What is the secret to designing a data system that provides useful information? Or getting great outcomes from a project?

People who know me know I think “public health is everything” (a blog for another day). In recent years I have added another maxim: “think in reverse”.

We have a tendency as humans to think in natural order of time. We start at the beginning and make our way through the middle to get to the end. For example, first you build the data system, then you build the report from the system and then you use the data. The problem with this approach is when you get to the part about using the report from the system it invariably isn’t quite right. We now want data that was never designed into the system and have to retro-fit what we really need into the system, if at the late stage it is even possible.

Thinking in reverse would have us think first about the information we need from the system, which is grounded in the purpose of the system. We start at what information we want, then go to the report we need. Once we have this, we start at the beginning and design the data system through to building the report we started with. Systems built on the basis of the clearly articulated end use will be properly tailored to the end use.

We can also use examples from programmatic or policy interventions, for example. These are often complex and hard to implement because there are many unknowns in the real world. Interventions are often designed by government or other large entities, who have a usual way of working and staff that must be gainfully employed. While there may be a problem to be solved (e.g. prevent chronic disease) there is often a drift off course in service to the busy work. When this situation also includes hard to measure outcomes, attention can be turned to measuring what we do rather than what was achieved. These are things like how many consultations did we hold, how many events did we attend, how many reads of our paper were there.

If instead we start with a hyper-focus on the outcome we want, we make sure that whatever we do in implementation is in service of the outcome. In other words, we build back from the ideal state, ensuring all our work is in reference to the ideal state. Every step we take is aligned with the ultimate goal. This thinking is not limited by the usual inputs we have – 5 FTE, a small goods and services budget – but expands to how we might actually contribute to solving the problem considering all available options.

Possibly one of the biggest problems with thinking in natural time is when we have a technology, or product or a program, and then we search around for a problem to solve it with. It sounds unbelievable, I know, but we will all have seen data that is mined without a purpose to see what can be found. I am sure we have all also seen AI proponents avidly search around for problems their amazing technology can solve. The major problem with this kind of approach is the solution options are limited to the one we started with, “AI can solve this one!”. Thinking in reverse would start with the problem and the outcome we want and allow problem solving that can consider all potential options.

Thinking in reverse has application everywhere (see point about public health is everything above :)) and works very well in fields outside of health. Businesses that start with what the customer wants and builds back to a product will have products that are tailored to the customer’s needs.

In the world of public health, our metric of success is sustainable positive outcomes for the community. Thinking in reverse is a critical way to make sure this success can be achieved.