How to Use Expected Value Thinking in Business Decisions

Expected value thinking is a method for weighing uncertain decisions by assigning a probability to each possible outcome and multiplying it by the value of that outcome. You do not need a finance background or a spreadsheet; the process of making your assumptions explicit is usually more valuable than the final number. For most business decisions, a back-of-envelope EV calculation takes under 20 minutes and surfaces the assumption that actually drives the bet.

What Expected Value Actually Means

Most business decisions are not coin flips with known odds. You are deciding whether to hire a sales rep before revenue can support the role, whether to invest in a new marketing channel before you have conversion data, whether to delay a product launch for another round of polish. In every case, you are betting on an uncertain future.

Expected value (EV) is a way to put a rough number on that bet. Multiply each possible outcome by how likely you think it is, then add everything up. The result is a single figure that lets you compare options with different shapes of risk and reward.

The formula:

EV = (Probability of Outcome 1 × Value of Outcome 1) + (Probability of Outcome 2 × Value of Outcome 2) + ...

That is it. What the formula forces you to do is name your outcomes, estimate their likelihood, and assign a value to each. Those three steps are where most of the insight comes from, not the arithmetic.

Why This Beats Pure Gut Feel

Gut feel is not the enemy. It encodes genuine experience, and for fast, low-stakes calls, it is often the right tool. The problem is that intuition handles low-probability, high-consequence events poorly. We overweight outcomes that are vivid or recent and underweight outcomes that feel abstract or distant.

Cognitive biases in business decisions covers several patterns that distort these estimates, including optimism bias (we overestimate our chances of success) and availability bias (we anchor on the most memorable outcome rather than the most likely one). EV thinking does not eliminate these biases, but writing probabilities down forces you to defend them. That friction is what intuition alone does not provide.

A Worked Example: Should You Hire a Sales Rep?

You run a B2B SaaS product at $30,000 in monthly recurring revenue. You are considering hiring a junior sales rep at a fully-loaded cost of $7,000 per month. The question: does the expected value of hiring justify the spend over a 12-month horizon?

You identify three plausible outcomes.

Outcome 1: The hire works well. The rep closes enough new business to add $15,000 in new MRR within six months. That compounds over the year to roughly $90,000 in new revenue against $84,000 in cost. Net value: +$6,000.

You estimate this probability at 35%.

Outcome 2: The hire is mediocre. The rep closes some deals but not enough. You add $3,000 in new MRR after six months, generating roughly $24,000 in new revenue against $84,000 in cost. Net value: -$60,000.

You estimate this probability at 45%.

Outcome 3: The hire fails. The rep produces minimal revenue. You spend $84,000 and net $5,000 in new revenue. Net value: -$79,000.

You estimate this probability at 20%.

Now calculate:

• Outcome 1: 0.35 × $6,000 = $2,100
• Outcome 2: 0.45 × (-$60,000) = -$27,000
• Outcome 3: 0.20 × (-$79,000) = -$15,800

EV = $2,100 - $27,000 - $15,800 = -$40,700

The hire has strongly negative expected value on these estimates. But the exercise has already done its job. It has surfaced exactly which assumption drives the result: the 45% probability of a mediocre outcome. If you can reduce that to 20% by using a 90-day contract-to-hire structure, or by building a proper onboarding track before day one, the numbers shift substantially. You now know where to focus before committing.

This is why EV thinking is worth doing even when the answer is "don't." The calculation tells you what would need to be true to make the bet worth taking.

How to Apply EV Thinking Step by Step

Use this process for any decision with meaningful financial stakes and at least two plausible outcomes.

Step 1: Write the decision as a single sentence with a deadline. "Should we launch a paid newsletter tier by end of Q4?" is a decision. "Should we monetize more?" is a direction. Clarity here saves time later.

Step 2: List two to four plausible outcomes per option. Include the realistic downside, not just "it fails." There is a difference between "the campaign generates zero leads" and "the campaign generates refund requests and damages customer trust." They have very different values.

Step 3: Assign probabilities. Your probabilities must sum to 100%. Use reference points: how often have similar bets worked for companies at your stage? What does your own track record look like on comparable calls? Making decisions under uncertainty has useful framing on calibrating confidence levels over time.

Step 4: Assign monetary values. Translate each outcome into dollars or a close proxy: customer count, months of runway, hours of team time. The goal is comparability, not precision. An order-of-magnitude estimate beats no estimate.

Step 5: Calculate the EV. Multiply probability by value for each outcome and sum them. A positive EV means the bet earns on average across many trials. A negative EV means you need to either restructure the bet or gather more information before committing.

Step 6: Check whether the downside is survivable. Before acting on a positive EV, ask whether any downside scenario would be catastrophic for your business. If yes, treat that scenario as a hard constraint, not just a number in the model. More on this below.

The Biggest Mistake: Ignoring Ruin Scenarios

Expected value assumes you get to run the same bet many times. If you ran the sales rep experiment 100 times across 100 parallel businesses, the average outcome would approach the EV. But you have one business.

The classic trap: a decision has positive EV because the upside is large, but the downside scenario, even if unlikely, would end the company. Consider a 70% chance of generating $200,000 and a 30% chance of losing $400,000. The EV is: (0.70 × $200,000) + (0.30 × -$400,000) = $140,000 - $120,000 = +$20,000. Positive. But if losing $400,000 kills the business, the positive EV is irrelevant. You should not take that bet.

The fix is to separate EV analysis from ruin analysis. Before you run the numbers, ask: "Is any outcome here an existential threat?" If yes, that outcome becomes a constraint that overrides EV. Reversible vs irreversible decisions covers this distinction in depth. Decisions where the downside is recoverable deserve more risk-taking than decisions where it is not.

A related mistake: using a negative EV to park a decision permanently without asking what would need to change. The number is a tool for inquiry, not a verdict.

A Simple EV Table You Can Copy

Use this structure for any decision with two options and up to four scenarios each.

Option A has a higher EV despite a smaller best-case upside. That is the kind of insight that disappears when you focus only on the headline number.

Fill this out collaboratively before a major commitment. Disagreements about probability estimates surface early and make the real debate explicit, before you are arguing about conclusions rather than assumptions. For a process to track these estimates against actual outcomes, the decision log template is a practical starting point.

When EV Thinking Is Worth Your Time

EV thinking earns its keep when:

• The decision involves material financial stakes (roughly $5,000 or more in committed spend or opportunity cost)
• You have at least some base rate to calibrate probability estimates against
• Multiple people need to align on assumptions before resources are committed
• You expect to face similar decisions again, so you can compare estimates to outcomes

It is less useful when:

• The decision is truly one-off and no base rate exists (scenario planning and downside analysis matter more here)
• The right call is determined by values or ethics, not outcomes
• The cost of the analysis exceeds the cost of being wrong on the decision itself

Most growth and investment decisions that founders and team leads face fall into the first category. The tool applies broadly, even when the inputs are rough.

Key Takeaways

• Expected value is probability multiplied by outcome, summed across scenarios. You do not need software; you need explicit assumptions.
• The main value of EV thinking is not the final number. It is the process of naming outcomes, estimating probabilities, and identifying the single assumption that drives the result.
• Never use positive EV to justify a bet with a ruin scenario. The framework only applies when you survive to run the experiment again.
• The assumption with the most weight in your EV calculation is where to focus information-gathering before you commit, not after.
• Run EV calculations collaboratively when alignment matters. Disagreements about probabilities are easier to resolve than disagreements about conclusions.
• Track estimates against outcomes over time. That feedback loop is what turns a one-off calculation into a reliable decision-making habit.