Glu Blog

Can we define productivity?

Previously I discussed my recent efforts to measure my productivity, with the view to finding the magic levers that would improve it. The process of measuring various aspects of my life raised a more existential question: What exactly is productivity? Can I objectively say something is productive?

It's work completed, isn't it?

It seemed obvious at first. If I was being productive work was getting done. Therefore the amount of work being completed, or the speed and efficiency of completion, would be the way to assess how productive I'd been.

If you've discussed the issues around estimating task effort you'll see one of the problems with this approach: it's really hard to be accurate. It's also hard to compare one task to another unless they're almost identical. So estimation is sometimes ok if you can maintain a consistent level of pessimism/optimism and average it out over a period. However trying to tweak "productivity" on an intra-week basis means that inaccurate estimates have a bigger impact on my productivity than anything else.

Actually… I meant the ability to complete planned work

This is where my approach to measuring productivity really started to breakdown. It had become constrained to decisions I made at the beginning of each week. Fixing a bug a affecting a customer? That's "unproductive" because it hadn't been planned for. Spending more time implementing a feature than expected because feedback highlighted UX issues? Unproductive. Responding to an alert and improving system availability so you don't get paged again at 3am? Unproductive.

External interrupts, iceberg features. Things happen. Having a definition of productivity held hostage to a preconceived plan was severely limiting. Following it to it's logical conclusion would mean prioritizing work that adhered to the plan over improving the experience of our customers. A terrible outcome for everyone.

It's subjective, and temporal!

The final nail in the coffin of this definition was a moment of enlightenment thanks to both iDoneThis and Retroospect. With reminders of the work I'd completed last month, and how I'd felt about it at the time, there were quite a few occasions where I'd go "Really? That's what I thought?". Time had and interesting way of altering my perception both positively and negatively.

There would be days that felt incredibly productive with an impressive list of work, that when I looked at it a month later seemed to be of questionable value. And conversely I'd have days that felt mediocre because I'd invested more time than I should have fixing some issue, which now felt like it had paid for itself tenfold.

If I can't reliably assess the value of something immediately upon it's completion, how could I possibly predict it prior to starting it?

This had broad ramifications. "Productive" or "unproductive" is not something I can objectively forecast. I can't even reliably assess it after the fact. It's a transient and subjective assessment.

It's the vibe

It all started with a feeling. A feeling of being less effective than I could be. On some level I realized it was a largely subjective measure but I'd been hoping that I'd find some objective output that was correlated with that feeling.

That was the wrong way to look at it.

My sense of productivity had a symbiotic relationship with the nature of how I was spending my time. And how was I spending my time? Thankfully RescueTime had the answers.

The days and weeks where I felt most productive according to iDoneThis and Retroospect, RescueTime agreed by giving my a high "productivity score". So it was just a matter of pulling apart what the RescueTime productivity score was made up of and then doing more of that. And it's pretty simple. Productive time is measured by what app or tab I'm in. Time spent in iTerm coding is highly productive. Time spent on Facebook or Twitter is highly distracting.

The biggest impact on my assessment of productivity was the percentage of time I spent doing specific types of tasks. Productivity wasn't an output that could be measured, it was the nature of a particular task or action.

There's also a potentially subtle aspect to this I want to point out: it wasn't the aggregate amount of time it was the percentage of time.

I finally felt like something to aim for, a needle to try and move. Next up we'll look at how efforts to increase my output were actually sabotaging my productivity.

Glenn Gillen

I'm the founder of Glu. I'm also an investor and advisor to early-stage tech startups. Previously worked at Heroku on the Add-ons Marketplace.

We're currently accepting a limited number of new users for our first product, ContentFocus.