Let’s build the future that we want to live in.
Designing Products from Scratch: An Interview with UX Leader Josh Porter
Josh Porter is the founder of What to Wear Daily, a service that helps you prepare for your day. He is best-known as the creator of 52 Weeks of UX and was also head of UX at HubSpot, by way of being co-founder of Performable (acquired in 2011). Below, Josh talks about designing features from nothing and how to make important design choices during the pre-seed and seed-stage phases of startup growth.
You Found a Problem to Solve. Great … But Now What?
NextView Ventures: Let’s start at the beginning. Once you identified that What to Wear Daily was worth building, what worked in acquiring your first 100 users — especially since it was focused on just Boston to start?
Josh Porter: In the very, very beginning I started with just 10 people, all friends from HubSpot. I woke up each morning at 5 a.m. and created the report manually in Sketch (similar to Photoshop), and hand-delivered individual emails through my Gmail account to those initial users. My goal at this very early stage wasn’t to figure out how to unlock growth but to create a product people loved.
To figure out how to make something people loved, I went after feedback with a club. At the top of each daily email, I asked those 10 initial users questions like, “What do you think of the new clothing icons for today?” Because my emails were personal, I got a lot of great feedback and started what I now call the “virtuous feedback loop,” which is when people realize that you’re actually using your feedback to improve the product, so they keep giving you feedback. If I didn’t feel like I got enough feedback over email, I followed up with people in person.
So What to Wear grew from 10 to 100 mostly organically through people telling their friends. I also recruited some users by just talking about what I was working on with friends at HubSpot. By being so close to users from the very beginning I was able to really understand what the “must have” experience of the product was, and I used that in my pitch.
NVV: What about the next few hundred to 1,000 users? Any new or different approaches compared to acquiring the first handful?
JP: So I love this question because it gets to a very important issue in product design and growth: user bias.
Initially when you start, you can talk to friends and get them excited about what you’re working on. These people are your co-workers, friends, and family so they’re easy to find and talk to. But they are also completely biased. They want you to succeed and they’ll overlook almost anything because of it. The first 100 people are relatively easy from a user standpoint, and even if they are using your product, you can’t really be sure its not just their bias in wanting you to succeed.
Then, however, you need to grow beyond your friends to your friends’ friends. This is where a tougher test happens, since your users aren’t as biased to see you succeed but they are biased a lot by their friends (so there is still bias). This is still a better test because they don’t really care about you as a friend; they just want a valuable service. So if they start using it, you know you’re adding real value.
So one part of the new approach (up to 1,000) is to make these strangers into friends. I try my best to respond to each and every email that I get to show people that I’m just a normal guy listening to them closely and using their feedback to build something valuable. In some cases, people actually start to become fans as well. My goal is to be personable enough that every new user has the feeling that there is a real person behind this who cares (because there is).
In terms of channels, word-of-mouth has always been the most powerful channel for What to Wear. Over 60% of people subscribe because someone told them about it. But Twitter was another channel I used during that time. I’ve been active on Twitter for a long time and have a good readership there.
NVV: What changes in the way you should approach product design when you’re past the initial 100 or so alpha testers and now want to serve a larger group?
JP: Well, the biggest thing is that you have to start saying “No” a lot. Everybody has great ideas and everyone sees the product in a slightly different way, but as a product designer you only have so many hours in the day. You can’t build along all the vectors that people want to take you, so you have to get clearer every day on what exactly you’re building … and what you’re not.
NVV: Along those lines of what you should build or shouldn’t, you recently tweeted that if adding a new feature is your approach to getting people to love your product, you’ve failed. Why? What did you mean by that?
JP: What I mean by that is, if you have a product and people don’t love it, then you should redesign the product. If you’re hanging your hat on some additional feature to your product in the hopes that this will finally be the feature that gets people to use your product, and you’re adding it without redesigning your core, it’s probably not going to work.
I’ve seen this many, many times. A product team building a product with flat or no usage pins their hopes on the new feature they’re adding. “If we just add this new feature, then people will use our product,” they say. It almost never works, because the core product is still the same. Instead, the product team should redesign the core product until people are using it well.
NVV: If a startup’s product or feature isn’t sticking, how can they tell if it’s the feature itself or if it’s the design of that feature? In other words, maybe it’s what users actually do want, but it was implemented improperly. How do you learn that?
JP: Yeah, this is related to the previous question and is a core product design problem. When you have a feature and it’s not getting used well, you have to question its existence. Should you keep it around or should you cut it? The worst decision is to keep it around because it’s already there and it would take work to remove. A close second-worst decision is to kill it off and just assume that the feature was a bad idea.
I’ve seen this a thousand times too: “Oh yeah, we tried that, and it didn’t work.” In 95% of those cases, though, the initial manifestation of the feature just wasn’t the right one. Think about it: The first time humans do anything, we’re not very good at it! I think the same goes for features.
