Quick Answer: The ability to learn quickly depends on shortening your entire iteration cycle — not just build time, but also data collection and decision-making. As product managers, we need to identify which specific bottleneck (production, marketing, or decisions) is the constraint and fix that first. Adding more marketing won’t help if you can only deploy weekly, and faster deployment is wasted if you only have 10 daily visitors. If no information is collected at all, your iteration length is effectively infinite.

Concept: Iteration Cycle - ability to learn quickly Skill Goal: Decrease the time required to learn. “The act of repeating a process, usually with the aim of approaching a desired goal.” - Wikipedia

• The goal of a startup is to discover a successful business model.
• Iterations can refer to small learning exercises, but is more often used to describe the amount of time required to learn something significant about the business model or ability to lean quickly.
• The iteration length is therefore not the amount of time to build a new version of the product, but also the amount of time required to gather significant information as well as the amount of time required to make a decision.
• If no information is collected to base a decision on, then the iteration length is infinite because nothing is learned.
• Companies can decrease the iteration length in a variety of ways:
  • Use continuous deployment with automated testing to increase production.
  • Reduce product requirements to create Minimum Viable Products.
  • Increase the amount of customers sampled for each test.
  • Simplify the decision making process.
  • Focus on qualitative data instead of quantitative data.
  • Etc.

Example: Production Bottleneck A team has 10,000 visitors each day, but can only iterate once per week. Additional marketing may increase sales, but will not increase learning. Example: Marketing Bottleneck A company with a 5 person development team working on a simple survey application and practicing continuous deployment has demonstrated and ability to iterate every day. However, there are only 10 people visiting the website each day. A sample size is 150 people, so the team has 14 days of wasted development per cycle. The team needs to focus on marketing. Additional readings: Cash is not king Iteration Time = Time to Learn, Not Time to Build Individual Exercises:

• A company XYZ wishes to optimize it’s on-line sales. It has two new home page designs it can test. It takes the engineering team three days to code, test, and deploy a new website home page. The website currently has 150 visitors each day and the CEO has decided that 150 visitors is a sufficient sample size of information to make a decision.
• • What is the iteration length?
  • What is the bottleneck?
  • How would you reduce the iteration length?
• Determine your average iteration length.
• • What are the bottlenecks?
  • How can you reduce iteration time?

Thought / Discussion Exercises:

• How does iteration length differ in larger established companies? Is this good or bad?

Frequently Asked Questions

What does “ability to learn quickly” mean for startups?

The ability to learn quickly refers to how fast a startup can complete full iteration cycles — not just building a product, but also gathering significant information and making decisions based on that data. As product managers, we need to optimize the entire loop because if no information is collected, the iteration length is effectively infinite and nothing is learned.

How do you calculate iteration length in lean startup?

Iteration length isn’t just the time to build a new product version. It includes the time to build, the time to gather a significant sample of data, and the time to make a decision. For example, if coding takes 3 days but you need 150 visitors at 150 per day, the bottleneck is the 3-day build — making the full iteration length 4 days (3 to build + 1 to collect data).

What are common bottlenecks that slow down startup iteration cycles?

There are three main bottleneck types: production bottlenecks (slow development or deployment), marketing bottlenecks (not enough customers to gather data quickly), and decision-making bottlenecks (slow or overly complex analysis). For instance, a team with 10,000 daily visitors but weekly deploys has a production bottleneck, while a team practicing continuous deployment with only 10 daily visitors has a marketing bottleneck.

How can startups decrease iteration time?

We can reduce iteration time by using continuous deployment with automated testing, reducing product scope to create Minimum Viable Products, increasing the number of customers sampled per test, simplifying decision-making processes, and focusing on qualitative data over quantitative data. The key is identifying which specific bottleneck — production, marketing, or decision-making — is the constraint and addressing that first.

Why is more marketing not always the answer to faster learning?

If your bottleneck is in production — say you can only deploy once per week despite having 10,000 daily visitors — adding more marketing increases sales but does nothing to increase learning speed. As product managers, we need to diagnose where the actual constraint lies before investing resources, because optimizing the wrong part of the cycle won’t shorten iteration length at all.

Comments

Loading comments…

Leave a comment

Name *

Email * (not published)

Website (optional)

Comment *

Leave this blank

Post Comment