At my startup Filter Squad we build mobile apps such as Discovr where turn around times can be long (due to the Apple review process etc) and where App Store rankings and discovery make and break companies.
After reading a lot about "Lean Startups" and "Minimum Viable Product (MVP)" I have always been keen to give it a try.
I believe web is the ideal platform for MVP, it's open to everyone, you can release at will and you can measure analytics easily.
AppBot came about to solve a pain point my own workflow, trying to keep up to date with our iOS reviews and features. If you are having an problem in your own workflow, chances are others are as well.
The initial version of AppBot was built in about 4 hours, it would run as a daily cronjob on my laptop and email me the reviews daily.
After I knew it was working, I showed it to a few colleges who wanted in on the emails. Next came small iterations of adding a web interface, deploying etc.
For me this is the easy part.
AppBot has very little code, most of the functionality comes from amazing frameworks like Bootstrap, Ruby on Rails and a long list of gems built and tested by countless others.
I initially reached out to 6 iPhone developers I knew who I thought might find it useful, 4 signed up immediately, 2 didn't. 66% strike rate I thought was pretty good, then another 4 people signed up who were either colleges or friends of the initial test group, good signs.
I received some great feedback from this small group, things such as :
As a pretty targeted tool I wanted to tell iOS developers about it, without spending any money. So I wrote a blog post How we track our app reviews, and why you should too, tweeted it and put it on Hacker News.
Luckily it got a little bit of press:
In the first 48 hours or so it lead to 372 users tracking 801 apps. Not exactly Instagram, but solid for the market size. Proof enough to push on.
As I have seen signups from apps and companies I recognise I have reached out to them and asked for their feedback. Every person I have reached out to has been thankful and given great feedback.
In every email and on the website I provide my email and Twitter account so people can get in touch. This has resulted in calls with some of the biggest names in the App industry, free shirts (I love a free shirt if you are listening) and a lot of great feedback.
Many people have even emailed to suggest potential ways to monetize the product (more on that later).
While there is a lot of great feedback, much of which has been implemented, not all the feature suggestions are necessarily right.
I was sent a complete rewrite of the copy for the homepage, although it was much better than the original it was also much longer. I took it blindly. The result was a massive drop off in conversion percentage over the next few days, a quick backflip to a cut down version and the conversions returned.
It's difficult to say no, especially when you have asked for the feedback, so be prepared to say no (nicely), or at least work the idea to suit the aim of the product.
Track and watch everything, Google Analytics, conversion rates, users, open rates, A/B testing etc. This is the key to picking the right marketing and measuring your success.
We have found a very typical pattern where somebody from a large company will sign up, then a few days later there will be a number of others sign up from the same company. I have confirmed in a number of cases this is where the person has received one or two days worth of reviews, then passed it around to the rest of their team. There are now 8 companies with more than 5 users, with two companies equal on 15 users each.
These type of evangelists are key to spreading the news about your product, and great validation of the idea.
What do you have to lose? A few dollars in hosting and a bit of time. It's a lot of fun and you will learn a lot.
Two months in we are at 2019 apps tracked by 717 users, with slow, but steady growth.
Only 11 users have cancelled. Emails each day have a open rate of between 30% and 40% (only counting compatible clients, so most likely much higher), which is more than double the industry average.
Feature wise AppBot does everything I ever wanted it to do, users come up with great new ideas all the time. But it sure could use a bit of design love.
The next level of validation would be to find the business model for AppBot. I have been amazed at the way users have reached out to offer ideas on how it could be monetized, as they want it to succeed. This is not something I had ever considered doing, but to complete the MVP experiment it could be worth it.
Watch this space.