WindTunnel

As promised in the last update, the portfolio has recently been expanded with not one but two smaller (but nonetheless very useful ;-) plugins: https://marketplace.atlassian.com/apps/1223553/metadata-maniac-for-bamboo and https://marketplace.atlassian.com/apps/1223693/image-badges-for-bamboo. The most important lesson from launching these plugins is that Atlassian is really slacking off in terms of their app approval process recently. It took them respectively 30 (!) and 20 days to review the apps. That is a very long time to wait, and a significant anticlimax after submitting :-) Given they already earned 3000$ commission on the sales of my apps, I would expect them to do better.

On the upside, business has been steadily growing. As the title boldly states, I hit 10K all-time revenue a week ago, and a total of 125 evaluations. My original business plan (made in April) projected 4K in revenue for this year, so that milestone is definitely hit.

All was good till the following announcement from Atlassian was made: https://www.atlassian.com/blog/announcements/journey-to-cloud. This essentially boils down to them saying self-hosted products (and their users) will be handled as second-class citizens from now on. Given my focus on Bamboo (which is already one of the lesser-used products in the Atlassian stack), a server-only product with currently no data-center equivalent, this makes me quite uncomfortable. My current portfolio is very risky in the light of this announcement. I should probably expand into other products in the stack (preferably cloud) to secure the future of this endeavour. The main problem: competition is fierce! Jira and Confluence combined have almost 7000 apps in the Marketplace at this point, which is mind boggling compared to the less than 200 apps for Bamboo. Also, Atlassian's cloud API's are seriously lacking compared to the hosted products. Most of the idea's I have I simply cannot realise on a cloud product... Needless to say, I'm not decided yet what my next move should be :-)

Anyone else impacted by this Atlassian decision? Please let me know.

The 'Secret Managers for Bamboo' plugin has been gaining some traction lately. I think this is a combination of the plugin documentation site ranking very good on Google for relevant keywords, and me being present on the Atlassian user forums to answer Bamboo related questions.

I'm now at an all-time total of 50 evaluation licenses and $1K in (hopefully recurring) sales for a total of 5 paying customers.

I'm only working part-time on this, but was (as always) hoping for a bit more traction 6 months into it. Point is, I have no reference and data about Atlassian Marketplace mechanics is hard to get. I know I should be doing more on the content-marketing side, but I have a hard time coming up with good topics for my niche.

Nevertheless, I'm going to expand my portfolio for now and start working on a third plugin... stay tuned.

In parallel with releasing a few updates to the initial Atlassian Bamboo plugin, I have been working on a more complex integration which allows launching and running Bamboo build agents on Kubernetes. The plugin is feature complete but needs additional testing on different Kubernetes environments. The plugin was accepted in the Atlassian Marketplace today, and can be found at https://marketplace.atlassian.com/apps/1222674/kubernetes-agents-for-bamboo. The documentation has been completed as well at https://windtunnel.io/products/ksb.

I'm happy that the Atlassian vendor page now shows two available plugins at https://marketplace.atlassian.com/vendors/1217722/windtunnel-technologies, which will hopefully add credibility and signals long-term commitment.

This one caught me off guard.

The normal sales process for an Atlassian plugin is to have a 30 to 90 day evaluation period, after which a company might or might not decide to buy a commercial license. This morning however, a sales transaction appeared in my Atlassian vendor page from a company I did not see before. They apparently decided to proceed with buying a commercial license without having a trail first, or the trial at least did not show up in my records.

Needless to say this was a very pleasant surprise and the first real validation.

After releasing Secret Managers for Bamboo in the Atlassian Marketplace, I did not know what to expect. However, I chose to build plugins in an existing ecosystem because I figured that would save me from having to implement stuff like billing, and I could lean on an existing distribution channel (the downside of course is that Atlassian takes 30% of your revenue 😬) .

Turns out this was a good guess. As soon as the plugin was released, I saw 5 to 10 daily visitors on the plugin marketplace page (Atlassian allows linking google analytics). It only took 6 days before the first two evaluations appeared in the Marketplace vendor reporting page. This meant two companies deemed the page interesting enough to take the effort to install the plugin on their server.

An evaluation grants them a 30 day free trial, which can be extended up to 90 days. As such, even if these evaluations would ever turn out to become customers, revenue was not to be expected before the end of May.

The initial version of the Secret Managers for Bamboo plugin took 3 weeks to build, including learning the Atlassian SDK, Bamboo API's and brushing up on my Java skills, as this is the first time I wrote an add-in for an Atlassian product.

The MVP demonstrates the core functionality: connect to an external secret manager and have secrets ad-hoc resolved during a build. I chose to support only Azure Key Vault and HashiCorp vault, because these were the secret managers I knew about at that point.

I started working on the MVP in the 2nd week of January, and had set the deadline at January 31 for submission to the Marketplace for review (to prevent myself from initially building too much). As part of this submission, I also had to prepare screenshots, marketing texts and a documentation website (available at https://windtunnel.io/products/smb). The review itself took over 10 working days but luckily did not result in any remarks 🥳, so the plugin was eventually released mid-February at https://marketplace.atlassian.com/apps/1221965/secret-managers-for-bamboo?hosting=server&tab=overview.

"Let's do something new", I said to myself at the start of 2020.
What that something would be was unclear at that point, but it came with quite some restrictions from the get-go: flexible working regime, working from home, solo-founder, bootstrapped, focus on technology... These restrictions are a consequence of my personal situation and motivation, and any idea or venture would have to tick all the boxes to be considered.

I eventually settled on building plugins for the Atlassian product suite. I knew the tooling very well from an end-user perspective, and topics like productivity, software craftsmanship, ALM, CI/CD and DevOps are near and dear to my heart. My assumption was there had to be a lot of uncharted territory in the Atlassian plugin ecosystem, and the Atlassian Marketplace would give me quick access to a broad range of (international) companies.

I had the domain name windtunnel.io laying around, so I built some branding around that, including the company name, a spartan landing page and a 5-minute effort logo.

I embarked on the journey of building my first plugin for Atlassian Bamboo: Secret Managers for Bamboo.