This is Your Tech Leader Speaking
This coming Friday 21st May at 8am (Sydney) I’ll be talking with Gretchen Scott at Cloudflare TV on their show “This is your Tech Leader Speaking”.
Come join our conversation, with topics ranging from my journey from Front-End Developer to CTO at Digivizer, tooling to bring a business together in a pandemic, to hiring in a hybrid workplace, with the show likely interspersed by my kids visiting me before going to school!
Bring your cup of coffee over to https://cloudflare.tv/live and I’ll see you Friday!
The episode is still up: APAC Heritage Month: This is Your Tech Leader Speaking on Cloudflare TV. Notes on what we actually covered, written up later, below.
What we talked about
Digivizer is two businesses stapled together, a SaaS product and an agency, which sounds messy but is actually pretty useful. The agency folks are hitting real customer problems every week, so the product team never gets to drift off building metrics nobody asked for.
Most of the conversation ended up being about the year we’d just had. Our customers had moved a long way from the Microsoft and Lenovo end of town toward smaller Australian businesses, a lot of them on Shopify or some blend of retail and online. The gap we closed was website analytics. Folks could see exactly what they’d spent on paid, and then nothing at all about how many sales came out the other side. Joining that up end to end was the missing piece, and it quietly became the thing that won demos.
On remote work, engineering barely noticed. We were already spread between Sydney, Bathurst and Melbourne and already collaborating online, so not much changed. The teams who’d always sat together at desks felt it much harder. The real casualty was onboarding, because the informal version of onboarding is sitting near people and pairing constantly, and that just evaporated overnight.
We reopened the office in early 2021 and almost nobody came in! Turned out what people wanted it for was workshops: book a room, stand at a whiteboard, go home again. So when the building changed hands we let the permanent office go, took shared space about a day a week, and put on a monthly all-hands with food and drinks. Being around people matters even when the work doesn’t strictly require it.
Then a good chunk on ensemble programming, which is pair programming taken up to team size. VS Code’s collaboration plugin does the heavy lifting, and pleasingly it lets me keep my vim bindings while everyone else types like a normal person. We used it to onboard two juniors we’d hired in April and June 2020, mid-pandemic, and the thing that made it click was picking genuinely non-trivial problems. Trivial ones just dispirit people. Two seniors and two juniors on one problem, seniors driving until they hit a fork, then opening the trade-off up to the floor. Rising tide, all boats: the juniors ask the clarifying questions, and the seniors have to go back and check what they’d assumed. Three or four hours in an afternoon is about my limit before I’m wrecked. Both juniors went permanent, so two for two.
Hard agree on hiring juniors, by the way. Skip them for four or five years and you’ll wonder where your mid and senior pipeline went. One of ours came from CS and the other from a science background (analytical, coded at uni, loved it), and that difference in how they come at a problem is the whole point. Stack a team with similar seniors and you get similar output and similar blind spots. The trait worth recruiting for is curiosity. If you can find folks who are genuinely curious about how things work and who’ve already proven they can learn to code, we can teach them the rest.
There’s a nice university angle too. Our software ended up being used in a digital marketing course at a Sydney uni, and one marketing intern found us that way and came looking. Beats a job ad.
And the obvious objection, which is a fair one: aren’t you paying two people to solve one problem? Yes, and on paper it looks slower. It stops looking slower once you count true cost of ownership, the debugging, the changed requirements, the production issues. Solving it early is cheaper than solving it in prod, and the customer gets a better product either way. Writing juniors off as too hard is short term thinking the moment you’re past being a three person startup.