James left his cushy enterprise job to start Appsurify, which helps software development teams release faster with fewer defects.
What’s your background, and what are you working on?
My name is James Farrier and I’m from New Zealand. I wrote my first program when I was 10 and then sort of forgot about programming until I hit university. There I rediscovered programming but found that my favourite part of software development was actually debugging, which I know is slightly masochistic. But debugging reminded me of solving puzzles which I loved doing as a child. Somewhat naturally my career path has moved towards Test Engineering and I’ve been in that space for the last 10 years mostly at large enterprise companies.
I quit my cushy enterprise job about 2 years ago and decided to start my own company called Appsurify. Basically, Appsurify analyzes your repository and test history, from which we help you create fewer defects and find them faster. A recent study by Stripe estimated that developers are only 58% productive and over 40% of a software development team’s time is spent on testing. Our tool helps you better understand your team’s strengths and weaknesses so you can distribute your tasks in a more effective manner. Your team’s output improves and the number of defects created reduces. Then we analyze the risk associated with your commits so you can target your testing and find bugs faster. We prioritize your tests based on the likelihood that they will find bugs and we detect flaky/unreliable automation and quarantine the results to prevent them from breaking your builds, in a similar way to how Microsoft does.
Visualize the risk associated with the files and commits of your projects.
So far our clients have seen a ~30% reduction in the number of defects created and a ~40% increase in testing effectiveness.
What motivated you to get started with your company?
While working at other companies I had noticed almost all of them were really struggling with their software development practices. They were creating too many bugs, they had no idea what to test and their test automation was often unreliable or took too long to run. These struggles often caused the projects to be delayed and teams to miss deadlines. I hated my team working weekends or being blamed for this and thought I could build something to help. I had been to conferences and seen companies like Google had far better tooling to support their development and testing. I thought I could build out similar functionality that could be used at any company. I hoped that by utilizing AI and automation to detect unreliable tests, assess risk and prioritize tests we could really help companies.
Automatically detect flaky/unreliable tests and categorize your tests
I had been building out test tooling at previous companies and seen some great improvement internally. So thankfully I had saved up some money and decided, rather than investing in property like almost every New Zealander and much to the dismay of my family, that I should start my own bootstrapped company. I reached out to my network and found some initial customers and we’ve worked with them since then to ensure that our app is helping them.
What went into building the initial product?
Our MVP took 4 months, I was the sole developer. I focused on the basics — detecting unreliable automation and preventing them from breaking the build. We were fully bootstrapped at this time and continue to be so. We use Trello for all of our task management and for development we use Python for all our backend work. As we’ve grown and added more features I think we have been too slow at releasing and tried to add too much at once. It’s a tough line to walk trying to have a product that satisfies enterprise customers, meaning it has the features they want and with no bugs and good support, yet trying to release frequently.
All of our initial customers came from my network. I’m lucky enough to know people in New Zealand, Australia, and the USA who are all struggling with these problems. Although it took a number of meetings to get going mostly there was little pushback and they were just interested in trialing the product. The customers have really driven the product, asking for features. I’ve worried though that with a small set of customers who were at the time nonpaying they’ve driven our development in areas that may not interest other companies, I guess this is a downside again of not wanting to release to a wider audience until everything is perfect. I know it’s said really often but try to find a way to release as frequently as possible, it’s a hard lesson to learn but one I will keep in mind from now on
Group tests by failure reason, priority or category.
Our focus recently has been on making onboarding as easy as possible so now a team can get set up in under 30 minutes to use our app. This was one of the initial
How have you attracted users and grown your company?
So although we have launched at this time we haven’t launched on anything like HN or Product Hunt. I mostly have just written on LinkedIn to see what type of traction we can get and make sure that potential customers are happy with our messaging.
Thankfully like I said our initial beta customers have all stayed on post-launch and will continue to be our customers. Getting new ones we are focusing on reaching out to Managers and above via email and LinkedIn. We are really just starting this. We are trying to be active on forums and are producing content which will hopefully be interesting to our target audience.
I hate cold outreach, but I think I have a product that can really help teams so I’m prepared to do it. But I’m trying to at least get introductions from my network before going direct. So far all of the new companies that are currently trialing our software have all come from my posts on LinkedIn on what we do and how we do it.
To go more into the cold messaging we are trying different techniques and different messaging. I have a giant list of possible subjects and we are trailing all of them to see which are most likely to work. We have the same for the actual emails with over 30 different emails trialed so far. We’ve considered targeting Development/Test Managers and above which has not worked so far, especially CTO’s seem very unlikely to read and respond to cold emails. For us our best traction has been bottom up, messaging Software Developers in Test who are interested in being involved with an AI tool, and Innovation Managers who are looking for ways to improve their development/testing process.
