How to Make Decisions More Efficiently (or when you’re sick of arguing with your team or co-founders over a decision)
Making tough choices are part of the day-to-day at any startup or investment fund. However, how those choices are made varies greatly. Decision making within an organisation can range from hyper-gut-feel to hyper structured and rigid.
Whilst I don’t profess to be an expert on the subject, having now done over 350+ investments, one of the areas we needed to innovate to scale was around was how to best make good decisions efficiently and quickly. Within this post I will review some of the decision making tools we use at Seedcamp that allow us not only make decision quickly, but also to quantify & qualify them
To start with, when coming up with or choosing a decision making process, it helps to know what kind of outcome you want and when. Sometimes you THINK you want an absolute outcome (more on that in a second), but in reality you might benefit from having a stage before it that assists your absolute outcome by grouping possible choices into relative outcomes first, followed by the absolute choice.
To help explain the difference between absolute and relative, I’ll share the two types, and the variants we use internally:
- Relative Processes — These processes are ones where there is no need to make a decision at the end of running it, but helps where the elimination of options you don’t want is critical; it helps reduce the time you spend discussing real options of consideration later. Of course, you can use relative processes for choice-making as well, but these don’t work well if there is only one decision to make without other choices. In the words of my friend Devin Hunt he describes them as a decision making process that surfaces a number of _good_ options and identifies options that can be immediately discarded. By reducing the pool, it makes it easier _later_ to engage in an absolute process for a final choice.
- Absolute processes — these processes, unlike the curatorial nature of relative processes tend to mainly drive towards a binary outcome. They help organise the information that is spread out across the team to provide one choice. The downside of only using an absolute process is that it strips the team of a view of where an opportunity or decision sits in the universe of choices. However, as I mentioned, it usually works well in decision making where you don’t have multiple choices to choose from, rather one decision to make, yes or no.
To kick off with, our relative processes include:
1.1 Currency points system — when we want to collate the opinion of various people, this system works well. The larger the group the better this works too. Basically, the idea is simple, everyone is given a budget of ‘points’, we use ‘9’ but as long as it’s easy and odd it helps reduce too much burden (too big) or not enough granularity (too small). With these 9 points you can ‘spend’ them outcomes you prefer, whichever way you want. Say if you want to ‘spray and pray’ you could give 1 point to all the choices. However, most people tend to group their use of points into either high conviction (lots of points to a few) or a smallish cluster. You then weight two outcomes — the choices determined by the most points, and the choices that may have not got the most points, but that most people gave any points two. They are both interesting to discuss.
example: 5 choices, 2 get the most points, 3 get the most votes (at least one). The chat to have is, are the 2/3 the same? what’s the difference between the most popular one (most votes) and the most convicted one (most points)?
IMPORTANT NOTE: Sometimes what prevents good decision making processes are bias towards one individual’s opinion. A key thing to keep in mind in this one or any other is try to review the points and discuss them with the people’s individual scoring anonymised. When we do it, we hide the spreadsheet columns of how people voted so that the conversation is focused on the data, and not on team politics.
1.2 Attribute based stack system — - Similar to the above, this format’s ultimate objective is to create a top to bottom list of options from best to worst, however on this one, it is done on the basis of attributes that each choice has. For example, if we were talking about cars, it would be how each car fared on gas mileage, speed, etc, and then for each of these, from 1–5 how they did in each attribute. You then take the average score and stack and rank accordingly across the attributes so you can see how choices stack up across each. The key thing is to make sure you’re choosing the right attributes to focus on, otherwise it’s garbage in and garbage out. For some of our investment decisions, for example, we put extra weight on team related attributes as that correlates highly with success in our experience.
example: car 1, 2, and 3 have mpg, top speed, cost, and maintenance fee. If cost is the most important attribute, how do the cars stack up on the rest of the attributes?
Switching gears, our absolute processes include:
2.1 The Gap Points system— In order to force a decision within the team, we ask people to give a company a score from 0–10, but it cannot be 4,5,6 and the average of the team’s score has to be greater than 7.5 for us to really consider the option as viable. In effect, this allows the team to not cluster around hedged opinions, rather the system forces a polarity and also an extra level of commitment by someone to really get a decision over the line.
example: the team has identified a possible yes or no decision, and no further debate on attributes is helpful. A vote is required which synthesises people’s attributes into one score. Assume a team of 4. One votes 3, one 9, one 7, and one 7.5. The average is 6.63 (which as is below a 7.5 is a ‘no’), but that matters less than the obvious which is that most of the team is lukewarm about it, and one is vaguely supportive. This can expedite the true nature of the debate internally.. why is one person a no and one a strong yet and two stuck in the middle?
2.2 The Team Momentum System — Sometimes too many people are involved in a decision and that slows things down or adds too many variables to move things along to a point where a wider team can comment. This particular system moves things to a subset of the core team, relying on a critical group to drive momentum within the larger group. As part of this system you reduce the people that are accountable for a decision and then the remaining team members are consulted for a final decision, but with a view unless there is a strong no at the end of a final review, that the default answer is ‘yes’ as an outcome. Naturally there needs to be a high degree of trust and you need to make sure there is no ‘factionalisation’ of team members or otherwise you effectively circumvent the value of the system.
example: Two team members passionate about a sector, say HR-tech, engage with a company and both build conviction on a company, the remaining team members would each have a chance to veto a ‘default yes’ from those that have researched it, but likely they only would for some externality rather than a view on a sector that they’re not exploring themselves.
2.3 Lastly, The Pizza System is a variant of the above but requires more planning as its one where ‘we collectively cut the pizza, but someone picks the slice’— In effect, the team agrees on a predetermined criteria that qualifies a decision as viable and then someone has the ability to drive the decision through conviction, so everyone is happy with the outcome based on the criteria, but one person drives it to a yay or nay, which reduces analysis paralysis.
example: If the entire team agrees that any company that is in the space of solar technology that is capital efficient will automatically trigger a yes/no debate. In effect, you have decided which way something gets chosen for either final decision or final discussion.
As I said before, any one of these systems absolute or relative, can work for just about any decision you want to make, but the key to really improve efficiency within a mix of team sizes and severity of the decision making is knowing which system helps you move things along prior to having to debate things with your team. Don’t limit yourself to just one system, if you benefit from it, stage them to help reduce cognitive load prematurely as it can fatigue a team’s discussion time and can polarise choice prematurely. If you have a variety of choices prior, come up with a way of using a relative system before an absolute system is applied.
To illustrate the above, at Seedcamp, we usually have to make lots of decisions, but it’s hard to treat each choice as an absolute choice, so one of the things we do is set up a curation process by first using a relative decision system, which then allows us to focus our debate time on those choices that surface to the top only, and then once we have a narrower group, we switch to an absolute decision system for making final decisions.