ROSL84 — Reflections on startup life, week 84
I’m not sure if it’s just the winter setting in — cold days, not getting light until late and dark early, but things continue to just feel hard at the moment. It’s a real slog. One thing I’m doing to break that up is to try and introduce a bit more company into my day.
Over the last few weeks I’ve had a regular weekly appointment with Howard who is working with us as an Intern at the moment and is building new connectors for us. It’s been refreshing to sit down and work with him outside my little home office, meeting at Inspire9, a local Melbourne co-working place. The guys (and girls) at Inspire9 have been great and I think that I’ll be spending some more time there just to keep the human contact up.
I think one reason that this wasn’t such a problem in the past was that I spent a lot more time visiting and selling Tribalytic to agencies in Melbourne, so that kept me on the road more. With Trunk.ly it’s totally online and we don’t drive user numbers by meeting with people (at least not significantly enough to warrant making it the main user acquisition channel).
Another lesson learnt this last week is the value of maintenance. As the site gets bigger and bigger, the need to invest more and more time in maintenance grows. We’re beginning to gather a list of tasks that need to happen on a regular basis to ensure things keep humming along, for example the search index has to be re-optimised regularly (something we’ve neglected until recently), backups should be checked and so forth.
Things change. Early on we were rushing features in with a break-neck pace, doing something is better than it working effectively longer term. When you have 50 users, you shouldn’t worry too much about 50,000. Gradually over-time we’ve moved away from this approach which has also slowed us in some respects, but is paying off too. Significant changes are quicker with better test coverage, the structure of the code is improving and some integrations are becoming much simpler (some we still have to do the refactoring first to make it easy).
With this new found maturity, tools are increasingly important. For the first time in two years we’ve now invested in an IDE for Python / Django, selecting PyCharm. I think the biggest wrap I can give is that it was a “no brainer”. It’s a pleasure to use and after the free 30 days, I know that my code quality had improved significantly just because of the constant gentle guidance towards improvements and standards compliance. It makes me weep now to open some of my earliest efforts into PyCharm and see all the errors and warnings fire up!
So on we go.
Goals this week? Having finally nailed the performance issues, we’re now working to dramatically improve the group feature and make it even more usable and useful.