Principle-based decision making

Written on May 10th, 2008 by phirate

When building a site like Entrecard, there are dozens, sometimes hundreds of decisions, big and small, that need to be made every day. In general, no one decision is ever the difference between life or death, but the wrong approach will lead you into groups and chains of choices which can be.

It is important, before you get to the point where you’re making these decisions, to try and evolve a larger guiding set of principles to help ensure that – on average – the decisions you make are consistent with your vision of where you want the site to be. Some of these are primarily ethical choices (”I will not lie to my users”), some of them are practical (”I will not implement anything that needs my constant attention”), and of course they will change as time passes and you gain experience. The important thing is simply that they are in place, helping prevent you from getting lost in a sea of detail decisions and losing sight of the bigger picture.

Lets take a look at some of the core principles we try to live up to with the decisions at Entrecard, and some of the side effects.

We will not lie to our users

This is a fascinating one, mostly because the same positives and negatives exist with this as you get in any interaction. Deceiving people is, in general, counter-productive, and you can win a great deal of trust in a market in which players are generally deceptive, simply by being aggressively honest – to the good of both you, and your clients.

I’ve been delighted with how well we’ve managed this overall, especially given a number of tricky scenarios – for example, what do you do when the situation demands an immediate response, but you don’t know the facts? this is often a challenging situation that requires some quick thinking, but the important thing is that the quick thinking should always be with the interests of your users at heart.

Another area that qualifies as a semi-exception is security. Security is as much about deception as it is about solid design, defense in depth is fundamental to solid security and revealing information is counter-productive. At the same time, there is a difference between deception *for* security, and deception *about* security. The first is critical, the second to be avoided at all costs. An example might be the difference between a security system that lets a security flagged user continue to act while silently ignoring their actions, vs saying “Our system is completely secure” when you’re looking at an email with a known security flaw.

We will try and avoid making unnecessary promises

This is a key one, a generalization of the “Don’t announce upcoming features” concept. The problem with an upcoming feature announcement, or indeed any promise, is that you are expected to fulfill it. Failing to do so, no matter how good your reasons, impacts on your trust. In addition, when you have relatively few promises, it is easier to fulfill them in a timely fashion, and thus people can generally come to trust that when you say you’ll do something, it will be done.

This one is hard to keep. It’s so tempting to announce stuff that’s coming up, especially if you’re feeling a little bit like you haven’t released anything interesting recently. It’s also difficult to avoid promising something when someone is having trouble – I’m particularly bad at this, I say I’ll deal with something, and 80% of the time I do, the other 20% of the time something more urgent turns up and it slips my mind. Bad phi.

The bloggers blog is their space, we have no right to say what should go there

I throw this in as an example of something key to our particular niche. While we do have rules about what kind of site we’re willing to have an Entrecard on – something that is necessary to the functioning of the system – we try desperately to avoid any further imposition on the hosting blog. We consider the widget to be a useful guest, not something that is there by right.

This influences a number of our smaller areas of policy, for example we always give our users ways to cancel an ad that is currently showing – the space it’s showing in is on their blog, and therefore it’s theirs to decide what should go or stay.

If you’re currently running, or looking at building, a site – especially a community site – I highly recommend you take some time to sit and think your way through stuff like that. It’ll save headaches later, and it’ll give a consistency to your values that your long-time users will come to understand, making it easier for them to predict your actions in uncertain circumstances.

It’s not about you or the company, it’s about the people you’re interacting with who are taking part of their lives and putting it in your care.

12 Responses to “Principle-based decision making”

Leave a Reply