Forging Ahead 2019!

Lessons learned in 2018

Many of the valuable lessons we shared in the past (Back to school: lessons learned this summer, In the year ahead, Analysis of Runtime Behavior of Boutique Software-as-a-Service) were focused on the product and reaching the audience for it.

2018 was different. We received an enormous of amount of feedback from customers and users alike. There are enough features to fill a product roadmap for the next 5 years.

2018 was different. Inquiries and support requests spiked. By September, we started to use Zendesk.

2018 was different. The team was so heads down on replying to user questions, keeping the pace on a quick response rate, that we neglected to catch up with the rapidly increasing HTTP traffic. By December the infrastructure started to reply with dreadful "502 Bad Gateway" intermittently. These were only momentary solved by heroic efforts (and a timely government shutdown).

The lessons from 2018 were focused on people, processes and management:

• Trying to keep response time to within an hour is hard, especially when you are in San Francisco and have customers on the East Coast starting by 7am their time.
• Even twenty tickets a day, every single day of the week can swallow the whole time a team member is available. Measuring here is critical because staff costs scales linearly with the number of new tickets, and support will impact the profitability of a SaaS business as it grows.
• Even if you are lucky to recruit talented people, the amount of knowledge that needs to be transferred in a very short period of time is enormous.
• The bandwidth of the executive team quickly becomes a limiting factor. You can read about it (The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers) but you won't fully grasp it until you experience it first hand.

We are strong advocate of a World-wide distributed team at DjaoDjin but solving these problems required to talk them through face-to-face. We thus gather an all-hands in the office at Punspace Chang Mai for a month in order to come up with a plan of actions.

Roadmap

Since the beginnings, the source code has been open sourced, but the features planning process has been somewhat "agile". Partners and customers require a better understanding of what is coming down the pipe so they can plan accordingly. DjaoDjin ethos is about transparency and the Welcome Spirit^tm. It was time to be more transparent about the 2019 roadmap.

At the off-site (or is it "on-site" since the team is distributed the rest of the year?), we put together every issue, feature request, annoying quirks and other piece of feedback as a sticker on the board, then organized them into buckets (see photo). The buckets were finally prioritized to help partners and customers grow their business with DjaoDjin. Here the high-level schedule DjaoDjin's team is committing to:

• Feb 15th 2019 - Upgrade default workflow pages to bootstrap4 and vue.js
• Mar/Apr'19 - Themes management and versioning
• May'19 - Deployments in EU region
• Jun'19 - Make audit logs searchable
• Summer'19 - New features in the billing workflow
• Q4'19 - Third party integrations

First, we have been working on a major upgrade of the default workflow pages to bootstrap4 and vue.js for the past three months. We are testing and finalizing the branch at this point.

We have been upgrading workflows for all customers as soon as new features are available. The intent was to make it as painlessly as possible for all customers to always run the latest and most secure version of the dashboard and workflows. The technical decisions we have taken to achieve that goal did not interact well with custom themes, resulting in third-party developers confusion and ultimately broken flows. We are going to step back and fix that process before moving forward again.

To celebrate the one year anniversary of GDPR coming into effect, we will be deploying applications for European customers in AWS Europe regions.

DjaoApp generates quite a few of events (user registered, user subscribed, user interacted with a specific page, etc.) that are logged for audit purposes (the most common use being chargebacks). These logs are not directly viewable nor searchable by DjaoDjin's customers. This is a feature that has been requested by many support teams. We listened. We will be fully focused on getting audit logs searchable by the beginning of the summer.

A significant number of customers are invoicing offline instead of asking for a credit card payment online. A few users are starting to build products build of an IoT device talking to a subscription-based service. This requires shipping a device, re-ordering a lost device, etc. We will make significant upgrades to the billing workflow in order to accommodate innovative business models over the summer.

One of the key advantages of DjaoApp is that you can run a SaaS business out-of-box, right after signing up. A few customers are growing their support team, as well as integrating with sophisticated marketing tools. Two open browser tabs, or download from djaodjin / upload to third party tool is becoming cumbersome. We will have a third-party connectors program in place by the Thanksgiving time frame.

If you have any third-party tools you use and would really like to see connected with your accounts and billing workflows, please let us know.

To a wonderful and exciting year!

More to read

More business lessons we learned running a SaaS application hosting platform are also available on the DjaoDjin blog. For our fellow engineers, there are in-depth technical posts available.