your optimization problem has a fixed, hidden force called demand
You just shipped a SaaS tool that saves logistics teams 10 hours a week. Your Cost Per Unit is low, your code is solid, and three early customers love it. You set a goal: 100 paying customers by Q4. You run ads, write content, bring on a salesperson. By Q4 you have 31 customers. You did not have an Execution problem. You had a Demand problem - there were only about 200 logistics teams in North America that fit your target audience, and most of them were not actively looking for a solution.
Demand is the total quantity of your product or service that Buyers are ready and able to purchase at a given Pricing level. It is external to your business, partially hidden, and sets the ceiling on your Revenue - no amount of operational excellence can exceed it.
Demand is the total number of Buyers who will pay for what you sell, at the Pricing you set, in the Time Horizon you care about.
Two things make Demand tricky for Operators:
Demand is not the same as 'people who could theoretically benefit.' It is people who (a) have the problem, (b) know they have the problem, (c) are ready to spend money to solve it, and (d) can actually authorize a purchase. Each filter shrinks the number dramatically.
Demand sets the ceiling on your P&L. Every other optimization you make - Cost Reduction, Throughput improvements, Pricing changes - operates underneath that ceiling.
Consider two scenarios for a business with $2M in annual Revenue where $0.30 of every Revenue dollar reaches Profit ($600K):
Cost Optimization is valuable. But it is bounded by your current Revenue, which is bounded by Demand. Operators who only think Supply-Side (build better, build cheaper) and ignore Demand-Side (who wants this, and how many of them are there) hit a ceiling they cannot diagnose.
On your P&L, Demand shows up as the upper bound of your Revenue Line. When Revenue plateaus despite good Execution, the first question is: did you exhaust Demand, or did you fail to reach it?
Demand has a few mechanical properties that matter for Operators:
1. Demand varies with Pricing.
If you sell a SaaS product at $50/month, some number of Buyers will pay. If you raise Pricing to $500/month, fewer will pay - but each one contributes more Revenue. The relationship between Pricing and quantity is not linear, and you usually cannot observe it directly. You infer it from experiments, Pipeline data, and Close Rate changes when you adjust price.
2. Demand is segmented.
Not all Demand is the same. Through customer segmentation, you will find that some Buyers convert quickly with low Churn. Others take longer, cost more to acquire, and leave sooner. The shape of your Demand determines which segments you should target first and how you Allocate Marketing Spend.
3. Demand has a time dimension.
Some products face seasonal Demand. Others face one-time Demand (the Buyer purchases once and is done). SaaS products aim for recurring Demand, which is why ARR is the metric - it captures the ongoing flow of Buyer purchases. When Demand is one-time, you need a constant supply of new Buyers. When Demand is recurring, you need to prevent Churn.
4. Demand can shift.
New competitors enter and split Market Share. Buyer preferences change. Regulations appear. A cheaper substitute emerges. Demand is not a fixed number - it is a moving quantity you must keep estimating.
You should think explicitly about Demand when:
You run a SaaS product for independent veterinary clinics in the US. Your Pricing is $200/month. You have 400 customers, generating $80K in monthly Revenue ($960K ARR). There are roughly 28,000 independent vet clinics in the US. Your Churn Rate is 3% per month. Your GTM Team adds about 30 new customers per month.
Step 1: Model net monthly customer growth. In month 1, new customers: 30. Churned customers: 400 × 0.03 = 12. Net gain: 18. But Churn is 3% of a growing base, so net growth shrinks every month. By month 6, your base is roughly 500 and Churn is 15/month, yielding net growth of only 15. By month 12, Churn has risen to roughly 17/month and net growth has slowed to roughly 13.
Step 2: Project the trajectory. Each month, your base = (0.97 × previous base) + 30. This converges toward a steady state of 30 / 0.03 = 1,000 customers - the point where monthly additions exactly equal monthly Churn. After 12 months, your base reaches roughly 580 customers. Revenue: 580 × $200 = $116K/month (roughly $1.39M ARR). Note: a naive projection that holds net growth constant at 18/month would predict 616 customers - it overstates by 6% at month 12, and the error compounds from there.
