Bite size design operations | Design estimation
The Subtle Art of Estimating the Duration of a Design Project
A counterintuitive approach to mastering the effort estimation as part of the cognitive load-based design sizing model.
This post aims to demystify the duration estimation process especially for design projects. It introduces a technique called ‘middle-out’ that helps designers provide project estimations based primarily on the level of understanding of the project.
Estimation is my favorite color and why it should be yours too!
- Estimating a design project is a difficult task because it involves layers and steps that depend on many variables. Given the non-iterative nature of design, the more the complexity of a project, the harder it gets to make an accurate estimation.
- Most of the time, many design project estimates are off because the design teams don’t get the full picture of the project; missing briefs and a lack of understanding of the user problem are a few to mention.
- Having said that, getting the estimation right is really important as it’s impossible to get aligned with and influence the engineering and product timelines without accurate estimations.
- There are two common estimation techniques out in the wild, looking at you Scrum, when estimating project timelines:
- Top-down technique is a timeline-driven approach. It’s about allocating a fixed duration to a project (depending on product launch dates or engineering expectations) and then dividing it into manageable chunks. This approach works well when there’s clear insights into the product (project brief, user insights, etc).
- Bottom-up technique, on the other hand, is a task-driven approach. It’s about having full visibility into what goes into each type of design project. You’ll then estimate how long each task will take and roll them up into the overall project timeline.
- Given the agile world we are in and ever-changing conditions and user needs, using the top-down approach would be a mistake.
- On the other hand, bottom-up technique works well when we can’t plan for the future. However, it’s mostly inefficient and brings more chaos to our processes.
- The Middle-out is an alternative technique to the traditional approaches. It’s more effective, efficient and better suits our needs.
The Middle-out technique
The middle-out technique heavily relies on identifying the high-level milestones that are driven by the strategic product priorities. For the design project, the most common milestones are Discover, Design, Review, Crit, and Iteration. Keep in mind that designers can break down these milestones into smaller pieces, however, the point of this approach is to keep it high level.
There are two more elements that go into the middle-out technique: T-shirt sizes and Fibonacci sequence. T-shirt sizes help articulate the effort to deliver, whereas the Fibonacci sequence makes the effort to deliver estimation much more tangible by mapping them to comparable values.
Here’s an example that helps you better understand the middle-out technique:
Step 1: You have a design project and you identified the high-level milestones as
Step 2: The next step is to decide on how much effort goes into each milestone. We’ll tackle this by roughly estimating the t-shirt sizes
Step 3: Now remember Table 1 where we map t-shirt sizes to Fibonacci numbers. It’s time to put it into use
Step 4: The next step is to estimate the time it goes into one of the high-level milestones that you feel confident about. Let’s say you have a pretty good sense of the scope of the project and you can comfortably say you’ll spend 2 weeks drawing wireframes.
Step 5: Using the weeks vs Fibonacci ratio (2 weeks vs. 8), we can fill out the rest of the table
Step 6: We can now roll each effort up to the project timeline and estimate it as roughly 10 weeks. We can communicate this to the engineering team as ‘8 weeks’ as we estimated 2 weeks of engineering support.
This method, like any other methods, never estimates with 100% accuracy, unless you’re in a controlled environment. Having said that, it takes into consideration your previous experience and seniority. It doesn’t require an upfront commitment and it’s transparent.
Failure is the way forward
Keep in mind that you’re going to fail at estimating the durations of your project, multiple times. This is from my fav book: “Improvement at anything is based on thousands of tiny failures, and the magnitude of your success is based on how many times you’ve failed at something”. However, as you get more involved in this model, make countless estimations and gain experience, your outputs will get much closer to the actuals. The key is to keep a record of what you estimated and where you did end up. Talk to your design manager or design program manager to see how you can get better at keeping your tracker up to date. When you communicate your estimations with your partners, don’t be hesitant to share your thought process; whether it is the middle-out technique mentioned here or something else. This will give you peace of mind!
Net Step: Design Sizing Model
Duration of the project cannot be translated into capacity and workload. Multiple attributes should be considered to measure the size of your project in addition to the effort.
Before you embark on your estimation journey and start a new project, make sure you know the Priority of the project, the Type of the project and the Level of XFN Collaboration required. These three project attributes represent the complexity of the project in the cognitive load-based design sizing model. Using the complexity, you will take the full advantage of the model and it will eventually help you make data-driven estimations using machine learning.
Wouldn’t it be cool if you get a message from your favorite tool that reads “Based on the complexity of the project, it will take [X] weeks to deliver”?
Shout out to my favorite book The Subtle Art of Not Giving a F*ck and of course Mark Menson
