Why 2026 is the Year for Zero-Based Budgeting
I had coffee last week with a CFO of a mid-sized construction firm. He spent twenty minutes explaining why his company couldn’t possibly cut costs without “impacting service delivery.” Translation: “We’ve always done it this way, and I have no idea where the money actually goes.”
This is the accounting equivalent of my dad telling me erasers were for kids who didn’t know math. It’s not about the tool—it’s about the thinking.
Here’s what I told him: 2026 isn’t going to care about your incremental budgeting process. The companies that survive the next three years will be the ones who burned their P&L statements and started from scratch.
Let me explain why.
What Zero-Based Budgeting Actually Is (And Why You’ve Been Avoiding It)
Zero-based budgeting isn’t new. Peter Pyhrr developed it at Texas Instruments in the 1970s. The concept is brutally simple: every year, you start your budget from zero and justify every single dollar you want to spend.
No carrying forward last year’s numbers with a 3% bump for inflation. No sacred cows. No “we’ve always had a line item for that.”
Most companies hate this idea. Why? Because it exposes the reality that nobody actually knows what half the organization does or why we’re paying for it.
Traditional incremental budgeting—where you take last year’s budget and adjust it—is the corporate equivalent of rearranging deck chairs on the Titanic. You’re optimizing a structure that may have stopped making sense five years ago.
Here’s the uncomfortable truth: Incremental budgeting was designed for a world where competitive advantage came from operational efficiency. That world is dead.
Your Expense Ratios Just Became Fiction
Let’s talk about the metrics your board obsesses over. You know the ones:
Sales & Marketing as % of Revenue
R&D as % of Revenue
G&A as % of Revenue
Gross Margin targets
EBITDA margins
Here’s the problem: These ratios were calibrated for a world where scaling required linear increases in headcount and infrastructure.
That world is disappearing.
I track a different metric now: Revenue per Head and Cash per Head. Let me show you why traditional ratios are breaking:
Traditional thinking: Walmart does $303K per head. Delta Airlines: $572K per head. AECOM (construction): $269K per head.
New reality: Google: $1.6MM per head. Microsoft: $1MM per head.
Notice something? The technology companies aren’t just 2x better—they’re 4-6x better. And that gap is about to widen dramatically.
But here’s where it gets interesting: What happens when AI allows a construction company to operate at Google’s revenue-per-head ratio?
Why Ratios Embed Assumptions That No Longer Hold
The traditional approach—taking historical expense ratios and extrapolating them against growth projections—assumes a fundamental continuity of how work gets done. That assumption is now shattered.
When a company historically spent 12% of revenue on customer support, that ratio reflected a world where every support ticket required human attention. If AI handles 70% of those tickets, the ratio becomes meaningless—not just different, but categorically irrelevant as a planning tool.
More subtly, the ratio approach treats efficiency gains as incremental. You gradually reduce costs by 2% annually through better spreadsheets. Nice story. Wrong model.
AI creates step-function changes.
You don’t gradually reduce your accounting department’s costs by 2% per year. You suddenly don’t need half of them when AI handles reconciliation, audit prep, and anomaly detection. Then six months later, you don’t need another third when AI starts doing variance analysis and fraud detection.
The ratio dies. The work remains. The humans shift to something entirely different—or they’re gone.
The Top-Down Trap
Top-down planning says “we’ll grow revenue 30%, so expenses should grow roughly X%.” This preserves your existing organizational structure as destiny. It asks “how much will this department cost at scale?” rather than “do we need this department at all?”
The Deming paradox I wrote about in “Creating the Intangible Organization” applies here. Organizations spent decades training people to be highly repeatable process executors. Now the processes themselves are being absorbed by machines. Top-down planning optimizes the wrong thing.
You’re essentially budgeting for increasingly efficient execution of work that doesn’t need humans anymore.
Suddenly, your carefully benchmarked “industry standard” of spending 15% of revenue on G&A becomes irrelevant. If you can automate 70% of your administrative functions, why would you spend 15%? Why not 5%? Or 2%?
The ratio itself becomes the wrong question.
This is why zero-based budgeting matters in 2026. You can’t start with “we spent X% last year, so let’s spend X% again.” You have to ask: “What do we actually need to spend to deliver value to customers?”
And in an AI-enabled world, that number might be radically different from anything your industry has ever seen.
The Vocabulary Problem: When “Big” Stops Meaning What You Think
When someone asks “How big is your company?” the answers usually sound like this:
“We’re going for our Series A”
“360 lawyers”
“2,000 engineering professionals”
“$5B/year in construction”
“1,200 stores”
Rarely do you hear: “$100MM in revenue with 40% gross margin, $10MM in EBITDA, and $250K cash per head.”
