Re: Does everyone need to build?
I think this discussion is probably framed more at large businesses, where "inhouse IT team" may mean hundreds of people. But in SMBs where I've spent my career developing, much of it as the only developer, the description of "non income producing" is just not adequate. Every dollar the company doesn't have spend buying something (or spending on expensive outside help) ends up on that same bottom line. If you write systems which make it possible to hire less people thru efficiency, produce better products thru timely information when it is needed, and make it easier for your customers to do business with you than someone else, how is that non income producing?
Take a look at the ecosystem around the IBM midrange server now called i5 (formally AS400). You almost never see that environment without a "build" mindset behind it. The platform is just so productive, you don't need many inhouse people to achieve amazing systems. Doesn't mean you may not have a ERP system to integrate around, most do. But they usually have the talent to complete the integration with the business, fill in gaps and keep support cost to an obsolute minimum.
This article hits very close to home at my current company, a global business with 20+ business units around the world. Most were acquired, so for years they had decentralized model, each unit had their own system solutions. They have tried twice to implement common ERP at units, once in late 90's with BPCS and currently with Microsoft's AX. But they never had the inhouse talent to pull it off, used outsiders who got it "running" enough to collect their money and then left. No continuous improvement cycle was in place to keep integrating it thru the business. Only a couple of sites made it work, and that was because they had inhouse people who took ownership, could actually write code and integrate.
That is how I got where I am now, was brought in as contractor on the first BPCS effort at this unit. I spent first few years as contactor, filling gaps and making the software support the business, instead of other way around. Eventually I came inhouse with them and continued those efforts. We are now so effective that they can't even put AX in here, you can't build a business case. The only other unit in that category is a unit with $20 million SAP install with inhouse support team. Every other unit did not have inhouse talent, and every one is failing again with AX. Mgmt has actually placed further rollout of AX on hold, they have not come close to an effective ontime, on budget implementation yet.
As Coverlet says, talent is the key. And the only to make sure you have it is to hire it and keep it. Whether it is cost effective to have one hundred people like that is a whole different discussion.