Life, distilled by Will Critchlow

Wow. That’s a boring title. Could I have come up with a worse title?

You already know I’m a geek. Well, when I read Wil Reynolds’ recent post on issues with various bonus structures he’s tried, this was the first thing I thought about. You’ll want to read Wil’s post before continuing here in order for anything to make sense.

At university, I studied game theory (my thesis was on combinatorial auction theory - a brilliant, often NP-complete branch of auction theory). I try really hard not to apply it to everyone I meet all the time.

Having said that, I’ve had to learn how to act in social situations, and game theory is actually pretty great for this (see a beautiful mind).

In many situations, the key is to view life as a series of events taking place over time rather than one-off experiments (it is this that “fixes” the prisoners’ dilemma). And it was to this that I turned when reading Wil’s post.

(Before I get into the meat of my thoughts, note that I don’t believe we have compensation completely right at Distilled - but that’s a post for another day). Here are the game theory issues that sprang to mind for me when I read Wil’s post:

• Anything expected is part of compensation - you can’t decouple these things. If people know they are getting a bonus (and know the approximate size of it from the quarterly updates) then they will treat it just like salary. To the extent that money is what motivates them, they will choose where to work based on total compensation (with some risk premium thrown in)
• You get what you reward. If you reward longevity, that’s all you’ll get. There’s no particular reason to think that rewarding longevity will result in more effort / output / productivity except to the extent that you keep good people around
• Company performance-based bonuses (shared according to a pre-defined formula, rather than on personal performance) are (once people treat them as part of compensation, see above) ways of giving people pay cuts in poor years. This is how investment banks work. Bankers wouldn’t work for their base salary(*), and bonuses form part of compensation. It’s one of the few fields where pay cuts from one year to the next don’t (always) cause deep unhappiness / exodus.

(*) ignoring all the 1% debate for now - that’s another post

It seems like some of this can be fixed by tying bonuses to individual performance, but there are issues here too:

• Bonuses based on personal performance reward that particular metric, but tend to encourage selfish behaviour which may result in lower overall performance (damnit)
• It’s hard to couple meaningful bonuses to personal performance in most areas. To get enough upside to cause real behaviour change, you have to push such significant risk of downside that the individual becomes better off running their own thing and taking 100% risk
• If it’s not tied to overall company performance (and either isn’t purely financially-related or doesn’t clawback in underperforming areas) you run the risk of large individual bonuses in times when the company-wide prospects look relatively bleak

Ignoring the theory, the practicalities of determining appropriate reward systems are also complicated by the fact that the people in charge of small businesses skew towards higher risk / faster change / less certainty than average. I find it can be an area where it’s hard to put yourself in others’ shoes.

I’m going to be doing loads more thinking about this. Thank you Wil for being so amazingly open in your post and I hope to have more fruitful discussions about this stuff.

[Sidenote - I wrote this post today on the Distilled blog and the kind of reasoning I refer to above is the stuff I was talking about in the “heads-up” part of the leadership section.]

I have been hugely impressed by Launchrock. I initially thought it sounded like the perfect example of Feature Not A Company but I’ve been convinced that if what they have is a feature, it’s one that should be present in every email provider’s toolkit.

In particular, I’d like Mailchimp to buy them please.

Yesterday, we ran a “ship-a-thon” at Distilled (like a hackday, but designed to be more inclusive of everyone who isn’t a developer / designer - the idea being that we all spent the day shipping stuff to make the company better).

I worked with 4 colleagues to ship an internal alpha launch of a product we are calling “DistilledU” - a learning platform for internet marketing. The first thing we shipped (in good “lean startup” methodology style) was a pre-launch landing page. Obviously we could have built this pretty easily on our own, but I was very impressed by how easy Launchrock made it to create a great-looking page with all kinds of embedded social sharing etc. and a 43% conversion rate.

Yes. You read that right. So far we have had getting on for 1,000 sign-ups at a 43% conversion rate.

I was amazed by the tweets it generated and the great feedback loop that resulted in ever-more signups. When you combine that with the analytics in the back-end, really the only thing missing is integration with Mailchimp. So - to Mailchimp - I don’t care if it’s a feature, I’d like it integrated - even if you have to buy it.

(If you want to see more about our vision for DistilledU, you can read more about our efforts here and of course you can sign up to stay up to date here).