The vocabulary we use reveals what we think success looks like. And right now, most companies are measuring the wrong things.
Headcount used to be a proxy for capability and capacity. More people meant more work could be done. AI is severing that connection.
In 2026, a 50-person company might generate the same revenue as a 500-person competitor from 2020. What’s your “G&A as % of revenue” supposed to be then? What’s your “acceptable” burn rate? What’s a “healthy” sales expense ratio?
The benchmarks break.
This is why incrementalism is dangerous. If you’re budgeting based on industry ratios that were calibrated for pre-AI operations, you’re essentially planning to be uncompetitive.
The Moneyball Effect Is Coming for Your Budget
In my book “Creating the Intangible Organization,” I wrote about the Moneyball effect in baseball. Billy Beane revolutionized the Oakland A’s by using analytics to find undervalued players. For seven years, his approach was a massive competitive advantage.
Then everyone else read the book. Suddenly, every team had their own analytics department. Beane’s secret sauce became table stakes, and the A’s winning percentage dropped from .586 to .470.
This is exactly what’s happening right now with operational efficiency. AI isn’t just making operations better—it’s making best practices in operations a commodity that anyone can access.
When I ask executives about AI disruption, over 90% now tell me they’re not questioning if AI will transform their business, but when and how. They’re right to think this way. But most of them are asking the wrong follow-up questions.
The question isn’t “How do we use AI to do what we’re currently doing, but better?”
The question is: “If we were starting this company from scratch today with AI available, what would we build?”
That’s zero-based thinking. And that’s why 2026 is the inflection point.
Four Forces Converging in 2026
Four massive drivers are colliding to make zero-based budgeting not just useful, but essential:
1. AI Democratizes Operational Excellence
Remember when having a great ERP system was a competitive advantage? When supply chain optimization required hiring BCG? When customer analytics meant building a data science team?
Those advantages are evaporating.
AI is creating what I call “abundant operational efficiency.” Every company—from the Fortune 500 to the three-person startup—can now access best-in-class processes. ChatGPT can analyze your supply chain. Robots can optimize your warehouse. AI can manage your accounting close process.
This creates a fundamental shift: Operational efficiency is no longer a differentiator. It’s the baseline.
When I worked with Softwear Automation (the robotics apparel company), we saw this firsthand. The robots could execute “best practices” that previously required years of institutional knowledge. Once that’s automated, the only question that matters is: What are humans doing that robots can’t?
And that question requires zero-based thinking.
More importantly, it destroys your expense ratio assumptions.
If you’ve been running 25% gross margins because “that’s what our industry does,” but AI can automate your most expensive processes, you’re not looking at a 27% margin. You might be looking at 45%. Or 60%.
But you’ll never get there by incrementally adjusting last year’s budget. You need to rebuild from zero and ask: “What would this cost structure look like if we designed it today?”
2. Geopolitical Fragmentation Is Rewiring Supply Chains
The days of “design in California, manufacture in China, sell globally” are ending. Geopolitical tensions, tariffs, and regional conflicts are forcing companies to reconsider every assumption about their supply chains.
This isn’t an incremental adjustment problem. It’s a fundamental redesign problem.
You can’t take your existing supply chain budget and tweak it by 5%. You need to ask: If we’re manufacturing closer to end markets, what does our entire cost structure look like? How does regional manufacturing change our logistics spend? Our inventory financing? Our quality control?
These questions can’t be answered by looking at last year’s numbers. They require zero-based analysis.
And here’s where it connects to the expense ratio problem: Your competitor who’s rebuilding their supply chain from scratch will have completely different economics than you.
If they’re using AI-optimized regional manufacturing with 80% fewer managers, their cost structure won’t look anything like the “industry benchmark.” They’ll have lower costs and higher margins. Not because they negotiated better deals—because they redesigned the entire model.
3. Economic Uncertainty Demands Scenario Planning
We’re heading into 2026 with more economic uncertainty than we’ve seen in decades. Inflation rates, interest rates, currency fluctuations—all are volatile and unpredictable.
Traditional budgeting assumes tomorrow will be a variation of today. Zero-based budgeting assumes we might need to operate in a completely different economic environment.
When I ran RCMS, I learned this lesson the hard way. The 2008 crash didn’t just reduce our revenue—it fundamentally changed what customers needed and how they wanted to buy. Our budget assumptions from 2007 were completely useless.
Companies that survive 2026’s economic turbulence won’t be the ones who “cut 10% across the board.” They’ll be the ones who asked: “What must we spend to serve customers in this new reality?”
That’s a zero-based question.
