4 ms·
Over the last 2 decades I have switched from development to business owner/management/marketing/sales and back 3 times. Right now, I am doing a rewrite of my o
by ToJans 6y ago
Over the last 2 decades I have switched from development to business owner/management/marketing/sales and back 3 times.
Right now, I am doing a rewrite of my online 3d engine for consumer products, after being active for 3 years as an enterprise consultant for mid and large sized organizations, mostly initiating first roadmaps and identifying how to onboard new clients, reorganizing my client's product and solution portfolios, figuring out/maturing pre-sales tracks etc...
The reason I switch from development to other tracks and back, is that after a few years I feel like I am stuck in an echo chamber, where I tend to overestimate the value of the thing I am doing, and underestimate the importance of what others do.
The biggest disadvantage is that I could have probably retired from a financial POV about a decade ago if I had stayed on the same track, while this might now take another decade, but on the other hand, for me the chase matters more than the catch.
The biggest advantage is that - due to my broad experience - my kind of profile is extremely unique, and I can find a well-paid freelance gig in about a week without to much effort if I need some cash to bridge one of my many mini-sabbaticals.
This might not be for everyone, but if you like to spend time out of your comfort zone, I would strongly advise anyone who tries to understand the bigger picture to switch every n years; it is super valuable to understand the drivers from both sides and makes the job so much more interesting.
Edit: typo
- ChrisMarshallNY 6y agoThanks for that. I’ve found mixing disciplines has been very helpful; especially these days, when I’m the only coder in a startup team. But specialization is still fairly important. I am specialized in Swift (with a minor in PHP). This has been beyond invaluable, in the work I do. To be fair, I am involved in multiple platforms and tech within the Apple ecosystem (like writing stuff to work on all Apple hardware platforms), but that is still a fairly constrained environment. I have found that it’s quite possible to write acceptable software, quickly; especially if we are willing to loop in a great deal of “mystery meat” code, but it’s another matter, entirely, to create awesome code. I once had someone boast to me that they only hire people that can switch languages and environments in a matter of a couple of weeks. Wide and shallow worked 20 years ago. Not so sure it is as effective, these days, as the river is a lot deeper, and runs faster. The stakes for screwups, nowadays, are also pretty terrifying.