Making Qwilr
Hundreds of millions in transactions and hundreds of thousands of users.
The Product Bug
My first year out of high school I did something I’d advise almost no young person to do: I built my first start up as a solo founder.
Doing it by myself was less of a choice than a consequence of my social milieux.
Growing up in sunny Sydney, my friends either played in rock bands or played rugby. The gym, the beach, the pub. These were our destinations. Girls, drugs, music, sport. These were our subjects of conversation. No one I knew was much interested in computer programming or design.
So it was by default that I became the Everything Guy for my first product business. I did the design and code work, raised money, hunted down early beta testers, pounded pavement as a salesmen and ran the user feedback sessions.
By the time I left high school I’d become a decent programmer and a so-so designer but in the arena of business and product, I really had no idea what I was I doing.
I remember it being an exciting time, but also very confusing. The information I could find on the web about starting businesses was very much aimed at the traditional Small Business of yore. The word "startup" had yet to enter common use - and I had no idea that I was building one. "The Lean Startup" was still years away from being published, let alone popularised.
In the end the business didn't fly. The doors shut after a year. I was nineteen and quite exhausted, but I had caught The Product Bug.
I've never lost it since.
The Genesis Of A Product
Qwilr started out as a disobedient idea. One that simply wouldn't go away.
After my first start up died, I thought I had better try the work-a-day professional life, went to work in advertising, and got jobs at a few different agencies. The people were all lovely, but I found the experience to be intellectually and creatively claustrophobic. I lasted around nine months in total.
Consulting seemed like then seemed like a viable alternative path. A middle road between creative freedom and a reliable income.
I started my consulting career very small. My first client was a referral from the printers where I got my business cards made, but I slowly and steadily I worked my way up the rungs of clientele.
I spent the next eight years in the consultancy game and in that time got to experience the full gamut of digital. From identity and branding, to marketing sites and interface design and to front end builds and interaction design, to back end builds and questions of infrastructure and system design.
Fast forward to 2012 and business was good. It was really good. I was getting a steady flow of high quality inbound briefs and it was no longer just the Mom-And-Pop jobs, it was now product companies, agencies, multinationals and government projects.
At the time, my tools for creating proposals consisted of Excel for pricing, IA Writer for language and composition, and then a combination of InDesign, Photoshop and Illustrator for packaging everything together into a PDF document. This suite of tools certainly worked; but it was highly inefficient. Each and every document required a significant outlay of my time.
Every now and then, I’d receive a brief from a whale of a brand. The kind of logo that opens doors for future jobs. I knew that the best proposals were highly tailored and very polished. So if I really wanted to land the contract, I delivered my proposal as a purpose built website. It provided an opportunity to showcase my design nous, engineering skills, the quality of my personal brand and depth of experience. The website strategy had an excellent conversion rate - but was a risky time investment. Sometimes I didn't the land job. Days were wasted.
Therein lay the business challenge. While there were often numerous briefs I was interested in, I simply didn't have the capacity to respond to all of them in such a sophisticated manner.
Then one day, clicking away at yet another InDesign doc - I came to a realisation: "This sucks. This totally sucks and I need to do something about it".
I pulled out a gridded sketch book and started to doodle interfaces: "Now in an ideal world... what I want is a unified and automated way to produce these documents. I want the system to help calculate my costs, generate my copy, build up examples of relevant client work and include the contractual pieces and I don't want to always be repeating myself, I want to assemble new documents from fragments of ones, from a library of content I've already used (case studies, introductions, quotes, process explanations etc.) - and I want the whole thing delivered as an awesome looking self-hosting website...".
I put my pen down. It was a pleasant sketch - but not feasibly buildable. As an engineer, I knew just how ambitious this idea was. So I closed my sketch book, sighed and got on with my torturous proposal process.
And that was that. Or so I thought. Like one of those dogs that miraculously finds its way home across state and country, the idea just wouldn't go away.
A Bash Script Does Not A Product Make
This notion of automating my whole complex workflow took root. It tantalised with promise. Action had to be taken.
So I spent two days crafting a glorified bash script.
The outcome was by and large a success. Based on an arcane list of terminal commands, the script created a nifty website, and the process took about 15 - 20 minutes per brief. I got a 10x improvement in my time expenditure per proposal and I was now able to respond to a large volume of briefs with a bespoke mini-website. Hurrah!