Step 3: Estimate the Demand ceiling. 28,000 clinics exist, but not all are Buyers. Assume 40% have the problem your product solves, 50% of those know they have it, 30% of those are ready to pay for software. That gives 28,000 × 0.4 × 0.5 × 0.3 = 1,680 addressable Buyers. Where do these percentages come from? Customer interviews, industry surveys, or data from analogous SaaS products in similar verticals. They are estimates - the point is having an explicit model you can update as you learn, not pretending the numbers are precise.
Step 4: Compare trajectory to ceiling. Your steady state (1,000) is below the Demand ceiling (1,680), so growth will not stall from Churn alone. But your Pipeline will start shrinking before you reach 1,000 because the remaining Buyers are harder to find and more expensive to convert. At 580 customers after year one, you hold 35% of addressable Demand (580 / 1,680).
Step 5: To grow beyond the steady state, you need to either expand Demand (new segments, new geographies, new product features that convert non-Buyers into Buyers), increase Revenue per customer through Upsell, or reduce Churn Rate to push the steady state higher (at 2% Churn, steady state = 30 / 0.02 = 1,500).
Insight: Revenue growth that looks like an Execution problem is often a Demand problem. When your addressable market is smaller than your ambition, no amount of GTM optimization closes the gap. And when Churn compounds against a growing base, the naive 'net new per month' projection overstates your trajectory. Model the decay.
You sell an analytics tool at $99/month and have 1,000 customers ($99K monthly Revenue). You are considering raising Pricing to $149/month. Your best estimate from customer surveys and Churn data is that 15% of customers would cancel at the new price, and new customer acquisition would slow by 20%.
Step 1: Calculate immediate Revenue impact. Remaining customers: 1,000 × 0.85 = 850. New monthly Revenue: 850 × $149 = $126,650. That is a 28% Revenue increase despite losing 150 customers.
Step 2: Calculate acquisition impact. If you currently add 50 customers/month, the new rate is 50 × 0.80 = 40 customers/month. Each new customer contributes $149 instead of $99, so monthly Revenue from new customers is $5,960 vs. $4,950 - still higher per month of additions.
Step 3: Calculate Lifetime Value shift. If average customer lifespan was 24 months at $99 (Lifetime Value = $2,376), and at $149 the lifespan drops to 20 months due to higher Churn (Lifetime Value = $2,980), the higher Pricing still produces more value per customer.
Step 4: Check the Demand ceiling. At $149, fewer total Buyers exist. If original addressable Demand was 5,000 Buyers at $99, it might be 3,500 at $149. Your maximum Revenue shifts from $495K/month to $521K/month - a higher ceiling despite fewer Buyers.
Insight: Pricing changes reshape Demand. Fewer Buyers at higher Pricing can produce more total Revenue - but only up to a point. The Operator's job is to find the Pricing level that maximizes Revenue (or Profit) given how the number of Buyers changes with price, not the Pricing that maximizes customer headcount.
Demand is external and partially hidden - you can influence it through positioning and Marketing Spend, but you cannot create it from nothing. Estimating Demand accurately is one of the highest-value analytical skills an Operator can develop.
Revenue has a ceiling set by Demand. Cost Optimization works below that ceiling. If you want step-function growth, you need to expand Demand (new segments, new products, new geographies) or capture more Market Share.
Always separate Demand problems from Execution problems. If Pipeline Volume is healthy and Close Rate is dropping, you have an Execution problem. If Pipeline Volume is shrinking and Close Rate is steady, you have a Demand problem. The interventions are completely different.
Treating Revenue targets as purely internal goals. A team that sets a $10M Revenue target without estimating whether $10M in Demand exists is planning to fail. Always ground targets in an estimate of addressable Demand, even if that estimate is rough.
Confusing total market size with Demand. There are 30 million small businesses in the US. That is not your Demand. Your Demand is the subset that has the problem, knows they have it, will pay to solve it, and can actually buy. The gap between total market and actual Demand is usually 10x to 100x.
You are evaluating whether to build a project management tool for independent bookstores. There are approximately 2,300 independent bookstores in the US. Estimate addressable Demand using reasonable assumptions about what fraction have the problem, know they have it, and would pay $75/month for software. What is the maximum annual Revenue this product could generate?
Hint: Apply filters sequentially. Start with the total count and estimate what percentage passes each filter: has the problem, knows they have it, ready to pay for software. Multiply the surviving count by your Pricing and 12 months.
