Data remains a hot word in business already since quite some years. Whether we understand what it means or not, somehow we should all be machine learning and data strategizing. Seminars fill up with corporations trying to figure out their data roadmap.
Many are approaching the data game in a pragmatic and opportunistic way: What data do we have? What could we do with it? What questions could it answer? This is a practical approach for business and might lead to great accidental finds for revenue potential and learning.
For big problems that matter regardless of whether they are easy or difficult to solve, however, opportunism is not a very effective cure.
In the search for great answers, many lose sight of what they were trying to solve to begin with. Let’s say you wanted to do a data-driven study on what causes bullying in middle-schools in Northern Sweden. Halfway through the tricky data venture you come across interesting data about therapist sessions among middle-schoolers in Sweden. You get excited! That’s when you start to answer a slightly different question, perhaps one like “How many teenagers get treatment for depression in Northern Sweden?”
This is very human: there is a big lure in switching to a question our data happens to serve.
Concentrating on tweaking the answer is natural: it is answers, not questions, that we typically compare and reward. Whose Facebook advertizing optimization algorithm is the most accurate? Who attracts the most clicks on their news portal? Additionally, big data technologies have allowed us to handle larger and larger masses of data to answer sillier and sillier questions. Enticing!
Let’s take an example that I have gotten a bit familiar with through our work at Upright: impact of companies.
There are currently many projects aiming at somehow answering the big question: how do companies impact the world around us? What data is out there, and how could it help drive better decision-making for investors, companies themselves, consumers? Many start with a big aim in mind: making the world a better place.
However, questions about holistic impact of companies are simply very hard to answer. If you take an opportunistic approach, it is easy to get off-track. When you start to dig into what kind of data is out there, you quite quickly run into three main categories:
What are these three actually good at answering?
#1 is great for answering the question: What do companies want to report and communicate regarding their impact? This is useful input for communications and branding consultants, and perhaps for companies themselves as they build their identity and motivate their employees with their sustainability efforts.
#2 is great for answering the question: How well do different companies adhere to the compliance factors this particular data provider has chosen to measure? This might be useful when wanting to understand how much resources a company puts into adhering to certain standards.
#3 is great for answering the question: Where could I get some inspiration and guidance when I’m starting to think about my company's impact? What impact categories are there?
None of the three answer the question: what is the holistic impact of company X? As a consequence, none of them is really helping us understand the actual impact of businesses.
The question that most drives me professionally and personally is: how should we as humanity allocate our scarce planetary and human resources in the 21st century in order to not become extinct in the next 200 years?
That’s why I have founded Upright, where we are building a very imperfect answer to what we believe is the right question: what is the net impact of companies, and how data about it could be utilized to better allocate planetary and human resources at humanity’s disposal today. We want to help people with concrete data-driven tools when pondering things like: If I want to fight climate change, which companies should I invest in, buy from, work for? How does the answer to the previous question change, if I choose to maximize job creation, tax generation or health benefits instead?
This question is the most difficult one I have ever tried to solve. I hate it on a daily basis. However, sticking to this incredibly mind-bending question is what keeps me going, as I believe it is this question exactly that we should be answering.
No perfect answer even exists for this question. Our goal is to develop an answer that is significantly better than what is available today. Being able to reach our goal requires that we tolerate an imperfect answer for quite some time.
18 months into the project, our current answer to our question is less accurate than an individual company sustainability report's data is to the question “How much water did company X use during year Y in its internal operations according to how we define and measure water consumption”. However, already today our answer is better than any other we have been able to find so far in addressing the question we are addressing. That said, majority of our work still lies ahead.
It could be tempting for us to produce data that sounds somewhat similar, but actually answers a quite different question. For example, it would be significantly easier with our skill set and some basic machine learning to answer a question like: “What impact-related news are there about companies and products? Who looks good and who looks bad according to them?” However, that would merely make us emphasize and accentuate the information disproportions already existing in media coverage, not answer our original question.
If you share our passion of not compromising the question, you are welcome to explore the development of our imperfect answer on The Upright Playground. In the meanwhile: what is the question you don’t want to compromise?
Slush and Upright teamed up to quantify the net impact of all startups coming to the North’s #1 startup event in this December. The results? It’s time to bust some myths.
Unless you are filthy rich or a shopaholic, chances are you influence the world around you the most via your work. And that is good news.