There are times when the British Political system puts on a great show. You can tell it’s happening because if you go to The Guardian or BBC News website there’s a big picture of the Prime Minister standing in the House of Commons with a red LIVE button next to it.
When you press LIVE, you’re whisked off to a separate area of the site. You can feel it’s different from the usual static news article view: a bit more modern, a bit more swish. It’s like an extension built onto a period house. Maybe they’ve kept some of the same look and feel but you can see that the windows fit slightly better and there aren’t any cracks in the wall.
Inevitably, this new page features a timeline that auto refreshes every few minutes with a new nugget or tweet. What can I say, I build technology products for a living; I see our democratic system in terms of the web applications that present it to us.
When I see sites like this, I always wonder about the business processes behind the scene. There must be a moment when someone says “right, we’re deploying the timeline view now.” Who decides that? Is there a meeting and a process for making that call? How big a deal is it? Do they need developers on hand, or is it all configuration that can be done by editorial staff? And then, do they have rules on how often then need to send updates through? How much checking goes into each update? Is there a workflow engine that takes it through to an editor to sign off? Being a product manager is exhausting sometimes. You can’t just read a website without spotting all these decisions. “Oh,” I think to myself, “a top border on each panel in the accent colour. That’s nice.” Then I wonder about the meetings and mood-boards that went on before that came to be.
The other thing about product management is the politics. That’s what everyone calls it. “Oh that’s got caught up in the politics” they say, or “We could change the colour of this box but that becomes political.”
Watching the current Parliamentary excitement unfold via various frequently-updating web products, I’m struck by how much of Westminster politics is arguing about whether you have the authority to do something you want to do. “The British people,” they say, “want Brexit,” or “my constituents voted to remain”. And when it’s not a debate about what they’re allowed to do, it’s a debate about procedure.
So much of this comes from speaking on behalf of others. I don’t have to worry about policies on schools or hospitals or what country we should declare war on. Yet I have the equivalent conversations about mandate: “Users don’t like drop-downs, they prefer radio buttons.” Sometimes we call this appealing to a higher authority. You don’t actually bring that authority into the room and ask them; you just tell everyone that’s what the authority wants. In Parliamentary Politics, that appeal is to the constituents: “the people of Britain want us to leave the EU.” In software development the equivalent is something like user research: “Research shows users are less likely to complete a form if it has drop-downs”. The process for testing these assertions — calling someone’s bluff, you could say — is a general election or referendum for politics and A/B testing, user research or analytics for digital products. You can’t carry these out for every single decision, so between these tent-pole moments are a series of small battles, played out at the dispatch box or in the product planning session.
This triggers a series of attempts at proving a mandate. “Drop-downs are an accessibility issue,” or “the style-guide says we should use radio buttons” or “when we A/B tested it, we got a much higher success rate with radio buttons”. Maybe even: “I bumped into the CEO the other day and he said he really hates radio buttons”. (It would be a he, wouldn’t it?) Who has the bigger mandate? Accessibility requirements or the results of the A/B test?
Sometimes the people discussing the different mandates talk on behalf of groups they don’t quite understand: “I heard once that Information Security have an issue with us using radio buttons” or “the lawyers said to me that drop-downs make us non-compliant with GDPR”.
When I interview Product Managers, one question I often ask is: can you tell me about a time you persuaded someone to change their mind. What did you do? The answer varies in phrasing, but the content is always the same: I did the research, I gathered the data I presented it to them clearly, and when they saw the facts laid out like that they changed their mind.
That’s fine. Interviews are a strange heightened reality type-situation. But I can’t help thinking: when has that ever worked in reality? Was your research not disputed by people who disagreed with you who then presented their own contradictory research? Maybe on an unimportant decision that no one really cared about this is possible. But as soon as a decision takes more than two or three email exchanges or lasts for more than one meeting, the stakes get higher. All those fiddly emotions come into play and people get entrenched.
This week, watching Brexit-type things play out in a little box embedded on a News web app has made me think about this. There’s a very human pretense at the heart of these conversations: I want to do this thing for a reason that I’m not going to reveal to you. And I’m going to use whatever is available in my arsenal to prove that I have a mandate to do this thing and you don’t. It’s amazing really that this how we decide public policy — but I guess that explains why society hasn’t made as much legislative progress in certain areas as it should have done.
The real reason we want to do something might be because it’s easier to code. Or because we have hedged investments against the value of the pound. Or maybe just because we said it a few meetings ago and now we fear looking silly.
I don’t think we can change these behaviours. Certainly we can’t in others, but perhaps we can be aware of them in ourselves. And if we are, maybe it will result in different, slightly more honest decision making. After all, watching the government’s latest farce unfold, that hardly seems like a model we’d want to emulate.