What’s your business model, and how have you grown your revenue?
We have a Saas offering with a fee per user and we use Stripe as our payment gateway. We are currently at $15,000 MRR, but this should be increasing as companies move from our trial to paid (crosses fingers). Before our launch we were not looking to bring on any new customers so our revenue was static, now it should start increasing as we gain new users.
To grow our revenue we are looking at possible integrations that might quickly grow our audience. Because we are a small company we are focusing not just on selling but on how we can partner and help the companies we work with. Our goal is to make sure that we improve their software/development process, so that might mean we offer suggestions on ways they can improve which are outside the scope of our tool and it means that we will also build out features that would help them and other companies. Again I think it is a different mindset because as a partner we just want them to succeed, we think of ourselves as a member of their team and look for ways for them to improve.
What are your goals for the future?
Our biggest goal is to bring more customers on. If we can have 20 customers by the end of the year I will be really happy. That will mean we have the revenue to expand our team and to grow our product offering. On a personal note, I’m actually getting married in December and we would love to be able to settle down in the United States. That again means we need the revenue to support ourselves and to get the correct Visa. In terms of our product offering, this will really depend on the companies we are working with and how we can best serve/work with them. But we see ourselves either diving deeper into test tooling, developer productivity or task assignment, all of which I think are underserved by the market at present.
What are the biggest challenges you’ve faced and obstacles you’ve overcome? If you had to start over, what would you do differently?
Our biggest mistake was taking too long to launch outside of our main 3 customers. This has slowed us down and also meant we weren’t discussing our product with a wider audience soon enough. Taking too long to launch, but I’ve talked about that a lot already.
I would say our second biggest challenge was being a distributed team. Obviously being a distributed team there are communication challenges and also understanding your team is more difficult as a manager. Interestingly Appsurify itself helped me identify underperforming team members, one of which we no longer work with and the other I now know that complex SQL queries are not his strong point or what he enjoys working on and so now he works on the areas he does enjoy.
Personally, I find marketing and reaching out to potential customers difficult, and the only reason why I do it is that I really feel like we can help teams succeed and increase their overall happiness.
One aspect I didn’t realize would be so challenging is the mental side of being a founder. I’m a pretty happy person but the ups and downs have been tough to deal with.
This is where having a support system and a co founder is so important. I’m very lucky to have supportive friends and family (especially Sharon, John, Kate and Rick).
Especially when bootstrapping your first customers are essential. If they stay with you know you’re not going to completely crash and burn. Before we got our first customers I was constantly worried as to whether I was on track, and someone not replying to an email or getting no traction on a post would cause some serious self doubt. But then the highs when people are interested, when they like the product or you’ve built something you or they love makes everything worth it.
Finally, I would say as a founder you are pulled in multiple different ways and understanding where to focus is key to actually getting things done.
Have you found anything particularly helpful or advantageous?
Startup School by Ycombinator has really helped us focus on the customers and what we are delivering, as has reading articles by Stripe. A big help has actually been articles from other companies detailing how they’ve improved their process or the challenges they’ve faced, then I’ve been able to share these with my network and talk about how we’ve done similar things. For example, both Google and Microsoft have recently talked about how they deal with unreliable/flaky test automation and how they utilize AI to improve their testing and we offer a similar solution which can easily be integrated into most companies workflow. Similarly, Stripe released a report about developer productivity and our solution tries to improve some of the challenges that they feel developers are facing — you can check it out here https://stripe.com/reports/developer-coefficient-2018.
But by far the big advantage has been both my network and my understanding of the problem being faced by these companies. I know the challenges so well having dealt with them for 10+ years and I’ve seen the solutions as well. This means that at least for our initial product offerings we were solving a pain point I was super familiar with and we instantly got traction.
What’s your advice for entrepreneurs who are just starting out?
Make sure you understand the pain point and industry you are looking to work with, but that is not all. You need to understand more than just their problems but also how they’ve solved similar challenges in the past and what solutions they’ve said no to. It’s best to have worked for or in the industry before so that you know everything about them, this will help you build the right product, not just that solves the problem but that they will be able to use, and also how to market to them.
Where can we go to learn more?
I know it’s bad but I am horrible with social media other than LinkedIn…
But if you do want to contact me or have any questions or comments you can either write in the comment section below or positive feedback can go to [email protected] and negative feedback can go to [email protected] (: — obviously joking negative feedback is just as important as positive feedback.
Before I go I just want to thank my team, they’ve been amazing and I couldn’t have done it without them. Also a big shout out to my family and friends again.