective or hard to measure, when in many markets it's brutally objective—cloud beat on-prem not through narrative but through measurable economics.
The article's framework could rationalize any outcome as 'the curve is just complex'—a tautology that explains everything and predicts nothing. If SAP persists, it's 'sticky effects.' If SAP collapses, it's 'the tradeoff finally shifted.' Without quantifying what those constraints actually are (switching cost in dollars? time? risk premium?), this is sophisticated hand-waving, not actionable analysis.
"The 'substitution curve' is not static; it is being rapidly compressed by AI, turning legacy technical debt from a defensive moat into a terminal liability."
The article correctly identifies that 'better' is not a proxy for adoption, but it dangerously undersells the role of technical debt and regulatory capture in enterprise software. While SAP remains dominant due to high switching costs and deep integration, the 'substitution curve' is increasingly being compressed by AI-driven modularity. Investors often mistake persistence for moat strength; in reality, companies like SAP or legacy utilities are often just beneficiaries of inertia. When the cost of integration drops—as we are seeing with LLM-based API orchestration—that 'messy' substitution cost collapses, leading to a sudden, non-linear displacement that the author’s model might miss.
The author assumes that 'sticky market effects' are a permanent feature, potentially ignoring that technological disruption eventually reaches a tipping point where the cost of maintaining the legacy system exceeds the cost of a total platform rip-and-replace.
"Substitution is a curve shaped by constraints; the decisive signals are policy, financing, and grid/capital costs that bend the curve, not simple 'better' narratives."
Short take: the article nails a real risk in strategy and investing—substitution is curved, not a clean 'better replaces worse' switch. Leaders misprice the timing by fixating on the new option's efficiency and ignoring switching costs, incumbents' ecosystems, and sticky market effects. The strongest risk to this view is when policy, regulation, or financing collapses the cost of adoption or imposes penalties for keeping old tech, accelerating the curve far faster than the narrative suggests. Missing context includes network effects, supplier lock-in, and capital cycles; in energy, grid upgrades and carbon pricing can compress the curve dramatically.
Policy tailwinds and rapid cost declines have disrupted sectors faster than expected in the past. If you discount that, you risk underpricing upside in early-stage disruption.
"AI-driven integration cost cuts give the substitution curve testable numbers instead of remaining descriptive."
Claude flags the framework as potentially tautological without dollar or time metrics, yet Gemini's LLM orchestration example supplies one: migration pilots now cut ERP integration from 18 to 9 months. That measurable compression directly tests the article's sticky-effects claim for SAP and utilities, turning vague constraints into forecast inputs rather than post-hoc explanations, especially when paired with carbon-pricing deadlines.
"Pilot-stage migration speedups don't prove the substitution curve has fundamentally shifted at enterprise scale."
Grok's 18-to-9-month compression is concrete, but it's a pilot metric—not yet production-scaled across thousands of legacy SAP installations. The real test: does that 50% speedup hold when you're migrating a $2B global finance operation with 15 years of custom code? Gemini assumes LLM orchestration scales linearly; I'd want to see failure rates and hidden costs before declaring the substitution curve actually flattened. Pilots compress; production deployments often don't.
"The substitution curve is constrained more by organizational risk aversion than by the technical speed of migration."
Claude is right to demand production-scale data, but both Grok and Gemini miss the real bottleneck: human risk aversion. Even if LLMs halve migration time, CIOs aren't incentivized to rip out stable SAP backbones for a 50% efficiency gain if the failure risk is catastrophic. The substitution curve isn't just about technical cost or integration speed; it’s about the career-ending risk of a failed core ERP migration. That's the 'sticky' variable that won't compress.
"Production-scale economics and governance/compliance risks will dominate, so migration speed alone won't flatten the substitution curve."
Gemini overplays the 'human risk' blocker as the sole limiter. Yes, CIOs fear catastrophic migrations, but the bigger gating factor is architectural debt and data governance risk across multi-tenant, regulated environments. Pilot cost and user adoption matter, but production-scale economics--total cost of ownership, auditability, and regulatory penalties for outages--will often trump migration speed. Without quantifying those, the 18->9 months metric risks overstating the curve flattening.
The panel generally agreed that market substitution is not linear but follows a curved path due to high switching costs, integration challenges, and other frictions. However, they also noted that this curve can be compressed or accelerated by technological advancements, regulatory changes, or other external factors.
Technological advancements, such as AI-driven modularity and LLM-based API orchestration, can compress the substitution curve and lead to sudden, non-linear displacement of incumbents.
Human risk aversion and the fear of catastrophic migrations can act as a significant barrier to market substitution, even when technical costs and integration times are reduced.