Recently we had the pleasure of sitting down for an extended conversation with Lon Binder, VP of Technology at Warby Parker. We discussed a range of issues central to how Warby Parker approaches IT, including new store launches, what IT is meant to achieve for a business, cloud strategy, and much more.
This is the first of two parts. Check out Part 2 here.
Thoughts? Feedback? Let us know on Twitter @CloudGathering.
Gathering Clouds: So you just had your recent store launch. How did that process go and what was your involvement from the perspective of the technical head of the company?
Lon Binder: The process of opening the store was really like launching a whole new business. We were technically multichannel before, because we generate the vast majority of our revenue through the E-commerce website, and have showrooms around the country, which represent a pretty fair piece of revenue as well.
GC: Yeah, I’d been to the one at Bird in Brooklyn.
LB: Yes, exactly. So the primary difference between the showroom and the store is that our showrooms are all co-located in another space, and the one in our office obviously is in our office. Bird is a perfect example: it’s a preexisting store. Some of them sell apparel or fashion accessories, some of them don’t; some of them are antique stores or bookstores or something else. For us, an in-person buying experience is a really a new concept; we can control the point of sale experience. We started to look at what kinds of technologies are available and realized we had to think about the security of what would happen in the store: do we want to go just as simple as something that’s public Wi-Fi or something more complex?
There was also the question of what else we would do at the store. For us, our flagship store in Soho is also an event venue. We put in a telescopic projector that comes down from the ceiling and a display board transforms into a projection screen. We put in audio equipment so that our parties would sound as good as they look. All that stuff fell to technology as well.
When we think about anything at Warby Parker, brand and creative are critical. So, when you walk into our new Soho store we immediately thought about where people’s eyes go. The biggest element, visually, in the store, is the display board for eye exams. We have a fulltime optometrist in the store. The appointments only cost fifty dollars and you can go on the website to book it. It’s completely self-service. When you go into the store we have a HIPAA compliant secure system.
For aesthetic of the eye exam booking system, we wanted to go for an old-fashioned train station footboard kind of feel. We started to look at building our own, and digital development came to the forefront of our process.
Now the store is open, and like any new system, there are bugs that have to be worked at, and quickly. So, using agile development methodologies, we just iterate through some of the software issues. There are still some hardware issues and we’ve been able to get through those to get the store to be the experience that we want it to be.
GC: What characterizes your approach to IT and the outcomes that you’re trying to drive for Warby Parker versus other places that you’ve been?
LB: What’s different about IT at Warby Parker is the support that the business gives to technology. One of my big hesitations when I joined was whether the business units outside of IT at Warby were going to look at IT as just a service center or not. In e-commerce, that’s actually quite a common outcome. E-commerce typically, and surprisingly, doesn’t see technology as a strategic investment; they see it as a “must-have,” or just another channel. And I think that’s why the retail experience for most companies is so awful: it’s just because they think of technology’s role in user experience secondarily.
During my interview I was actually said technology is the most important thing that Warby Parker was looking to invest in, that it would be the thing that put the business ahead. And that was exciting.
That made me feel great and I think the entire technology team feels that as well. Nobody working in our engineering group right now feels like they are just getting told what to do or getting whipped around. Instead, it feels very empowering and IT-ers are really heard.
For example, we just did a workplace study. We were surveyed for employee happiness and one of the questions was “How do you feel about your ideas? Do you feel they’re being heard, listened to, or implemented?” And of all the questions they asked, that was the number one best metric our third party surveyor rated us against compared to the top one percent of companies in the country across all industries. We exceeded them by a wide margin in terms of innovation and I think that’s what’s different about our engineering team. So when an employee says, “Hey, I have an idea to build our own vacuum form slides and wire our own hardware,” we don’t shut that idea down, because it’s insane. We say “Oh my god, that’s awesome, let’s talk about it.” That idea, for example, went right through into production very quickly.
That doesn’t mean every idea gets taken; it just means we’re open to ideas, and that’s the way we run engineering. We try to be as objective as possible. People ask us whether we are building or buying everything ourselves – there’s no such rule at Warby about buy everything/build everything. Everything is an open question and we have concepts in mind – targets and goals, especially around speed and quality – but for the most part we’re open to it all.
GC: If it’s an all-of-the-above approach in terms of how you’re viewing IT, how does cloud play to that strategy? What’s your perspective on how that piece of your infrastructure technology’s meant to work for you?
LB: Cloud is amazing, first off. What we like about cloud is that it helps us with speed to market especially. When we think about these questions, like I said, speed is always there, quality is always there – those are two elements. We actually work with a number of Cloud providers already for a range of different purposes. Our entire ERP is based in the cloud, as one example.
When we think about new systems or revamps to existing systems, that same question that I mentioned earlier comes back: what’s the best way for us to solve this problem? In some cases cloud makes a lot of sense. In some cases cloud has some limitations. For example when we need to completely own something, or when we want all the data in-house and need to have 100% control over our systems, cloud has limitations. At that point we begin discussing whether we move to a private cloud. So if it’s a virtualization system, for example, we have to consider whether we go to somebody else’s virtualization platform or run our own.
