What I learnt in 2017 while running Feature Upvote
We launched Feature Upvote in early 2017. Here are some lessons I learnt along the way.
The feeling of getting the first paying customer for a new product is euphoric
It is hard to describe how good it feels when a stranger on the Internet chooses to pay for your new product for the first time. The second and subsequent sales are pretty good too, but nothing beats the high when the first notification email comes from Stripe (our payment processor) saying “Congratulations! You’ve just received a payment.”
Creating a product is an investment of time, money, and emotion. The first sale is a partial validation of the investment. It is when I first started to believe that Feature Upvote would succeed.
Almost as good a feeling is when the first customer’s subscription renews automatically after one month.
My typical customer is not who I thought it would be
I had a clear idea of our target customer. They’d be a small software company with revenue up to $50K per month. They’d have a successful product, and would be having trouble with tracking feature requests.
In other words, my own company would be our target customer.
I was way off. Our actual customers are diverse and often have nothing to do with software. They are often large and established organisations. Typically they are a small team within a company, and they tend to be focussed on product management and development.
Our typical customer is, foremost, a product person.
Quora is a great way to get initial customers for a B2B app
We got most of our first 10 customers via my answers to questions on Quora, a Q&A website. This surprised me.
An acquaintance suggested I answer some questions about “alternatives to UserVoice”. She even found two questions for me to answer. I was dubious, but it seemed worth spending a few minutes answering to see what happens.
Almost immediately after I did so, Quora became our main source of trial customers.
As the months passed and our SEO strategy started delivering results, Google has surpassed Quora as our main source of new customers.
Having a sysadmin from day one is good
I followed the advice from Starting and Sustaining to have a freelance system administrator as one of my first tasks:
“If you don’t have an experienced system administrator on your team, then one of your first tasks should be to find someone with that experience. This person or team may only work with you two to five hours a month, but these will be essential hours. It can be challenging to get your application’s architecture right at this point, but a good system administrator can help you do that.”
Our system administrator does the things that a software company should do but often doesn’t. He checks our database backups can be restored. He makes sure we don’t have suspicious activity on our servers. He deals with server problems while I continue with product development.
When Feature Upvote’s traffic suddenly went up 100-fold and buckled under the weight, my system administrator helped with a short-term solution and researched a long-term solution. He was able to do so because he knew our system infrastructure so well.
Having a sysadmin in a different timezone is great
Our uptime has been good so far. In our first six months we didn’t have a single second of downtime. But when eventually we did go down, I was sleeping. My sysadmin wasn’t. By the time I woke up he had dealt with the problem. Instead of a morning full of stress, I could first enjoy my breakfast and then calmly conduct a post-mortem investigation of the problem.
Some trial users convert to paid users on day 1. Some convert after several months.
In theory, our customer acquisition goes as follows:
- A potential customer finds our marketing website via Google or a referral link.
- Said customer signs up for a free 30 day trial and decides during that time whether our product is right for them.
- At the end of the trial, the customer either subscribes to a paid plan or goes away forever.
But customers do things according to their own timetable.
Some users subscribe to a paid plan on the day they start their trial account. Some wait until the last day of their trial period. Others finish their trial period, then go silent for some months before suddenly subscribing.
Usability testing has an outstanding return on investment
Usability testing involves watching as people try to use your software to accomplish specific tasks.
It is hard to tell what keeps potential customers from successfully using your product until you watch them in real life. You’ll see them stumbling over things that you didn’t realise were confusing.
Without usability testing, people stop using your product and you don’t know why. You can only assume it is because they didn’t want your product. But often it is because your product had fundamental design flaws.
My small team sat silently together as we watched volunteers try to use Feature Upvote. It was humbling and illuminating. As soon as we saw humans use our product, we could see that decisions we had made after much discussion turned out to be utterly wrong.
We changed our first-time experience in response. The result is a smooth workflow that seems to achieve the simplicity we want in our product.
Feature Upvote is a Viable Product
This is actually the most important thing I learnt. I started the year with a rough prototype and no customers. I now have a working product, paying customers, and a process for getting more customers. I didn’t know whether I’d find room in a crowded market. But thankfully I did.
Feature Upvote lets your customers openly suggest and upvote improvements to your product.
If you are already a Feature Upvote customer, thank you! If you aren’t, and you want to let your customers openly suggest and upvote improvements to your product, sign up for a free fully-functioning trial of Feature Upvote.
