Decision making at scale and without casualties
Have you ever experienced any of the the following?
- Decisions are made by whoever has the most important job title in the room
- Whoever has the most stamina for the disagreement wins out
- Disagreements rumble on and on without coming to a conclusion
- People back channel to campaign for the decision they want
- Decisions you thought had been made are actively undermined by people
- You are in disagreement with someone in another team and don’t know how to progress
If you answered yes to any of those then this article is for you. You’ll have come across the term disagree and commit and probably heard it misused. Disagree and commit doesn’t mean “stop disagreeing with me”. Or even in the worst instance “I commit to disagreeing with you when I’m proven right”. It means that everyone has agreed to proceed with a particular course of action even if it is not the one they would have chosen originally.
It’s worth understanding how large scale misalignments come about. As companies grow it becomes probable that any significant decision will affect a large number of people with vested interests. You get multiple people who are used to being the decision maker who are unable to agree or people from different disciplines (eg. engineering, privacy, product..) with different priorities. This is not a problem in and of itself — but the way we go about tackling it often becomes problematic. To resolve this you have a choice of either:
- Agreeing to disagree. This only works when the decision affects a small number of people. Choose your battles — if the impact on what you’re doing is small then it may not be worth conflict. If the impact is large or people are working at cross purposes then they may wind up spending more time in each other’s way (or worst case undoing each other’s work) than getting things done. This strategy only works for very minor misalignments.
- The team doing the work make the decision and everyone else has to suck it up. This works if the impact is small or there is a high necessity for change. I would argue you still want to broaden the scope of the decision if this impacts other teams. The end result of this is other teams will start working around the team in question because they can’t rely on them to work together. At which point you embrace the most brutal form of Conway’s Law. This is where you’re thinking “but, what about autonomy?”. Surely the team doing the work should have sole agency here. Let me give you an example — imagine a team that wants to go to market to test an idea out but the privacy team or the legal team has concerns about how data is being stored. If you breach privacy laws you can wind up with a big fine. So the short term interest of the team to launch and iterate quickly should be subservient to doing the right thing for the business. (And if you’re thinking this is a very artificial example — I’ve seen teams push forward in this instance and do it anyway. Which makes me sad because this is not in either the interest of the user or the company)
- You continue arguing until everyone agrees that you were right all along. Tempting because, let’s face it, it feels good being righteous. But wrong on so many levels. Jeff Bezos described this in Amazon’s 2016 board letter as ‘“You’ve worn me down” is an awful decision-making process. It’s slow and de-energizing. Go for quick escalation instead — it’s better.’
First of all there are very few black and white answers. There are usually many ways to solve a problem and pretending there is one true way is somewhat arrogant. The other downside of this is when you’ve bludgeoned your foes into submission — I have some bad news for you. They haven’t actually agreed with you and they’re not invested in helping you succeed. In fact they’ll happily sit back and watch you fail because “I told you so” is going to feel so good (but — it really doesn’t).
- You escalate to someone senior to get them to force the issue. The good news is you’ve recognised the need for external intervention. But, one-sided escalations are poison to the health of your working relationships and your company. Congratulations — you’ve just burned your working relationship with that person. Their cooperation is likely to be surface level only. Expect lip service, dragging of feet and even sabotage. And that will be the tone of your relationship from this moment onwards.
- You agree on how you’re going to make the decision. This includes the criteria for the decision, identifying who has a vested interest and who is the person accountable for making the decision. You also agree that you will commit to support whatever is decided. This is a healthy way to resolve conflict and what the remainder of this article is about.
The cost of all the approaches listed above, except the last one, are wanton expenditure of time on energy on things that bring no value and broken human relationships. So how do we go about doing it constructively?
Know when to escalate
You disagree with a decision or course of action and would like resolution. The decision or course of action has a large impact on either your work or the business. You have attempted to resolve with the other party (or parties) and cannot come to agreement. At this point it’s healthy to agree that you need assistance in resolving your disagreement and that you need to escalate together. It’s also worth noting that this is for exceptional circumstances — if you escalate everything then there’s something very wrong in your world.
Know whom to escalate to
Part of escalating together is coming to agreement on the decision maker for this escalation. Agreeing who is the arbiter is the first step. A good rule of thumb here is go to the least senior common ancestor in your combined organisation chart (this may be the CEO at the very worst case). You can also agree a respected peer who can mediate and make the decision.
Follow these guidelines
- First seek to understand the other person’s point of view. A little empathy can go a long way in preventing the escalation happening. A colleague used to get two people in disagreement to argue each other’s cases. The best escalation is one which doesn’t happen because people and come to consensus together and empathy is your most powerful tool here.
- The people doing the work should have agency over the majority of decisions affecting that work. If you don’t like the way another team does something then only escalate if it’s important.
- Understand who is affected by what you’re doing and seek to align with them as early as possible in the process. People surprised at the end of a process tend to react poorly. Failing fast also includes with decision making.
- There is a single decision maker. Decision making by committee just punts the problem elsewhere and can result in the cycle of misalignment continuing. You need a single person. This is non-negotiable.
- The commitment part is critical. Once the decision is made then everyone is committed to making it succeed. If this means altering goals for a team impacted then you do that because everyone needs to be able to succeed here. I’ve seen situations where people effectively committed to disagree but actively undermining the decision after the event. To be blunt — remove these people from your organisation. If you can’t commit to a decision you disagree with then it’s time to look elsewhere because you’re going to harm the progress of those around you.
- The decision making discussion happens in public. If you allow back channels to happen then you’re destroying trust between the participants. You escalate together (which may involve working on a draft together)
- Agree on the criteria you should use to make the decision. What are the characteristics of a good decision? How do you verify you’ve made the right decision. This might also include experiments or analysis to cheaply determine the decision was correct in the case of high impact decisions with a large time commitment.
- You can use a framework like DACI to manage this process.
- Time box the amount of time you’re spending on the decision. If it takes you months to decide anything you’re crippling productivity and all the toxic behaviour noted above will happen anyway. To quote Jeff again “Day 2 companies make high-quality decisions, but they make high-quality decisions slowly. To keep the energy and dynamism of Day 1, you have to somehow make high-quality, high-velocity decisions. Easy for start-ups and very challenging for large organizations.”. Speed of decision making matters.
Notes for decision makers
If you are the decision maker here are some things you need to avoid doing:
- Don’t tolerate back channels and one-sided escalations. If people come to you individually suggest they should escalate together (share this guide, point them at DACI). Your job is to be trusted and neutral.
- Be very clear about the timescale for the decision. Your job is to make decisions quickly.
- Change your mind after making the decision. If people have committed to a course of action and you pull the rug out from under them then you’ve just undermined everyone involved. There may be times where new information arrives which renders a decision obsolete — your job is to facilitate people so they can choose a new decision.
The myth of consensus
It would be great to live in a world where everyone agrees on every decision. If would be great if we lived in a world where people had strong opinions that were loosely held. The reality is the more people are involved in a decision the more likely it is that people will have strongly held contrary opinions. Not making a decision leads to consequences which are undesirable and far worse than deferring or remaining in conflict. Relationships where disagreements are resolved quickly and without drama are enjoyable and productive. Making high quality high speed decisions inclusively and transparently is how you continue to focus on the things which really matter — solving your users’ and each others’ problems.