Guest Post: On Being a Beginner Again: Structuring Problems with Design, Systems, and Decisions
Every once in a while, in every career, we’re faced with a completely new situation, whether it’s choosing a startup idea, pricing a new product, deciding where to live, or selecting a company to acquire.
I’ve observed that, in general, people aren’t very systematic in approaching these situations. We don’t know what information to gather, what processes matter, and what are the constraints. The result, I’ve observed, is a chaotic process, where we fall back on intuition and “gut” decision making. We can do better: I’ve faced these situations often enough that I’ve developed a framework to tackle them, and Decision Intelligence is an important part of my approach.
I start by distinguishing between the problem space and the solution space. This may seem obvious, but you’d be surprised how many people straight to a solution without taking the time for carefully thinking through a problem. A problem space focus explores everything about a problem, including its root cause/pain points, desired results, and interdependencies. The latter takes all this information into account and helps to find the best solution. This includes all the steps from strategizing and formulating the solution to a fully developed solution. And, depending on the situation, solutions can take many forms: in my world it might be a mobile app, an analytics dashboard, a gap analysis, or a one-off decision like a merger.
The framework below combines Design thinking and System thinking for the problem space formulation. I use Decision intelligence to develop the solution space.
1. Discover Problem (Double Diamond, see below) / Problem statement and key features
- What is the business problem we are trying to solve
- Who has it (target persona)
- When (specific period or event over which the problem occurs)
- Where does it appear (sector, country)
- What conditions make it more probable for this to happen (economic growth, recession, seasonality)
2. System exploration (System Thinking)
Once I’ve structured my problem along the above lines, I then structure the solution as below:
- Who are the key players related to the target persona under these conditions (competitors, partners, customers, board of directors)?
- What are the processes being involved (reports, transactions, interactions)?
- What is the information currently used (data points, key events, news), and also what other information could help? This can include:
- indirect effects
- impact of time
- social network
- political/economical/technological conditions and maturity
- impact of the system on the problem
- emergent behaviors
The aim is to redefine the problem statement under system conditions (“define” – the 2nd step in Double Diamond). After all, when developing a solution, we basically re-engineer a system into a more effective (even fun) way.
For example, Facebook initially didn’t provide economic value to its users, but gave them a more fun way to spend their time by personalizing the content in their home page. When users increased significantly, Facebook allowed marketing companies to target specific users based on their preferences (personalized content), thus narrowing campaigns to fewer but more suitable potential customers and reducing costs significantly. So, the solution wasn’t just to allow users to find their friends, but to create a system of interactions with more personalized content and to use a feedback loop to enhance this system (better content brings more users which will increase and improve future content and so goes on).
3. Develop potential solutions (Double Diamond) / Causal Decision Diagram (Decision Intelligence)
Ask people with different expertise/points about the solution. This will help to create a robust CDD, which breaks down silos, reduces errors, and uncovers new ideas (Collaborative Decision Mapping). This also helps to identify hidden goals and/or actions, discover synergies (actions that benefit multiple outcomes), and to support tradeoffs.
4. Deliver solutions that work (Double Diamond) / Implement & Review Decision (Decision Intelligence)
After implementing the decision, do a post-event analysis:
- Don’t check only final results (outcome), but also all intermediate calculations
- Investigate for secondary and tertiary (unexpected) effects
- Ensure whether the solution contradicts other important qualitative / quantitative aspects (ESG metrics, Customer Satisfaction, Branding, etc.)
Haris Papavasileiou is a Data Engineer at KPMG with five years of experience in the data industry, working with cross-functional, multi-national, and geographically distributed teams delivering end-to-end solutions to complex data and analytics problems to customers in the US and Europe. Prior to his current role, Papavasileiou held a number of related positions including Data Analyst, Risk Analyst, and Data Scientist in various sectors spanning from Real Estate and Asset Management to Telecommunications and Consulting.