| You pretty much exactly described my job. I started last year and inherited a horrific ERP system (and ancillary code for things like production scanners). They realised they needed an ERP but they where unlucky in who they got to build it, things I take for granted as simple things to do (an interactive timeline showing job start stop, per user/department in d3) has had a massive impact on how they run things (it surfaced issues around job tear down/turnover). Efficiency against estimate tracking (it's a table that's red/green if over under) in real time again made a huge difference since it surfaced issues with optimistic estimates from some operators and pessimistic estimates from others). That matters because it throws off capacity planning and pricing on jobs (we go in too high, too low, too high we might lose the job, too low we take the hit on profits). All vanilla stuff really. None of this stuff is particularly difficult (refactoring parts of the existing shitpile are but that's because it's one gigantic pile of mud). The fundamental problem is that there was a disconnect between the business and the specification of what they wanted, they didn't know enough to know what they could have asked for and the original devs didn't think about how they could improve on the existing business flow at all (I'm not even convinced they even considered the business in a lot of cases..). In terms of users, about the same 50 on a typical day and I have a similar level of control, from the hardware we buy to the OS to the framework to the programming languages to well everything, it's an awesome feeling of freedom to be able to pick the correct tool for the job. Technology applied properly can still have a multiplicative effect for a lot of SME's, there is still a vast amount of low-hanging fruit out there. |