What to Wear Daily

Our story, or how I went from identifying a problem to creating a product

The initial idea

The idea for What to Wear came from a simple observation I made several years ago: people tend to check the weather to figure out what to wear. I am not the first person to make this observation and I won't be the last. I'm not sure what percentage of weather-checks are for this reason, but it's a relatively high percentage...I would bet well over 50%.

Not really a big deal, right? But what is interesting is that this simple observation leads to other observations ...one of them is that most weather services are poorly designed for what people actually use them for. Really poorly designed! To show you what I mean, here is a simple critique of the Weather.com homepage.

And thus all of the work of figuring out what to actually wear is left to the person checking the weather…no weather service that I know of answers the simple question of what to wear for you. So I thought…here is an opportunity to solve a very real problem that many people have.

The job-to-be-done

I also realized that this was a great example of a “job to be done”. The JTBD framework is a product marketing framework invented by Clayton Christensen (who wrote the seminal book The Innovator's Dilemma) and is helpful when thinking about why people do what they do…and helps product folks focus on a real problem they are trying to solve. From the JTBD website:

”The jobs-to-be-done framework is a tool for evaluating the circumstances that arise in customers’ lives. Customers rarely make buying decisions around what the “average” customer in their category may do — but they often buy things because they find themselves with a problem they would like to solve."

This intrigued me…the job people had to do was to figure out what to wear that day and the way they decided was to first look at the weather. There are other factors like what clothes are clean and what mood someone is in but weather is the starting point. I wondered…is there a way to test the validity of the JTBD framework and build a service that simply answers this question for people?


Now that I had identified the job-to-be-done I had to figure out how to design something to test it in the simplest possible way. One way of describing this is the minimum viable product, or MVP: "The minimum viable product is that version of a new product which allows a team to collect the maximum amount of validated learning about customers with the least effort." I first looked around to see if someone was already trying to solve the problem. If so, how were they doing it? The answer was yes…many people have made the same observation and there existed several apps/services trying to solve the problem.

One of the more popular apps trying to solve this problem is Swackett. However, after using Swackett and several others for a week or so I concluded that although they started with the same initial observation they were not testing the idea as clearly as I wanted to...they had all basically built services to sell clothes. This is where product design gets tricky...when looking at competing services you need to be able to figure out if their design is actually working the way it is intended to. My hunch was that Swackett was not very effective in becoming was verified after talking to a few people who had used Swackett...none of them used it after a few days. They found it annoying and not useful.

That's one of the most difficult things to figure out...how can you really test an idea? So I decided to test it without trying to make money from selling clothes, which everyone else seemed to be doing. This would allow me to focus on the recommendations themselves. That felt like a good test...can I create a set of recommendations of what to wear that people actually used as part of their daily routine?

After working with a friend on one for a few weeks I also decided not to make an iPhone app as it was a relatively big investment for an untested idea.

Just an email to friends

I settled on a simple product: a daily email. So in June 2013 I started sending out an email to several friends who worked at HubSpot with me. I started sending to mostly women initially because that was who my initial conversations started with, but I also had talked to enough men and women to know that women, in general, tend to have more sophisticated clothing strategies than men.

Without thinking too much about it I had become an email-first startup, an idea I first heard about from Ryan Hoover of Product Hunt. It makes sense...if your idea is testable as an email then just do that at first...it's really inexpensive and will give you an idea of how valuable it is very quickly. (I was also working full-time at HubSpot then so a lightweight email was something I could actually squeeze into my day)

I initially focused the email on just the absolute minimum information that people needed to get dressed. I didn’t recommend “green North Face hoodie” like Swackett did. Instead I just focused on what type of clothing was appropriate for given weather, such as “lightweight long sleeve shirt”. This type of recommendation wasn’t as detailed as some people wanted, but it was enough to make a decision on so that’s what I stuck with. Early on I even left out the weather information itself…but quickly learned that people wanted the weather information to understand how the clothing recommendations were related. Having weather information helped build trust in the clothing recommendations.

And I sent out this email manually, every morning at 6am or so. I created the report body in an image editor and then attached it to an email that I send out to a growing email list. When I first started it took me about an hour each morning. Then as I got good at it I reduced the time down to about half an hour and even 20 minutes.

The power of feedback

The initial email was really rough...in greyscale and not very informative. But it did have enough of a pull that people started to read it every day. And then they spread the word for me and starting giving me feedback. This was invigorating! I realized that feedback is the oxygen of design...without that initial feedback (very critical in some cases) I would have stopped publishing the email. But the feedback gave me a way forward...things to improve. I learned pretty quickly to continuously ask for feedback and just keep getting as much as I can. In many cases people wouldn't have feedback until I asked them about it and told them that I want as much as possible. I got emails like "you asked me two months ago for feedback...I didn't have any then but I have some now". Every single feature I've added to the product has been a result of direct user feedback like this.

A fully realized product

After sending the email for a few months something funny happened...I actually had a real product on my hands! I didn't realize it until other people started sharing it for me...they told their friends that I had a cool email they should sign up for. I wasn't even taking the project that seriously until other people took it seriously first. Another lesson: your product is only as good as people say it is.

Fast forward a year later and I'm working on What to Wear full-time and growing faster than ever! From a single job-to-be-done I've created a product that normal people can use every day and get value from. The initial email testing was critical...and was sticky enough that I simply continued to send the email every day and it became the product! My guess is that many ideas could be tested and improved in the same way.

So that in a nutshell is how I went from a simple observation all the way to a real product. In the future I'll write up a post on the evolution of the What to Wear email UI...I have actually kept a copy of every single email that I've sent out. It's pretty amazing to see how far it's come!

You can subscribe to the free What to Wear email here. Let me know what you think!