Hope Is Not a Strategy: How to Stop Guessing and Start Forecasting Like a Pro.

Hope Is Not a Strategy: How to Stop Guessing and Start Forecasting Like a Pro.

Discover how top-performing businesses remove bias from planning.

Introduction: Why We Get Forecasting So Wrong (And How to Fix It)

If you’ve ever written a business plan, you’ve probably fallen into the same trap I have: confidently predicting how much revenue you’ll make next year, how long a project will take, or how many customers you’ll have in six months. You plug in the numbers, build a tidy forecast, and think, “This looks solid.” But then real life happens, and the plan doesn’t survive first contact with the market.

I’ve seen this play out repeatedly, not just in small businesses, but on a massive scale. Look at the Sydney Opera House. It was supposed to be finished in four years at a cost of $7 million (Australian Dollars). It ended up taking 14 years and blowing out to $102 million. The UK’s HS2 rail project, originally budgeted at £37 billion, is now projected to exceed £100 billion, with huge delays and massive public backlash.

Even in the world of tech startups, it’s the same story. I once worked with a SaaS founder who was certain they’d hit £1 million in ARR in 18 months. Two years later, they were still circling £400K — not because the product was bad, but because the growth assumptions were pure optimism, not grounded in data.

We all want to believe our projects will be on time and on budget. But here’s the truth: most forecasts are wildly optimistic. And that optimism can be dangerous. It leads to overcommitment, under-resourcing, and missed expectations that can wreck both morale and cash flow.

So what’s the fix?

In this blog, I want to introduce you to a smarter way to forecast — a method called Reference Class Forecasting, developed by Bent Flyvbjerg. It’s not magic. It’s not guesswork. It’s about looking outward — at what’s happened in the real world — instead of inward at your best-case scenario.

If you’re serious about planning, scaling, and managing risk like a pro, this concept could change the way you run your business. Let’s dive in.

1: The Forecasting Trap in Business.

I’ve worked with dozens of business owners over the years, and one thing they all do — without fail — is create forecasts that are more fiction than fact. I’ve done it too. We sit down to plan next quarter, next year, or a new product launch, and the spreadsheet starts filling up with best-case numbers: “We’ll close 30% of leads.” “We’ll onboard that hire in two weeks.” “This campaign will return £10K in new business.”

It feels good. It looks impressive. But it’s rarely how things actually unfold.

This is the forecasting trap — the belief that you can predict the future with confidence just because you’ve laid out a nice plan. But forecasting isn’t about certainty. It’s about managing uncertainty. And most people forget that.

I once worked with a consultancy that forecasted a 50% increase in recurring revenue after hiring two new team members. The logic seemed sound. But they underestimated onboarding time, overestimated sales velocity, and missed how long it would take those hires to become revenue-positive. Six months later, their margins were down, not up — all because the forecast didn’t reflect reality.

Another client forecasted a product launch for Q1. Everything looked great on paper: budget, tasks, timelines. But manufacturing delays, content issues, and a contractor who dropped out pushed them back two full quarters. Their original forecast assumed a perfect run. But no plan survives contact with the real world — especially not when built on hope.

The truth is, most forecasts reflect our ambitions, not our constraints. We want things to go right. We assume we’ll hit deadlines. We forget how often things go sideways.

That’s not a reason to avoid forecasting. It’s a reason to do it better. The answer isn’t to give up — it’s to shift the way we forecast entirely.

That’s where Reference Class Forecasting comes in. Before we get to that, though, let’s look at some of the biggest and most expensive forecasting failures in history — because they show how even experienced teams with massive budgets fall into the same traps we do in small businesses.

2: Famous Forecast Failures (Real-World Examples).

Forecasting failures are not the exception — they’re the rule. And it’s not just small business owners who get it wrong. Some of the most expensive and well-resourced projects in the world have been crippled by overconfidence, flawed assumptions, and sheer wishful thinking.

