This is a reproduction of a guest blog I wrote for Harvest at:
How long before the robots attack?
While getting drinks one Friday after work, some friends and I got round to discussing whether it would be such a bad thing if robots took over the world. We didn’t come up with any answers, but the discussion did leave an impression on me. It got me wondering whether my job as a project manager at White October could eventually be fully automated.
I wanted to go a bit deeper than just speculation and really try to find out whether a robot could do my job. To do this I needed to understand what it was I actually did every day and whether any of these tasks could be automated or at the very least delegated. Laboriously logging all this information felt like it would be a step too far for such an idle fancy until I realized that I was already doing this very thing as part of my job.
Timesheets are often seen as a chore, a necessity for agencies, freelancers, and professionals, something we have to do in order to get paid but not a part of our job in which we enthusiastically engage.
However, timesheets give you a minute-by-minute account of your day. There is so much potentially rich metadata and knowledge that we overlook, focussing instead on what we bill. Timesheets can give us insight into what we do.
So I started to add my own metadata to my Harvest timesheets. In each entry description I would add my own code depending on whether I had enjoyed the task and whether I felt it had potential for automation or it could be delegated.
My code was very simple:
- (M|+) – meant a manual task that was positive
- (D|=) – meant a task that could be delegated and was emotionally neutral
- (A|-) – referred to a task that could be automated and was emotionally negative
Various combinations of the above gave me a rich set of data to work with.
I was then able to use Harvest’s detailed export function to get all of my data in one place. Then using simple COUNTIF statements (e.g. =COUNTIF(F2:F19,”*M|*”), I could aggregate and analyze the data I had built up over time.
This opened up new possibilities for how I used my timesheets. In theory I could track my mood, the type of work I was doing, and even my stress level – anything I could quantify could then be tracked.
I found a number of interesting things from my subjective analysis but here are four key findings:
- I like my job – It’s reassuring to know that task by task, I’m broadly happy in what I do. I had assumed this before but this was a neat way of confirming it and a personal confidence boost. Maybe I shouldn’t be worrying about the robots and just enjoy my job.
- I could automate and delegate a whole bunch of stuff – Especially the things I don’t like doing. If I could automate these tasks away surely I’d be happier in my job.
- I could spend more time doing what I like – If I automated and delegated everything that I could then (based on my projections) I’d end up spending more time doing the tasks I enjoyed. My task breakdown might look something more like this:
Automating with Harvest
Now that I had a better understanding of what I did on a day-to-day basis and how it made me feel, I began to explore ways of automating the tasks I didn’t enjoy. Here are a couple of examples of where this worked for me:
Recurring appointments and timesheets
I have a few appointments that occur regularly every week. They occur at the same time and I feel the same way about them every week. Putting these entries into Harvest feels a bit like a waste of my time. It’s non-billable, internal stuff but I need to add them in or my weekly time won’t add up.
However, there’s a Zapier integration that can scan my Google calendar for these recurring appointments and automatically add them in as new timesheet entries.
The time this saves is maybe only 5-10 minutes per week, so it isn’t amazing in that respect. However, this was a task that I regularly felt negative about, so not having to say to myself, ‘Why am I doing this?’ in a dejected manner for 5-10 minutes each week is of great personal value to me. Potentially I could expand upon this to have all internal meetings and appointments automatically appear in my timesheets and only change them if they run over.
Giving clients access to Harvest
If you can prep your clients by telling them that they are seeing raw timesheets, and thus things may be subject to change, then there’s a great amount of time to be saved by doing this. If the client wants a budget update, then they just go to their site at any time of day. Being transparent with the budget from the start saves you time. The earlier a client can see that a budget is being stretched the more likely it is you can both come to an agreement on how to deal with it. Saving up the bad news and letting the client know that there is no money left a week before launch leaves you both in a difficult position and less likely to find a win-win situation.
You can also augment this with Harvest’s own built-in alerts function. You can trigger alerts to be sent out to you when project budget reaches a certain point. You can use this as a trigger to set up a meeting or just automatically forward the alert to the client so they can check the budget and call you if they are concerned.
Slack and Harvest
Slack and Harvest offer some really interesting possibilities. You only have to do some light googling to find out all the interesting integrations from places like IFTTT, Zapier, and Cloudpipes. I also admire Neverbland, which has started doing Harvest tracking via Slack commands.
The bit that would save me time is to have a weekly report generated so I can see really quickly where time is being spent (or not logged). Essentially having the information pushed to me in Slack would be really helpful, as opposed to having to pull it out, which I invariably forget to do.
I hope this gives you some insight into how Harvest can be used for more than just time tracking and how the use of automation and integration tools can save you time and take your use of Harvest to a whole new level.