4. Innovation Requires New Business Models, Not Optimized Old Ones
Here’s the thing about AI, robotics, and advanced technology: They don’t just make existing processes better. They enable entirely new business models.
Look at what happened with Webvan versus Amazon Fresh. Webvan raised $800 million in the dot-com era to do online grocery delivery. It crashed and burned. Same concept, re-launched by Amazon in 2007, became a success.
The difference wasn’t execution—it was timing and business model. The world wasn’t ready for Webvan. By the time Amazon Fresh launched, the infrastructure, consumer behavior, and unit economics had shifted.
In 2026, we’re entering a similar moment. AI isn’t just making your existing business 20% more efficient. It’s making entirely new business models viable.
But you can’t discover those new models by looking at last year’s P&L and asking “what should we cut?”
You have to start from zero and ask: “What business are we actually in? What value do we actually create? What would we build if we started today?”
The Real Reason Your CFO Hates This Idea
Let’s be honest about why zero-based budgeting is resisted: It’s threatening.
When you start from zero, you have to justify everything. That means:
The VP of Operations has to explain why they need 47 people instead of just saying “we had 45 last year”
The IT department has to prove their $2M software renewal is essential, not just automatic
The CFO has to admit they’re not sure why accounting takes 12 people to close the books
Zero-based budgeting exposes organizational bloat, political fiefdoms, and the reality that many expenses exist because they always have, not because they create value.
This is why it rarely happens. It’s not because it’s hard—it’s because it’s honest.
But here’s the thing: AI doesn’t care about your org chart politics.
If a competitor can deliver the same service with 1/3 of your cost structure because they used zero-based thinking and you didn’t, the market will decide for you.
What Zero-Based Budgeting Actually Looks Like in 2026
Let me give you a practical example from my world.
Traditional budgeting for a software development firm might look like:
Engineers: $2M (same as last year + 5% raises)
Sales: $800K (planning to hire 2 more reps)
Marketing: $400K (industry standard is 15% of revenue)
G&A: $600K (last year’s number adjusted for inflation)
Total: $3.8M
Zero-based budgeting asks different questions:
Engineers: “What if AI can generate 70% of our code? What if our developers become 10x more productive? Do we need 10 engineers or 3?”
Sales: “What if AI can qualify leads, schedule demos, and handle basic objections? What if our reps only focus on closing and relationship building? Do we need more reps or better enablement for existing ones?”
Marketing: “What if AI can generate all our content, optimize our campaigns, and personalize outreach? Do we need a marketing team or a marketing strategist with AI tools?”
G&A: “What if AI handles invoicing, expense management, and routine HR questions? Do we need a full back office or one person orchestrating AI tools?”
New total: Maybe $1.2M. Or maybe $800K.
But here’s the critical insight: That’s not just cost-cutting. That’s a fundamental reimagining of what your company is and how it operates.
And it completely breaks your historical expense ratios. You’re not spending “15% on marketing” anymore. You might be spending 3%. But you’re getting better results because you rebuilt the function from scratch around new capabilities.
But Wait—AI Isn’t Free (And Nobody Wants to Talk About It)
Here’s where most “AI transformation” narratives get dishonest. Let me break the bad news:
Organizations that deeply integrate AI don’t just pay licensing fees—they incur real, ongoing costs:
Compute costs that scale with usage (API calls, tokens, processing)
Integration costs for connecting AI to existing systems
Prompt engineering and orchestration expertise (yes, you need humans for this)
Quality assurance overhead (AI outputs need verification, especially in high-stakes domains)
Data infrastructure costs to feed AI systems effectively
A company using AI at scale might spend $50K-500K+ annually on compute alone, depending on intensity. That’s real money that doesn’t appear in traditional ratio planning.
I call this the “stellar adopter tax.” Early adopters of AI don’t just get benefits—they get a new cost structure that looks nothing like the old one.
The Hidden Labor Shift Nobody Mentions
Here’s the part that really messes with your ratios: AI doesn’t eliminate labor—it relocates it.
The work shifts from execution to:
Designing prompts and workflows
Reviewing and validating AI outputs
Handling exceptions and edge cases
Managing the AI systems themselves
This labor requires different (often more expensive) skills than the labor being displaced.
A company that fires 10 junior analysts making $60K each and hires 3 senior people at $150K to manage AI-driven analysis now has higher labor costs for that function, even with dramatically higher output.
So your “labor as % of revenue” goes UP even as your headcount goes DOWN. Try explaining that to your board using traditional ratios.
This is why ratio-based planning is dead. The categories themselves are unstable.