Take HS2, the UK’s high-speed rail project. It was originally announced with a budget of £37.5 billion and a promise of completion in just over a decade. As of today, costs have soared past £100 billion, timelines have stretched indefinitely, and large portions of the route have been scaled back or cancelled altogether. The project has become a case study in political overreach and forecasting failure. See my previous blog on Action Bias for more on this.

Or look at the iconic Sydney Opera House. The plan? Complete it in 4 years for $7 million. The reality? 14 years and a staggering $102 million. That’s a 1,400% cost overrun. The design was revolutionary, but the planning was disconnected from the practicalities of construction. The forecast wasn’t a plan — it was a hope.

Even modern examples haven’t fared better. The California High-Speed Rail project, pitched in 2008 at $33 billion with a 2020 completion date, is still under construction, now pegged at over $100 billion and counting. Every year, the gap between plan and reality widens because the forecast was built on ambition, not evidence.

In business, I’ve seen similar patterns repeat at a smaller scale. Founders planning to build an MVP in three months — it takes nine. Agencies forecasting smooth six-figure launches — but underestimating lead time, resource constraints, or market readiness.

These failures have one thing in common: they relied on inside views — isolated projections based on internal assumptions. No one stopped to ask: 

“What actually happened in similar situations before?”

This is where Reference Class Forecasting shines. It forces you to look at the real track record of similar projects, so you stop fooling yourself — and start building forecasts rooted in reality.

3: Why We Get It Wrong – The Psychology of Forecasting.

“The biggest forecasting problem isn’t your spreadsheet — it’s your brain.”

We’re wired for optimism. It’s part of what makes us entrepreneurial in the first place. If we didn’t believe we could pull off bold ideas, we’d never start anything. But that same optimism also blinds us when we’re planning the future.

Psychologists call this the planning fallacy — a term coined by Daniel Kahneman and Amos Tversky. It means we consistently underestimate how long things will take, how much they’ll cost, or how difficult they’ll be. Even when we’ve done something before, we assume this time will be different. Better. Faster.

I’ve sat with small business owners mapping out a new product launch. Despite delays on the last three projects, they’ll confidently tell me, “We’ll have it ready in four weeks.” They’re not lying. They believe it. But it’s not based on history — it’s based on hope.

Another common trap is anchoring. Once we come up with a number — say, “We’ll sell 100 units in month one” — it becomes our benchmark, even if there’s no evidence to support it. We get emotionally attached to that figure, and when reality doesn’t match, we either panic or rationalise the shortfall.

Then there’s confirmation bias. We selectively look for data that supports our forecast and ignore signs that contradict it. If three customers express interest, we treat it like a trend. If five say “not now,” we call them outliers.

And let’s not forget overconfidence. Business owners — myself included — often think we’re the exception. “Sure, that other guy struggled to launch on time, but we’ve got a better team.” Or, “Yeah, the market is tough, but our offer’s different.” Sometimes that’s true. Most times, it’s not.

The problem isn’t that we’re wrong. It’s that we’re predictably wrong — in the same ways, over and over again.

This is exactly why traditional forecasting methods often fall apart. They’re built on flawed assumptions and emotional bias rather than grounded data. Which leads us to the next step: how do we fix it?

The answer isn’t to throw out forecasts entirely. It’s to shift the foundation from subjective guesswork to objective precedent. That’s where Reference Class Forecasting makes its entrance.

4: Introducing Reference Class Forecasting.

When I first came across the concept of Reference Class Forecasting, I had one of those “why didn’t I learn this years ago?” moments. It’s simple, powerful, and — once you see it — hard to ignore.

Developed by Bent Flyvbjerg, a Danish economic geographer and expert in megaproject planning, Reference Class Forecasting (RCF) is a way to de-bias your forecasts by grounding them in reality, not guesswork.

Here’s how it works.

Instead of starting with your project and asking, “How long will this take?” or “How much will this cost?” — you start by asking a better question:

 “What happened with similar projects in the past?”