And it’s the same types of considerations for the services we use: if the service is a cloud service, maybe we can launch with it. But if it’s really a major component to our process, perhaps we decide we should own the whole thing. Is there an option to bring it home eventually? Those are just a few of the issues we consistently think about and questions we try to ask.
GC: You were talking about time to market, but what goes into that for you? I ask because your time to market is more than just a product launch or being able to produce a physical item. What does cloud do for that to support both aspects of time to market?
LB: So as you say, time to market refers to both our frame products as well as our software products. At Warby Parker we have over 65 systems in production, so we consider all of those to be technology products that we manage. So time to update a feature, time to resolve an issue – those are all time to market for us.
Here’s a perfect example of how this process works through technology: we’re thinking about frame design, which is a physical product that has to get into someone’s hands. We do a lot of frame design and so we start breaking down the production life cycle and looking for where there are areas to improve. One area where cloud doesn’t help, but technology can, is 3D printing. We started looking at 3D printing as a way to cut down production fat and turnaround time when we’re shipping back and forth. Putting a 3D printer in the office makes that a lot easier, and it’s very difficult for a Cloud provider to help with that particular aspect of our production process.
There are cloud-ish service providers for 3D printing, where you can use a website and they’ll ship your product back to you. But the value you receive is not truly cloud.
An example of where cloud has helped us is when we needed a very specific project management tool around the life cycle of production and for integrated collaboration, which includes parties all around the world for us. In that situation we went to Smartsheets, which is a hosted project management tool.
We were able to customize it exactly to our needs and we were able to get that launched in days. It’s been great.
GC: What are you able to craft for the user experience via quick time to market and the highly agile development cycle, that cloud enables?
LB: That’s a great question. When we think about technology delivery adding value to the business, it starts with customer experience; it’s the number one metric that we measure in the company to judge our performance (via Net Promoter Score). What we do with Net Promoter Score is, if a customer tells us he or she really loves the company, we ask why and what elements are working.
If they’re neutral, we’ll ask them what we can do to improve the company and their experience with our brand. We’re always focused on how we can improve our buying experience to help people move from neutral to promoters.
So when thinking about time to market, typically we’ll work to incorporate input from our customers, and turn it into a feature or an enhancement to any of our software systems or the website in an effort to make it a better shopping experience for the customer.
One example of that is our home try-on feature. A lot of our customers use this feature in particular. They go on the website, they’ll pick five pairs, they’ll have them shipped to either their home or office; they’ll try them on and maybe their friends or coworkers try them on and then they pick the one they like and they send them back. Or if they want to try more they can ask for more. To do this, we ask for a credit card, just so that if you keep all our glasses we can be reimbursed. But we don’t intend to charge you, it’s totally free.
GC: And I’ve done it, it’s very cool.
LB: Oh, you’ve done it? Cool!
The vast, vast majority of people who do a home try-on want to purchase glasses, they’re not just experimenting. So that’s a feature, for example, where having our own control over the system is really important, because we can turn around that feature very quickly and add that to the website.
In a different circumstance, our phone system, for example, is voice over IP, but we don’t have full control over it. One of the things that we want to do is use similar feedback from our customers to get data about the way our call center works in an effort to improve that feature. It has been a little bit challenging for us to get the turnaround times that we need, because in that particular example, the cloud provider for our phone system hasn’t been able to do it as quickly as we want. But over time they’ve understood our needs.
And that’s one of the nicer things about our growing brand: the companies we work with understand that as a growing business we have a really good understanding of their platform and how we need to use it. We’re really taking manageable auto-features and making them work for our model, and our vendors have asked for our feedback, which they then incorporate into their development roadmap.
GC: Do you see a point at which the speed that you’re matching the demand of your growth levels out, or is it built into the very culture of your organization to always be pushing it in terms of rolling out new products and feature sets? Is that actual demand derived from your customer base?
LB: It’s both. We listen to customers, because we want to make sure that they’re satisfied.
Where we’re most innovative is in the office. Our team thinks about our customers 24/7 and constantly come up with ideas. In the end it’s a combination of inputs from our team and our customers. But we never want technology to be first; it’s always customer experience first. We’re never thinking how to come up with like a whiz bang feature that looks really awesome; we’re always thinking about what we can do to improve the customer experience. The best way to do that is with transparent technology, where the customer doesn’t even see it.
We have some features that are very tech-centric – our virtual try-on, for example. Customers use it a lot and feel it’s really amazing. You just look into a webcam, and see the glasses on your face. It just looks awesome and who cares how it works? It’s just a really great experience. It’s very obvious technology in the sense that you realize there’s something crazy going on with your webcam.
But then we have other features that are a lot more casual. For example if a customer didn’t select five frames for their home try-on – they selected less – we could populate an additional frame into the assortment, based on an engine that we wrote statistically to figure out what customers might enjoy seeing in the box as a fifth frame. That’s an example where there’s heavy technology and data at work, but completely not visible to a customer.
We’re trying to make technology the enabler of magical experiences for the customer, and again, the goal is always customer experience.
When we’re hiring engineers, that’s one of the most interesting qualities we look for: the ability to understand how the tech will work for the end user. When people are interviewing with us, a lot of times they’ll say things like, “Wouldn’t it just be amazing if there was this one other thing.” When they do that, we see it as a really good sign that this is somebody who’s very thoughtful of customers.
By Jake Gardner