Office Not Using Some of My Tools
Posted by AmustyG@reddit | ExperiencedDevs | View on Reddit | 15 comments
Made a slight career shift and have been at this new job for about 6 weeks now. I was previously at a big fortune 100 company where anything I did would have to go through many layers and approvals so I never saw my work used in real time. This new opportunity allows me to create what I want, for as long as I want, and implement it when I feel it's ready. I'm the first "technical" person they've hired on this small team of 12. So far I have 1 tool that got heavily implemented and i'm constantly updating/making changes to per requests from users in the office, but the other 2 tools I spent time on have not been implemented at all.
I really want to show them what i'm capable of but am worried that it seems like im wasting time on things that arent even used. It's easy for everyone to say "we want this automated" but then when I sit down to ask them it seems like they can't think of a single thing they'd like.
Is it normal for this to happen? Is it a bad look to have this many failed projects already? Being a 1 man show for this, how can I have a higher success rate with my tools being implemented?
Soggy_Grapefruit9418@reddit
The fact that one tool already became heavily used after only 6 weeks is honestly a very strong signal that you’re providing value.
WhateverHowever1337@reddit
Pick a book or any other good resource on user research, and learn how to actually gauge and interpret what people need (hint: they usually don't know what they need, or their bottlenecks) so you will have to study/interview them well beyond just asking what they want for you to do.
AmustyG@reddit (OP)
I really like this idea, it's definitely a skill im lacking. Any recommendations for me to read?
StevenJOwens@reddit
One classic is Nielsen's "Usability Engineering". It's very readable and a good start.
Fish150@reddit
The best automation you can build is no automation.
If you can suggest a change in process that turns a 10 step process into a 5 step process, you get a free 50% speed up. Or maybe a 99% speedup if the slowness was one of the eliminated steps.
And now you just eliminated a bunch of future technical debt.
AlexFromOmaha@reddit
I used to run an R&D lab for a telephony company, and I had a team to do this sort of work.
Shipping a dud happens, but two-thirds isn't a great ratio, especially since you're at the phase where the work is easiest. I'm also kinda suspicious about the timeframe. A new tool every two weeks, even allowing for onboarding? Are you making things you think you'd like in their shoes, or do you know what their actual problems are? I'm not saying ask them for solutions (they'll offer you some without asking, don't worry), but definitely ask them for real problems.
I wouldn't bother with things like A/B testing or usage tracking at your scale. You lean back, say "Hey Brian, is that change helpful?" and iterate.
For the first few months, piecemeal will be fine. It'll help build relationships and trust. Once you settle in a bit, you want to get in touch with the Powers That Be, get real familiar with the company-wide KPIs, and make a concrete bet on how some of them relate and what you can do to drive it. In our case, we focused mostly on internal tooling for the non-technical white collar workforce because our bet was that eNPS and time-to-implementation were actually the same problem, and the automation was less about doing the work and more about clearly defining the inputs and outputs from each step and making it easier for each team to know when they were done without the opportunity for office politics to derail handoffs. Yours probably won't look anything like that, partly because of the size of the company, but partly because their biggest needs are probably external-facing. Vendors, customers, maybe banks/investors, maybe onboarding employees.
AmustyG@reddit (OP)
This was super helpful thank you! A bit of context, it's a small advisory firm in real estate. They have no automation implemented AT ALL.
First 3 projects were things that they wanted before I ever got hired. I have 1 big project that's much bigger in scope and will take time though. The 3 that I put out were pretty small things and just simple dashboards that adjust based on uploaded data. I was putting in 65 hour weeks to start strong and try to get started immediately. They ended up loving 1 of them and use it regularly, other 2 immediately dropped. The difference (in my opinion), was that I was regularly updating/bringing it up in meetings to show progress and new features to get their input for that 1 success.
My question for you is, aside from shadowing, how can I get better at finding things to build? What's the easiest way to implement things into people's workflows that they've had the same for so long? With nobody else being technical, what can I do better to let them know that some projects are just going to take much longer especially because i'm 1 person?
gefahr@reddit
After shadowing: If you're given the latitude and time, and you intend to be there for awhile, taking the time to understand the business and everyone's job functions is going to be valuable for you.
And then get an understanding of how the person who signs your check feels about these functions. You may find their view differs from the persons doing the jobs today.
Where you find a disconnect, you ask your boss how to reconcile. Could have been a disconnect on their part, could have been they want things done differently.
AmustyG@reddit (OP)
I really really appreciate both of your replies and would love to DM you if thats okay
gefahr@reddit
Np go for it
Consistent_Photo5064@reddit
Yes, it’s part of your job to detect what they actually need.
The easier way IMO is to spend time with them, detecting actual pain points and solving them instead of trying to come up with a tool.
AmustyG@reddit (OP)
Completely agree and have started to slowly do that! Do you think I should be expecting 100% implementation for the first few months or is that unreasonable?
gefahr@reddit
Unreasonable. These people have ingrained workflows likely. Shadow them as suggested above. It'll help you discover requirements and what's actually going to move the needle for them, and give you the chance to catch things they do that they'd forget to tell you about.
Importantly, it'll also build trust with them that you care to understand their pain points. The last thing you want to be seen as right now is someone who will force tools on them without fully understanding their problems.
If you want organic adoption, you need to be seen as empathetic and genuinely interested in understanding how you can improve their day to day.
Altruistic-Bat-9070@reddit
You are talking about changing processes. You have to find out what processes they think are bad, what processes are actually bad, then find the ones that overlap on the venn diagram and make solutions for those things. Those are the only things that will be successful. Changing human behaviour is something is bad but good enough and a habit is impossible
SoggyGrayDuck@reddit
Damn you're so lucky. I had that for the last 7+ years and now in the corporate cog machine.
One thing to keep in mind is that you'll slowly lose those processes. You'll learn a TON but depending on the job market might have difficulty getting back into the traditional corporate culture.
It was interesting because I could list all these high end tools, was the official admin of multiple AWS environments and the only person who would stand up new services for people to use. Architect stuff BUT very little was built dynamically, we didn't have a dev environment in one of the jobs so it was pointless.
I just started a new tech lead job that hired me for the background and is willing to let me learn the process and work with me learning how to properly interact with multiple environments, use git and pipelines properly and etc.
Side comment: our industry really needs to standardize. You say pipeline to 10 people and ask them to define it and get 10 different answers. I feel for the people running the interviews and get why they're now 5+ rounds (or contract to hire where they simply test you out) but there has to be a better way.
Other industries have a union that helps define all this stuff but we don't have anything like that.