You define a reference class — a group of completed projects or businesses similar in scope, size, and complexity to yours. Then, you look at how those performed. What was the actual timescale? What were the costs? How close were their original forecasts to reality?

Only then do you place your own project in that distribution. It’s called taking the “outside view” — and it forces you to confront reality before optimism.

I started using this approach when helping business owners plan new initiatives. For example, a client wanted to launch an online course. Instead of asking how long they thought it would take, we gathered data from 10 other course creators in the same niche. The average build time? 4.5 months. Not the 6 weeks he was planning for. That insight alone reshaped his entire launch strategy — and saved him from blowing his budget.

RCF is especially useful in financial forecasting, too. Instead of creating revenue targets from thin air, you look at what similar businesses achieved in their first 12 months. How fast did they grow? What were the drop-off rates? What churn did they face?

And here’s the magic: RCF reduces your risk of being wrong in predictable ways. You’re no longer betting your future on rosy assumptions. You’re building your forecast on solid, historical evidence.

It’s not perfect. But it’s more accurate — and far more honest — than 99% of business plans I see.

5: How to Use Reference Class Forecasting in Business Planning.

When I first started applying Reference Class Forecasting (RCF) in my business planning work, the impact was immediate. No more spreadsheets built on fantasy. No more “we’ll just figure it out.” It gave my clients — and myself — a concrete way to anchor plans in reality, not wishful thinking.

So, how do you actually use it?

Let me walk you through it step by step.

Step 1: Define the Project or Forecast.

Be specific. Are you planning to launch a new product? Hire your first team member? Scale your revenue to £500K? The more clearly defined the objective, the easier it is to forecast accurately.

Let’s say you’re building a SaaS product and want to forecast development time and cost. That’s your project.

Step 2: Identify the Reference Class.

Find 5–15 real examples of similar projects. The key here is similarity. Look for businesses or founders with similar resources, skill sets, industries, and ambitions. If you’re building a course, don’t benchmark against Apple. Benchmark against someone who launched a similar course on a similar platform.

Sources? Look at:

• Industry benchmarks and case studies.
• Conversations in mastermind groups or forums.
• Data from peers, competitors, or prior clients.
• Your own past projects (if applicable).

This part takes effort, but it’s where the value is created.

Step 3: Analyse the Distribution of Outcomes.

Here’s the reality check. How long did it actually take them? What did it actually cost? What % hit their revenue targets? What % failed?

When I did this with a client launching a new service offer, we found that the average conversion rate across 10 similar businesses was under 3%, much lower than the 10% they had assumed. That completely reframed their expectations and pricing model.

Step 4: Position Your Project Within That Distribution.

Now ask: Where do you fall on that curve?

Are you more experienced than the average? Less funded? Have better systems? Worse? Be honest. If the median result was 4 months and £15K to launch, and you’re newer or less resourced, assume a worse-than-average outcome.

This step strips away ego and makes the plan credible.

Step 5: Adjust Your Forecasts and Build Buffers.

Now, you revise your forecasts, grounded in the reference class data. You add contingencies, buffers, and fallback plans.

If your original forecast was £10K in sales in Month 1, but the reference class shows £3K is typical, adjust accordingly. Or keep the £10K as a stretch goal — but base operations, cash flow, and marketing decisions around the more realistic figure.

Using RCF won’t make your plan perfect. But it will make it robust. You’ll spend less time explaining missed targets and more time hitting achievable ones.

6: What Henry Mintzberg Got Right About Planning.

Before we dive into how the 365/90 Planning System complements forecasting, it’s worth pausing to acknowledge one of the most insightful voices in strategic planning: Henry Mintzberg.

When I first read Mintzberg’s work — particularly his critique of rigid, over-formalised planning systems — it was like someone had switched on the lights. He wasn’t anti-planning. He was anti-false certainty. And that resonated with me, especially after seeing how often entrepreneurs mistake a detailed spreadsheet for a strategy.