With such short turn-around times on the proposals, the companies recieving them started to ask questions. “This is pretty cool! How did you do that? Could you do something like this for us?”. With my indecipherable bash-script the answer was "well, kind-of but not really".
But it got me thinking: maybe there's a genuine product here, perhaps there IS real demand for something like this?
Version 1.0
I hadn't forgotten my first startup experience. I did the due diligence.
Before my design toolkit was opened, or a line of code written, I undertook extensive customer development interviews. The results challenged a lot of assumptions I had made about the requisite feature set. I pushed and pulled, added and removed.
Within a few weeks, I was satisfied that there was a genuine problem to be solved, with a breadth of market applications. So I set about turning that The-Idea-That-Wouldn't-Go-Away into a nuts and bolts, designed-and-coded, web-accesible reality.
I called the product "Qwilr". (Etymological Digression: the "quill" was made obsolete by the huge efficiency improvement that came with the invention of the printing-press, and since I was aspiring to obsolete the current approach to document creation with a similiarly massive increase in efficiency: "Qwilr").
Those early product creation days were a real adventure. The learning curve was almost vertical. Over the course of that first build, every wall in my house was overrun with database schematics, design notes, interface sketches, class wiring diagrams and notes-to-self. One friend described the interior decor as "Rain Man Chic", another less kindly as "Serial Killer Chic".
After 3 months of hard work my prototype was stable. I signed up my first early customers and began consulting with users and gathering feedback.
It wasn't long before I realised I had made a fundamental mistake. A faulty assumption at the very core of the product design...
Tearing It All Down
The quoting tool? People loved it ("so simple to use!"). The overall product design? Users thought it was great ("super intuitive!"). But something wasn't right. Questions kept rolling in about how to customise the output website. Questions for which I didn't have any satisfactory answers.
I'd assumed people wanted rigidity and cookie-cutter consistency in the output. Every document cut from the same mold. A corporate production line. That was my mistake: I'd built something more akin to Wordpress, than to a design tool. Wordpress is open and hackable, which is great, but you really have to be a developer to dive in and start extending it.
Qwilr 1.0 made the calculation of time costs and the assembly of copy and content really easy, but fell short in crafting the presentation. Tailoring the output website was simple... with design / development skills in-house. Few of my early-adopting businesses were developers or designers - and those that were, didn't want to spend precious engineering or design time. Especially with an unproven product.
Users were happy enough. But I wanted them to be delighted. I wanted them to tell their friends and spouses. I wanted Qwilr to be the answer to their productivity hopes and dreams. Not just another dusty tool.
I was faced with a dilemma...
Product design instinct told me: "tear everything down and start from scratch - merge the content creation experience and the content output-styling experience into one!".
But did I really want to go right back into R&D mode, after months of arduous work to finally get this product out into the world? Did I want to take a wild, hairy, ambitious idea and make it even wilder, hairier and more ambitious?
My little start-up product was already in a comfortable position: I was speaking to a number of potential investors. Some had expressed interest in investing. I had businesses using the product, and some willing to pay for it. It didn't seem like a practical time to go to ground all over again.
But as Bob Dylan says: "If ain't right, it's wrong". Words I do my honest best to live by.
In the end, I tore everything down. The entire project was rebuilt from scratch and it was the best product decision I've ever made.
No Man Is An Island
If the experience of my first startup had taught me anything, it was this: "going it alone is really, really, really tough". It was never going to work trying to do everything myself. I knew I needed a team of top-quality people to help me chase after this big idea.
So, while I began the mammoth task of rebuilding Qwilr from the ground up - I also began searching for Qwilr's founding team.
The particulars of finding each of our team members is a piece unto itself, a whole series really, so suffice to say: it involved months and months of coffees and fourth, fifth, sixth hand introductions, as well as a dollop of luck and social serendipity.
From backgrounds at Google, Dropbox and Microsoft - technical leads and business development - I think the Qwilr team are about the highest calibre folks I've ever worked with.
If you take your time to choose the right people to work with, they'll propel you forward with a velocity and energy that would be impossible to acheive under your own steam. You can't choose your family but you can choose your team!
Update: 2014
About a year later and Qwilr is now alive and well. We've got thousands of users, an incredibly low churn rate, an amazing core team and a group of excellent and supportive investors.
Its been a long and circuitous journey to this point. At times hugely challenging and stressful - but for that, equally rewarding and gratifying. I imagine the road ahead to a sustainable long-term business, will be just as long and hard; but I'm incredibly excited to be walking it.