The Laws of Large Numbers Problem
Here’s a concept that should terrify every CFO who relies on industry benchmarks: Historical data assumes historical conditions.
When someone says “companies at our stage typically spend X% on Y,” they’re describing a world where the fundamental mechanics of work were stable. That world is gone.
The construction industry data I’ve worked with illustrates this perfectly. Historical productivity metrics for construction reflect decades of basically unchanged processes. Those numbers tell you nothing about productivity in a firm using AI for estimation, BIM for coordination, and robotics for fabrication.
Historical data is useful for understanding where you’re coming from. It’s terrible for understanding where you’re going.
An Intangible Organization should expect its ratios to look “wrong” by historical standards. That’s actually the signal that transformation is working.
If your cost structure looks like everyone else’s in your industry, congratulations—you’re average. And average is the new death.
The Intangible Organization Advantage
In “Creating the Intangible Organization,” I wrote about how competitive advantage is shifting from operational efficiency to customer intimacy and human-centered capabilities.
Here’s how that connects to zero-based budgeting:
In the future, your budget shouldn’t be organized around functional departments. It should be organized around value creation.
Instead of:
Sales: $X
Marketing: $Y
Operations: $Z
Try:
Customer Acquisition: $A
Customer Success: $B
Product Innovation: $C
Brand Building: $D
Notice the difference? The first structure is about managing resources. The second is about delivering outcomes.
Zero-based budgeting forces you to think in outcomes, not inputs. And in 2026, when AI can handle most inputs, outcomes are all that matter.
The Activity-Based Framework
Rather than “we need a 5-person finance team because that’s what companies our size have,” the question becomes:
What financial operations require human judgment?
Strategic analysis
Stakeholder communication
Unusual situations
M&A negotiations
What can AI execute autonomously?
Routine reconciliation
Standard reporting
Compliance checking
Anomaly detection
What requires human-AI collaboration?
Complex forecasting
Board materials
Strategic scenario planning
Risk assessment
Then you resource from the ground up based on that analysis.
This isn’t about headcount. It’s about capability deployment. Some capabilities cost $150K in labor. Some cost $5K in compute. Some cost both. The total matters, not the category.
The companies that will thrive aren’t the ones who figure out how to do what they currently do with 10% less. They’re the ones who reimagine what they should be doing entirely.
The New Metrics That Actually Matter
If traditional expense ratios are dying, what should you measure instead?
Here’s what I track now:
1. Throughput Per Dollar of Total Cost (Labor + Compute + Infrastructure)
This acknowledges that AI shifts costs between categories. A company might reduce headcount but increase compute spend. The question isn’t “how many people?” but “what’s our total cost to achieve this output?”
Traditional ratios hide this. “Labor as % of revenue” dropping looks great until you realize compute costs quintupled.
2. Cash Profit Per Head
Not revenue per employee. Not gross margin. Actual cash generated per person.
This tells you if you’re building a machine that creates value or a machine that creates jobs. There’s a difference.
3. Value-Add Labor Ratio
What percentage of labor costs go to tasks that require human judgment versus tasks that could theoretically be automated?
This reveals transformation opportunity and organizational drag. If 70% of your labor spend is on automatable tasks, you’re sitting on a gold mine—or a time bomb, depending on how fast you move.
4. Exception Rate
In an AI-augmented operation, what percentage of work requires human escalation?
This measures both AI capability and process design quality. A well-designed AI system should handle 80-90% of routine work autonomously. If your exception rate is 40%, either your AI is poorly implemented or your processes are too complex.
5. Speed-to-Capability
How fast can you deploy new capabilities?
In an AI-native company, adding a “function” might mean configuring a new AI workflow, not hiring and training a team. The cost structure is radically different, and the speed is 10x faster.
If it still takes you six months to launch a new product feature, you’re not AI-enabled—you’re AI-decorated.
6. Revenue from Products That Didn’t Exist 24 Months Ago
This measures innovation capacity. If it’s zero, you’re dying. If it’s over 30%, you’re thriving.
AI should accelerate your ability to experiment, launch, and iterate. If it’s not showing up here, you’re using AI wrong.
7. Percentage of Operating Expenses That Are Variable
How much of your cost structure can you scale up or down quickly?
In a volatile environment, fixed costs are death. AI-driven operations should be more variable because you’re paying for compute usage, not maintaining permanent headcount.
Notice something about these metrics? None of them are ratios to revenue. They’re measures of value creation, adaptability, and efficiency.
And none of them can be properly managed through incremental budgeting.
How to Actually Do This (Without Creating Chaos)
I’m not naive. You can’t just walk into the office on January 1st and announce “we’re starting from zero!”