Assume 60% have project management pain (1,380). Of those, 50% recognize it as a solvable problem (690). Of those, 25% would pay for software vs. using spreadsheets (173 addressable Buyers). Maximum Revenue: 173 × $75 × 12 = $155,700/year. This is a small market - you would need very low Cost Per Unit to make the Unit Economics work, or you need to expand your target audience beyond independent bookstores.
Your SaaS product has 500 customers at $300/month ($150K monthly Revenue, $1.8M ARR). Pipeline Volume has declined 20% over the last two quarters while Close Rate stayed at 25%. Your VP of Sales wants to hire two more salespeople. Your head of marketing wants to double Marketing Spend. Using what you know about Demand, which investment is more likely to help, and why?
Hint: Diagnose first. What does a declining Pipeline Volume with steady Close Rate tell you about where the problem is? Then match the intervention to the diagnosis.
Declining Pipeline Volume with steady Close Rate is a Demand-Side signal, not an Execution problem. The GTM Team is converting at the same rate - they just have fewer opportunities. Hiring more salespeople (a Supply-Side investment) would give more people less to do. Doubling Marketing Spend is closer to the right answer, but only if the spend targets new segments or channels that reach untapped Demand. If it just increases spend in the same channels, you are bidding more for a shrinking pool. The best move is to first estimate remaining addressable Demand in your current segment. If it is nearly exhausted, neither investment helps much without expanding your target audience or product scope to unlock new Demand.
You run a SaaS service at two Pricing tiers: $500/month (Basic, 200 customers, 5% monthly Churn Rate, $1,500 to acquire) and $1,500/month (Pro, 50 customers, 1.5% monthly Churn Rate, $6,000 to acquire). Addressable Demand is 800 Buyers at Basic and 120 at Pro. You have $300K in Budget for customer acquisition this year. Calculate the expected Revenue net of acquisition cost from each tier over a 24-month Time Horizon, accounting for Churn. Which tier should you invest in? Does the answer change if you also consider what your customer base looks like at month 25?
Hint: For each tier: divide Budget by acquisition cost to get new customers acquired. Then model expected Revenue per customer over 24 months - a customer acquired today has a (1 - Churn Rate)^n probability of still paying in month n, so expected months of Revenue = (1 - (1 - Churn Rate)^24) / Churn Rate. Subtract acquisition cost per customer. Then think about what happens after month 24 - how many customers from each cohort are still active and generating Revenue?
Basic: $300K / $1,500 = 200 new customers (within the 600 remaining addressable). Expected months of Revenue per customer over 24 months: (1 - 0.95^24) / 0.05 = (1 - 0.292) / 0.05 = 14.16 months. Revenue per customer: 14.16 × $500 = $7,080. Net of acquisition: $7,080 - $1,500 = $5,580. Total: 200 × $5,580 = $1.116M.
Pro: $300K / $6,000 = 50 new customers (within the 70 remaining addressable). Expected months per customer: (1 - 0.985^24) / 0.015 = (1 - 0.694) / 0.015 = 20.4 months. Revenue per customer: 20.4 × $1,500 = $30,600. Net of acquisition: $30,600 - $6,000 = $24,600. Total: 50 × $24,600 = $1.23M.
Over 24 months, Pro wins ($1.23M vs. $1.116M) despite acquiring 4x fewer customers. Now look at month 25: surviving Basic customers = 200 × 0.95^24 = 58, generating $29K/month. Surviving Pro customers = 50 × 0.985^24 = 35, generating $52.5K/month. Pro produces nearly double the ongoing Revenue from a smaller base. The naive Demand arithmetic (600 remaining × $500 vs. 70 remaining × $1,500) says Basic. The Lifetime Value math says Pro. This is why Churn Rate is a Demand-Side variable, not just an operations metric - it determines how much of the Demand you actually convert into durable Revenue.
Demand connects to Churn in a way Operators often miss: when you lose a customer, they return to the pool of unmet Demand, available for a competitor to capture. Churn is not just a retention metric - it is Demand recycling. Demand also directly constrains your Budget process: Revenue targets set without a Demand estimate are fiction, and every Allocation decision downstream inherits that error.
Disclaimer: This content is for educational and informational purposes only and does not constitute financial, investment, tax, or legal advice. It is not a recommendation to buy, sell, or hold any security or financial product. You should consult a qualified financial advisor, tax professional, or attorney before making financial decisions. Past performance is not indicative of future results. The author is not a registered investment advisor, broker-dealer, or financial planner.