Good product design is the result of persistence as much as anything else. That’s why you hear companies like Airbnb and Evernote and others say that it can take years to figure out user onboarding or free trials or other sophisticated flows. It takes a lot of time to get your message right, get the user incentives right, get the interaction right, etc. It is easy to become impatient and throw out some great ideas prematurely.
But some features really don’t work. So as a product designer, you have to continually assess the distinction between your product vision and its execution. You need to be able to tell the difference between the right feature implemented poorly and a bad feature.
NVV: So how do you approach that tactically and strategically?
JP: Like most user experience issues, you do this by gathering multiple types of feedback, both quantitative and qualitative. You look at quantitative usage data to see how things are actually being used. For those features that aren’t being used as much as you expect, you need to dig deeper with qualitative feedback to understand the problem.
So interview folks, and find out why they aren’t using it. They will typically know — they might not care or they might not understand the value of the feature. In some cases, they will have a very specific reason why it’s not working for them. Here is the opportunity: You may be able to fix the problem and kickstart usage of the feature.
Quantitative feedback tells you what is happening (or not happening), and qualitative feedback helps you understand why. Never trust just one of these types of data. Use both together to really understand your product’s usage patterns.
At the Intersection of Intuition, Feedback, and Data
NVV: You seem to incorporate small changes and tweaks that over time change the complexion of the product. Two questions along those lines:
(1) How do you know how long to keep the product the same before tweaking/adding? And (2) how do you prioritize what to actually change, especially when you probably have some intuition and vision that needs to now co-exist with user feedback?
JP: To answer the first question: After you launch a new feature, you have to give it a while. It takes time for people to notice and try your new thing. We try to wait for actual usage numbers, and while we’re waiting, I’m also talking with folks about how they’re liking the feature, how it can be improved, etc. This is where the feedback loop becomes virtuous. When people know you’re listening and designing based on their feedback they will go out of their way to help. In general, though, I wait until I’m confident I understand what’s really going on with the feature before tweaking — enough so that no new feedback surprises me anymore.
As for how I prioritize change and balance intuition with feedback, it depends on your current product focus. Is it improving product satisfaction, usage, or growth? If you’re measuring all of these over time, then one easy way to choose is to simply focus on the one when it drops below an acceptable threshold. For example, if your monthly growth goal is 5%, and last month was only 3.5%, then that’s what you should focus on (and improve features/campaigns related to growth).
The interplay between vision/intuition and data/feedback is that you need both. You need to be able to assess how you’re currently doing with data but also not let data stop you from taking risks with your product. Many teams become conservative over time as they rely more and more on data to make decisions. They ultimately become paralyzed and unable to build something really new.
So intuition plays a big part in knowing when to make a big leap. I try to divorce the behavior from the product: What are people doing that they’re going to do regardless of whether they’re using my product to help them? If you truly understand the behavior then you can more easily imagine another solution (that might be a huge leap beyond your current product). So think about problems like that: “I’m solving for this particular behavior,” not, “I’m creating this particular product.”
It’s easy to get stuck thinking your product is the only hammer worth using for this nail.
NVV: What are the top lessons you’ve learned about product design and development in your time doing What to Wear?
JP: One lesson is that people will root for you at the beginning like they never would later — and you need to take advantage of that. People love to see their friends working on a product, whether it’s because they use the product or whether they’re just impressed that you took the risk to go out on your own. So it’s important to tell your story and take advantage of that support. I have friends who ask me how it’s going every time I see them, and it’s great.
Another big lesson is just how tricky user feedback is. Some people (even friends) will just tell you what you want to hear. They’ll say they use your product and it’s great. It’s so easy to hear that and feel good about yourself and go on assuming everything is fine. But often that’s just not true; they have very real problems with your product and you have to be a really good interviewer to suss it out.
I often have chats with users where I don’t actually get into the meat of the situation until I give them an opening to share negative feedback. I’ll say something like, “You know, I have a feeling that this part of the UI just isn’t working 100%. How are you finding it?” And just that little opening allows them to unburden themselves. They might just say, “Yeah I don’t use that feature because I don’t find it valuable,” or, “I was thinking the same thing, and here’s why it doesn’t work for me.”
In general the lesson is to always be skeptical of positive feedback, don’t take it at face value, and help prompt people for giving you negative feedback.
NVV: What’s next for your product and its design? More importantly, why is that next?
JP: So the next step is just increasingly personalizing the daily email. We have minimal personalization now, and our product and growth numbers are good. But I think we can take it to the next level with some really innovative personalized features that continue to do what we’ve been doing: help people prepare for their day.
The reason why this is next is because we’re product-centric. We fundamentally believe that creating a great product is the best way forward. Any success (growth, revenue, or profits) we have will follow from that. Personalization is going to make our subscribers happy, and that can only lead to good things.