Here’s how to approach it:
Phase 1: The Pilot (Q1 2025)
Pick one department. Just one. Run zero-based budgeting for that function. Ask:
What outcomes are we trying to achieve?
What’s the minimum spend required to achieve them?
What could we accomplish if we rebuilt this from scratch?
What percentage of current work requires human judgment?
Where could AI create step-function improvements?
Phase 2: The Learning (Q2 2025)
Share what you learned. Be honest about what was uncomfortable, what was illuminating, and what you discovered.
More importantly, share the cost shifts. If you reduced headcount by 4 but added $80K in AI infrastructure, say so. If productivity doubled but so did compute costs, acknowledge it.
This builds organizational muscle for the broader exercise and prevents the false narrative that “AI is free.”
Phase 3: The Commitment (Q3 2025)
Start planning for a full zero-based budget for 2026. This gives you time to:
Train managers on the process
Set clear outcome metrics
Identify areas where AI/automation can fundamentally change operations
Budget for AI infrastructure costs explicitly
Prepare the organization for honest conversations
Map activities to human-essential vs. automatable
Phase 4: The Execution (Q4 2025)
Build your 2026 budget from zero. Not from last year’s numbers. From zero.
For every function, ask:
What value does this create?
What activities are required to create it?
Which activities require human judgment?
What’s the total cost (labor + compute + infrastructure)?
How does this compare to historical spending?
If it’s radically different, why? (And is that good or bad?)
Yes, it’s more work. Yes, it’s uncomfortable. Yes, it will reveal things you wish you didn’t know.
But it’s also the only way to build a budget that reflects reality in 2026 instead of history from 2019.
The Practical Implications for Leaders
If you’re building an Intangible Organization, here’s what this means in practice:
1. Plan from activities, not ratios
Map every function to its essential activities. Determine which are human-essential. Resource accordingly. Ignore industry benchmarks that reflect pre-AI operations.
2. Budget for AI as infrastructure, not magic
Compute costs are real, ongoing, and scale with sophistication. Plan for them explicitly. Don’t hide them in “technology” line items. Make them visible so you can optimize them.
3. Expect volatility in ratios during transformation
Your cost structure will look “wrong” compared to historical benchmarks. That’s the point. If you’re transforming and your ratios still match industry standards, you’re not actually transforming.
4. Measure what matters now, not what mattered before
$Revenue/Head and $Net Cash/Head cut through the noise of shifting cost categories. So does Throughput per Total Cost. Traditional ratios obscure the real economics.
5. Accept that some historical “efficiency” was actually dysfunction
Needing 30 people to do accounts payable wasn’t a feature of scale—it was a limitation of available tools. Don’t optimize for maintaining limitations.
The fundamental shift is from “how do we resource the work we’ve always done?” to “what work actually needs doing, and what’s the right way to do it now?”
Zero-based, activity-driven planning forces that question. Ratio-based extrapolation obscures it.
The Question You Need to Answer
Here’s what it comes down to:
Do you want to manage your company based on what it was, or what it needs to become?
Incremental budgeting is about preservation. Zero-based budgeting is about evolution.
In stable times, preservation works. But 2026 won’t be stable. AI is rewriting the rules of work. Geopolitical shifts are rewriting the rules of trade. Economic volatility is rewriting the rules of finance.
Your expense ratios from 2023 are artifacts of a world that no longer exists.
The companies that thrive in 2026 won’t be the ones who managed to cut costs by 7%. They’ll be the ones who rebuilt their entire operating model around new realities.
That’s not an incremental adjustment. That’s zero-based thinking.
Starting From Zero Means Starting With Purpose
Here’s the final piece: Zero-based budgeting isn’t really about money. It’s about purpose.
When you start from zero, you have to answer the most fundamental question: Why do we exist?
Not “what did we do last year?” Not “what’s our market share?” But: What value do we create that wouldn’t exist without us?
In my work with startups and established companies, I’ve found that the most successful organizations are the ones who can answer that question clearly. They’re the ones who know what they’re for, not just what they do.
When you know your purpose, budgeting becomes simpler. You fund the things that serve that purpose. You eliminate the things that don’t. You rebuild from zero around what matters.
That’s the real reason 2026 is the year for zero-based budgeting.
Not because AI makes it possible. Not because the economy makes it necessary. But because in a world of radical change, the only sustainable competitive advantage is knowing who you are and what you’re about.
Everything else is just expense ratios.
And those? Those are fiction now.
What’s your company’s sacred cow expense? The one that everyone knows is bloated but nobody will touch? What ratio are you still tracking that stopped being meaningful two years ago? Email me. I’d love to hear your stories as we all navigate this transition together.
-KP