How to Estimate More Accurately

Estimation accuracy is a crucial skill for startups and entrepreneurs. However, this is a vast area in which the majority of individuals suffer. Consider how many estimates you’ll need on a daily basis as a startup founder:

  • When will the next edition of the product be released?
  • What are our sales predictions for the coming quarters and years?
  • When and how many people should we hire?
  • What is the most appropriate amount to raise?

The number of items to estimate is nearly endless. And you’ll be way off on a lot of things if you can’t estimate accurately. And, surprise, you’ll almost always UNDERestimate (be overconfident) instead of OVERestimate. Improper estimation can have severe consequences for you, your startup, and your reputation!

So, what’s the solution? How can you drastically improve your ability to deliver more accurate estimates on a regular basis?

Allow me to introduce you to our mutual friend:

CBE stands for Confidence-Based Estimating. It’s a simple strategy that’s simple to implement.

When using confidence-based estimating, I want you to get into the habit of thinking about the confidence level (represented as a percentage) of your estimate anytime you create one.

Let me give you a specific example of what I mean by “confidence level” and the psychological impact before I go any further. So, let’s talk about something else: the weather!

You’d assume “okay, it’s going to rain” if I told you my forecast for tomorrow’s weather (let’s call it a “weather forecast”) was that it would rain. However, if I told you, “Hey, there’s a 50% chance of rain tomorrow (also known as the probability of precipitation or “POP”),” you could have a different opinion of my predictions. And if I told you that there was a 25-30% chance of rain tomorrow, well, that’s a whole other kettle of fish. I simply went from “it’s going to rain tomorrow” to “there’s a 25-30% chance of rain,” and you’d definitely leave your house without an umbrella or rain jacket now. So you heard not just my predictions but also my level of confidence, both of which have a big influence on how you think and act.

Let’s get back to talking about startups from talking about the weather.

Here’s an experiment I’d want you to try. Ask your product engineering team, “What is your level of confidence, on a scale of 0-100 percent, that you’ll fully deliver that functionality in two weeks?” the next time they give you an estimate for when something will be done (e.g., we anticipate the next release will be ready in two weeks). It’s as if you’re putting your entire salary on the line if the percentage is 95 percent or greater.

Wait for a brief pause… and then you’ll most likely hear a confidence level of 60-80 percent. Take a deep breath… then respond with “okay.” So, given a 95% confidence level, what is your forecast?” and then wait for the response. But hold your breath. Prepare for the next question to increase by 50-100 percent!

There’s also a potential that someone will remark, “I can’t give you any more confidence.” I can’t say I’m certain.” That’s OK. Just remember to take their estimate with a grain of salt because there’s a good probability it won’t be correct.

Consider the following… When you asked “when will it be done?” you meant “done.” You didn’t ask for an estimate with a confidence level of 70 or 80 percent. When you ask “for realsies,” you may now safely ask, “When will it be done?” Normally, you’ll be given a significantly higher estimate.

I’d like you to take a time to consider this >> Have you ever observed that most things take twice as long and cost twice as much as they should, and no one seems to notice? Do you know why? Because the majority of individuals are terrible at estimating and no one forces them to think critically about their estimates. So, should anyone be surprised at all?

Okay… So, let’s keep this CBE thinking going and add another layer of complexity to our estimate process.

Follow up the “at a 95% level of confidence” query with another: “what are the primary assumptions you’re using for your new estimate?” when you ask for more accurate estimates with higher degrees of confidence.

Let’s have a look at an example. You can hear assumptions like “well, with a 95% level of confidence, our top assumptions that we can get this done in 2 weeks are…” in a startup when you’re building cloud software.

  • “The new outsourced design team consistently meets deadlines.”
  • “With the addition of two new people to the testing team, the tests are completed on time.
  • “Plus, the development team won’t have to deal with any complications, allowing them to focus exclusively on the upcoming release.”

Take a look at your assumptions and think about them. Examine whether or not they are feasible. Each assumption puts the timeline in jeopardy. If you suspect assumptions are incorrect and could affect the accuracy of the estimate, don’t be afraid to dispute them. According to the preceding, YOU should anticipate troubles, and the development team may need to commit up to 25% of their time to emergency solutions. As a result, you should discuss that assumption, determine the best way to handle it, and understand the estimate’s realistic impact.

I’d like to go over some of the greatest places to employ the CBE technique before I add another stage to the process. Consider the following questions to see how you may improve your estimate skills:

  • When will the next version be released?
  • What sales forecasts do we have for the next month, quarter, or year?
  • When should we hire people and how many should we hire?
  • What is the best amount of money to raise?
  • When is that customer going to make a purchase?
  • When will we achieve our objectives?

Every single one of these is an estimate where adopting a CBE strategy will pay off handsomely.

A few further adjustments to the method and thoughts are now necessary:

Estimate “in the small” rather than “in the large” before making a decision. Instead of estimating the entire project, break it down into smaller chunks and “guess in the little.”

When you “estimate in the micro,” you could notice more detailed assumptions that you would have missed if you were only thinking about the big picture.

Don’t be tempted to think “that can’t be right” when you pull the small estimations together and roll them up. It can’t possibly take that long/cost that much!” Keep the number in mind.

Try making team estimates next, where each team member produces their own estimate. After that, gather everyone to discuss each estimate, including the level of confidence and important assumptions. This manner, you’ll be able to triangulate and get a more precise, nuanced estimate. By the way, when the team comes together, be cautious. People’s assessments and views will frequently disagree, so don’t let them “bully” others about theirs. Maintain a calm demeanour. Maintain a critical frame of mind. Aim for precision as well.

It’s as simple as that! This is the essence of confidence-based estimation…

  • When you get an estimate, inquire about the confidence level and insist on an estimate with a 95 percent confidence level.
  • Ask what top assumptions are being made when you acquire estimates with increased levels of confidence.
  • Make estimates at the smallest scale possible; remember, estimate in the tiny, not the large.
  • If you’re working with a group, have each team member complete their own estimations before gathering everyone together to discuss.

If you do that, your estimating accuracy will skyrocket!

Finally, a word on more precise estimates. You’ll appear to be a rockstar and have more credibility in the eyes of investors, not simply because your efforts and company will function more efficiently. You’ll gain a reputation as a business owner who does what they say and when they say it.