As I am always looking for better ways of working (aha!), I sometimes like to step back from an established process and analyse what I’m actually doing.
As I’ve become more confident and capable with my reporting, I’ve been challenging myself earlier in the processes I work with. The ‘big bad’ is usually the UK data sets, since the company I work in is UK based, the majority of their data is also UK based and HUGE.
Which brings me to my example for today. I needed to open several large Excel files and get a particular set of data out of each one and into another file that produces some calculations and outputs the results in a new format. So, with my more challenging attitude, I went straight in for the UK and Excel seized up and wouldn’t play nice with me. I quit my open files and took a moment to think.
Lately I had been exporting the data I needed for my Monday reporting into a small snapshot file with the format I needed and nothing else. I would then open the snapshot with my calculation file and copy the formatted data directly in there. It’s been working really well and has also given me a small, clean data set that I can share easily if needed (which I have). I’d been doing that to overcome the very same problem I was having with my task today, two large Excel files just not working well when open together at the same time.
And I already had another output further down the process that resembled a snapshot of the data set that I needed, so I swapped over a couple of steps in the process and did the snapshot first.
This wasn’t a magic bullet by any measure, as I now had to set up a new process to export the data from all the outputs into this smaller snapshot file, including a whole lot of checks to make sure I hadn’t gone wrong somewhere. Just setting up the checks correctly is a pretty lengthy process, just when you think there can be no way it could go wrong, you think of another way it could go wrong.
I spent the rest of the day setting up the snapshot file and doing the checks, but it worked out really well. My checks were better than before and I had one of my outputs ready much sooner than it usually would have been. I also found that it was easier to bring out inconsistencies because it was a simpler view than before.
The only downside is that I still have to go back and do the process step I skipped before for this one, but now that I know I have a full data set in this snapshot and I just need to match it, my job will be a whole lot easier.
So by stepping back and looking at other processes I’d changed, I realised that I could swap a couple of steps and simplify what had been a pretty convoluted task. Now I just need to automate it and that’s one more job off my list!