Mintzberg argued that most traditional planning approaches assume the world is stable and predictable. But as small business owners, we know the opposite is true. Customers change their minds. Supply chains break. Competitors come out of nowhere. Your brilliant idea turns out to be less brilliant in the wild.

Mintzberg warned that planning often fails when it becomes detached from doing, when forecasters are not the people implementing the work. He described this as the fallacy of detachment — and it’s something I see all the time. An owner creates a five-year forecast, but the team on the ground is firefighting month to month. There’s a complete disconnect between the plan and the real world.

He also criticised the illusion of control — the idea that the more detailed your plan, the more in control you are. In reality, that level of control rarely exists, especially in fast-moving markets or lean businesses.

What Mintzberg advocated instead was something much closer to what I use with my clients today: adaptive planning. Planning that recognises uncertainty, builds in flexibility, and keeps revisiting assumptions.

This is exactly what the 365/90 Planning System was built for.

It acknowledges that you need a longer-term vision — your 365-day outcome — but you also need 90-day execution cycles to test, learn, and adjust. It’s not planning once and forgetting about it. It’s planning as an ongoing discipline.

You’ll still forecast — but you’ll do it with more humility and more realism. And that’s the kind of planning Mintzberg believed in. One that’s alive, responsive, and tied directly to action.

7: Forecasting the 365/90 Way.

When I created the 365/90 Planning System, it wasn’t just about breaking big goals into smaller chunks. It was about solving a deeper problem: how do you plan and forecast in a way that reflects the messy, unpredictable nature of real business?

That’s where reference class forecasting and Mintzberg’s adaptive planning philosophy fit perfectly into 365/90.

Let me walk you through it.

Start with a 365-Day Outcome

Your first step is to set a clear, measurable annual goal. Not a fantasy number, not a vague ambition — a target you can build towards. But here’s the shift: before committing, you run your forecast through the reference class lens.

Ask: What happened when others in my industry tried something similar?
What were the real-world results for businesses of my size and stage?

This grounds your big goal in context, not just optimism.

Break it down into 90-Day Sprints.

Once the 365 goal is sanity-checked, you plan quarterly cycles of action and learning. These are tight, focused bursts of execution where you:

• Test assumptions.
• Track actual results.
• Compare forecasted vs. real outcomes.
• Adjust direction if needed.

At the end of each 90 days, you review performance against the forecast. Did you overestimate lead flow? Did a marketing campaign flop? Did delivery take longer than expected?

You don’t beat yourself up — you use that data to build a smarter next 90-day plan.

Review and Revise Forecasts Quarterly.

Every 90 days, your forecast gets sharper. Not because you have a crystal ball, but because you’re learning from real data.

You adapt. You tweak. You get better at predicting because you’re closer to the action.

Over time, your business becomes anti-fragile (read my blog on Antifragile here) — stronger, not weaker, when reality deviates from the plan. That’s the real value of 365/90: it turns forecasting from a guessing game into an evolving system.

Final Word: Forecasting With Realism, Not Illusion.

We all want to believe in our vision. That’s what makes us entrepreneurs. But belief alone doesn’t pay the bills, win the clients, or hit the targets.

Forecasting — real, useful forecasting — starts when you stop trying to predict the future like a magician and start acting like a strategist.

If there’s one thing I’ve learned over the years, it’s this: you’ll never be 100% right, but you can be 100% better informed.

When you use tools like Reference Class Forecasting, you’re borrowing wisdom from others who’ve already walked the path. And when you pair that with adaptive planning frameworks like 365/90, you’re building a system that helps you react fast, fix faster, and grow smarter.

That’s not guesswork. That’s strategic execution.

And if your forecasts haven’t worked in the past, maybe it’s not that you’re bad at planning — maybe it’s that you’ve been using the wrong kind of plan.