Worked with 3PL vendor on setting up real-time postage pricing
It turns out none of their other Shopify customers had ever asked before, so they were all just charging flat shipping fees and either eating the cost or pocketing the difference when the static price differed from the actual postage costs
But I had to upgrade from Shopify’s $80/mo plan to their $400/mo plan!
I should make some of the money back in lower (by 0.2%) credit card fees at this tier
The higher-tier plan also supports better international shipping (customer pays duty fees up front rather than getting caught by surprise later), so it may be worthwhile
We’re still working out some kinks in the integration, but it seems like we’re good enough to get up and running
Started gathering documents for an R&D credit study
I can potentially get $20-35k in tax credits TinyPilot qualified for through R&D, but it seems like it’s going to be several hours of unpleasant paperwork
Interviewed a new accountant
Got together documents for 2022 tax filing
Sent a short-term roadmap to the team
Worked with hardware partners to iterate on case design
One 1:1
Reached out to most of the contacts I received from my last retro
Worked with hardware partners on case improvements
Customer support
Realized email notifications for our support forum had been broken for a week
Sales
Talked with a customer about sponsoring TinyPilot to add a new feature
Increased base price by $30 ($399 -> $429)
I’m trying to slow down sales until the fulfillment team has a chance to catch up and build up a better inventory buffer
I actually mistakenly increased the price by $100 and customers still purchased, but it’s unclear if that’s meaningful because they only saw the +$100 price at checkout. Though it does make me wonder whether the price should be a lot higher.
I partially refunded those customers because the website showed them a lower price when they decided to purchase
IT
Upgraded Balena Etcher on office desktop
This turned out to be more complicated than I expected, and I made it unusable for a day
Some of the paths and flags had changed, and I hadn’t tested correctly on my VM
Wrote instructions for troubleshooting our new router
This was one of the first projects I worked on after I went full-time as a founder
I wanted a way for users to be able to test Zestful without signing up for a RapidAPI account, but I didn’t want them to be able to just skip ever paying
I implemented a proxy that mimicked the semantics of the real API, except that it limited each IP address to 30 requests per day
The original implementation was in Python 2.7 AppEngine and persisted data in Google Cloud Datastore
I got pinged because Google is sunsetting Python 2.7 AppEngine
Revisiting this project, all that infrastructure is way too complicated, so I’m in the process of reimplementing it as a single Go binary that just keeps everything in memory
Everyone’s quotas will reset on every deploy, but that’s not a big deal
If I ever want to do better, it will be easy to layer in SQLite
I was all excited to see how much faster I could stand this up than I could five years ago, but it looks like I’m not that much faster
Five years ago, based on git commits, I’m guessing it took me 2-3 full days
I’m two hours in, and it’s looking like it’ll take another two hours
I was expecting it to be like 2 hours vs. 2 weeks, but I’m not that much better