The SumUp Decision Making Framework

At SumUp, we’ve learned that how we make decisions often determines the quality of the decisions. We want to make decisions that have a significant, positive, and lasting impact, making SumUp an ever more effective organisation. How can we accomplish that in an agile environment?

The key is to make decisions collaboratively. If we make and communicate decisions in a low hierarchy / high autonomy environment that is iterative and includes all stakeholders, team members will better execute the decisions because they fully buy into them rather than because someone “said so.”   

The “Co-create, Consult, Test, Tell and Sell” framework by Peter Senge, is an agile decision making approach that is authentic to SumUp’s organisational culture. It identifies five basic decision making models along a “collaboration continuum.” The five models (from most collaborative to least) are: Co-Create; Consult; Test; Sell; and Tell.

The fastest, but least impactful decisions are made using a Tell or Sell model, with decision makers acting autocratically and receiving little to no input from others. The more collaborative decision making models (Co-create, Consult, Test), which we tend to favour at SumUp, take more time but result in better, more impactful decisions.

1. Co-create: 3~5 people discuss and prepare a proposal or prototype to initiate the decision making process. The proposal can be text (written summary), image (infographic, diagram, representative image, schematic, etc.) or analysis (behavioural, financial, statistical, etc.) aimed at tackling the objective. A prototype could also be developed as a physical or conceptual model. Regardless of the method, the group works together to demonstrate how they will resolve a topic or improve upon an existing challenge. The people involved can be very familiar with the problem or not familiar at all. In fact, it may help to include people unfamiliar to provide a fresh take or different perspective. Example: A group of leaders discusses how to financially support every SumUppers work-from-home setup. During the meeting, the group summarises the problem and suggests a solution in a short document.

2. Consult: A proposal or prototype is shared and discussed with 15~30 SumUppers to understand skepticism, identify frictions, and illuminate blind spots. Ideally, consulting includes people who have a direct stake in the problem or proposed solution. This gives an informed perspective to the process and starts to gain buy-in. The feedback is worked into the original proposal or prototype, solidifying it into a formal plan. Example: The co-created work-from-home plan is shared with a few leaders and a handful of people partners for asynchronous feedback. The feedback is worked into the proposal in several iterations until the feedback is either incorporated, resolved, or rejected.

3. Test: A plan is shared with a wider group of people, e.g. a specific tribe or team. It is communicated as a decision that has been made but might change if it yields different results or reactions than expected. If results are positive, the communication and implementation of the decision are gradually extended until it reaches the whole organisation. “Testing” and extending the test to include everyone keeps a door open for continuous improvement on the plan. Example: One or more squads decide to test the work-from-home policy. As some team members were already involved in the consulting phase, it was generally positively received. After a quarter, some minor feedback comes up that is worked back into the original plan. The success of the introduction test is then shared with everyone at SumUp and subsequently, many teams adopt the same approach. The decision has now de facto been made without it ever being very explicit.

4. Sell: A decision is made and will be implemented. Rather than accepting feedback and challenges in an iterative revision process, we try to “sell” people on the benefits of the decision to promote buy-in. “Selling” is used less frequently. Typically, only when the potential impact of the decision is very limited or there is a high level of confidence that the decision is the best one for everyone impacted and there is no time for a collaborative review. Example: The tribe leader has decided for the tribe offsite to be held at a mountain house near Sofia. She tells the tribe, explaining the rationale for the decision.

5. Tell: A decision is made and will be implemented without giving people the ability to give feedback and challenge it. Minimal to no effort is made to convince them of the benefits. “Tell” is the most hierarchical of the decision making models, and is generally something we try to avoid at SumUp unless we find ourselves in an emergency situation with no time to collaborate or sell. Example: The fire alarm goes off in the office. Some team members are sure that it’s a false alarm. The office lead tells them to leave the office immediately without further discussion.

It’s important to note that these models are not mutually exclusive, and at SumUp we tend to take an iterative approach that begins with co-creation and then expands to consultation and testing.

The iterative co-create/consult/test approach is more time-consuming than selling or telling. However, following it improves the effectiveness of the decision making process through the incorporation of feedback while promoting buy-in. Moving through the interactive phases (being mindful of the number of people involved in each phase) can save time in the long run. Efficiency can further be improved by the original co-creating group letting other participants know which phase they are in and what level of involvement they are expecting of participants.

While potentially much more time efficient in the beginning, selling and telling will typically lead to lower engagement and dissatisfaction with the end decision. As fewer viewpoints are taken into account, autocratic decision making can lead to suboptimal decisions. As fewer people have been involved in improving the decision, buy-in - and as a consequence - commitment and impact will be lower.