Sir Tom Dyson iterated 5,126 times on the vacuum cleaner design that would ultimately make him a billionaire. Thomas Edison famously said “I have not failed. I’ve just found 10,000 ways that won’t work.” These are daunting numbers for any product team or CFO trying to determine how many iterations a product will need to achieve the results they want.
The more iterations you can work through the more chances you have of succeeding. Assuming you can output something is just hoping for luck or planning for failure.
The reality is that endless iteration is hard to come by, no matter how big the company or the bank balance, so you need some sense of the number of iterations before you start to get your idea, feature, product, team or company funded appropriately.
This post aims to help to answer the question of how many iterations you need to make your product succeed. And how to plan for those iterations around budget, timeline and iteration length.
Funnily enough, the answer isn’t 42.
Let’s walk through more detail about how to determine the right number of iterations for you.
Definitely more than 1
You can be 99.98% certain that you will need more than 1 iteration. It is rare (if not impossible) to bring a product or major feature to market and have a perfect reception that requires no major changes.
This means we can immediately conclude that we need N (the number of iterations here) to be greater than 1.
So, now we know it needs to be greater than one. But how much greater?
Calculating iterations required for your product
When it comes to determining how many iterations a product needs there are three factors that you need to consider. These three factors can quickly provide boundaries for working through the question, particularly because one of them is usually fixed.
Iteration Factor: Budget
You may already have a budget allocated. If you’re building a product in a company, this may be the money remaining in the budget of your project or department. If you’re a startup it is cash in the bank. It may be the number of days effort available in your team over a certain time period before they get shifted to another priority.
Iteration Factor: Timeline
You may already have a timeline that you need to hit. The Google Docs, Sheets and Slides team knew that they needed to be in market before Office 365 launched. This set their timeline. You may have a fixed launch date due to an event.
Iteration Factor: Iteration Length
The nature of the product or feature may determine the iteration length. There may be something inherently complex that just takes time to work through.
Iteration length needs to be measured in actual calendar days from start to finish.
If you take 2 days redesigning a web page but then it sits in a review workflow for 5 days before it is shown to people and another 4 days for you to collect adequate feedback from users then your total iteration length is 11 days. I’m calling this out because often people would say “oh it took us 2 days”.
For digital products this might be relatively short, for products depending on relationships between companies this might be longer, for products involved in seasonal activities it may be longer.
To calculate your iteration length, you need to look at how long it will take you to test your product and get real feedback. This comes in different shapes and varieties but you really want to focus on feedback that involves people paying so plan your iteration length around how quickly you can get that type of feedback.
Some tips and considerations for calculating iteration length:
- How many stakeholders are involved? Generally, the more stakeholders the longer the iteration length needed. Can you cut some of them out somehow?
- How can you do it differently? Here is an example of how a team used photoshop to shorten their iteration length instead of coding up the site. Maybe you could use off the shelf products or software (like Excel!) to prototype what you need.
- What dependencies are there? More dependencies, like stakeholders, usually means longer iterations. Can you reduce these?
- Production timelines – some things take time to manufacture or produce.
- Seasonality and Events – some things only happen at certain times so you need to wait.
- Regulations – you may need to abide by regulations or regulatory processes that increase iteration length (e.g. medical devices).
Calculating iterations left
There are situations where you can calculate the number of iterations you have left instead of how many you need. Here are some quick formulas to help you:
Iterations Remaining = Time Remaining / Iteration Length
Iterations Remaining = Budget Remaining / Iteration Cost
Perspectives from others on iterations needed
Some more thoughts
- Marty Cagan advises “I tell founders that if you want to have any hope of solving problems and build the right products, you’re going to need to have at least 50 to 100 iteration attempts. And there is just no way to do that by using your engineers to build four-month MVPs. If you have your engineers build a four-month MVP and deploy it, and it doesn’t actually do what you want, you have just wasted four months out of an 18 month runway.”
- Dan Olsen, Group PM at Intuit says prioritize iterations based on metrics you want to drive versus required resources.
Quick Rules for Iteration
Hopefully the explanation above has shown you that working out the number of iterations you need or have left isn’t a straight forward exercise and is also a useful exercise in thinking through how to approach your problem or opportunity.
There are some general rules you can follow though which are:
- Try to fix one of the factors affecting the number of iterations you need/have left. E.g. fix budget, timeline or iteration length required. Then work from there.
- Follow Marty Cagan’s advice and plan to ensure you can complete 50 to 100 iteration attempts.
- Whatever you do, don’t do 1-2 iterations and wonder why it didn’t work.