Guardrails for the influencers
Reflecting on ethical guidelines tailor-made for insights practitioners
Regular readers know I've been reflecting lately on my read through of Kate Raworth's Doughnut Economics. It’s served as a helpful guide for me to think differently about common tasks and activities in the corporate data space. One case in point is Raworth's proposed ethical principles for 21st century economists, which I'll dive into below. Her proposal got me thinking: What are the equivalent ethical principles for insights practitioners?1
Don't get me wrong: Ethics in the data space are nothing new. Ethical principles for data and AI products are prevalent, and oft-discussed. Regulations on the usage and storage of data have also given rise to robust Data Governance roles and responsibilities. I don't mean to imply that this extensive work doesn't exist, nor am I looking to retread that same ground.
But as I consider the role of an internally-facing insights practitioner — who, yes, may build data products for an internal audience, but is largely focused on steering decisions to optimize revenue and other metrics — I can't help but wonder whether there's even more to articulate:
The existing ethical principles for data and AI products still apply, but perhaps indirectly. An insights practitioner’s work steers decisions, which in turn may influence those products, in line with ethical principles or not. How might we center ethical principles in the case of this indirect impact?
And what about the core work of the insights practitioner? Are there ethical principles that speak directly to their day-to-day?
I don't have an answer to these questions yet. Instead, I'll focus this post on my exploration spurred by these questions, with the aim of supporting deeper discussion around what a set of ethical principles might look like for the 21st century insights practitioner.
Ethical principles for the indirect impact of data work
As I mentioned up top, the impetus for this exploration was Kate Raworth's proposed ethical principles for the 21st century economist. They are:
First, act in service to human prosperity in a flourishing web of life, recognising all it depends upon.
Second, respect autonomy in the communities that you serve by ensuring their engagement and consent, while ever aware of the inequalities and differences that may lie within them.
Third, be prudential in policymaking, seeking to minimise the risk of harm — especially to the most vulnerable — in the face of uncertainty.
Lastly, work with humility, by making transparent the assumptions and shortcomings of your models, and by recognising alternative economic perspectives and tools.
As with many ethical principles, Raworth's aims to bring in considerations for economists that may seem a step removed from — or not directly relevant to — their day-to-day modeling efforts. With a focus squarely on driving growth, the impacts of those growth paths may not come to the fore unless explicitly accounted for.
I see a lot of overlap here with internally-facing insights practitioners. While extensive thinking has gone into ethics for data and AI products, they'd be forgiven for not centering them in their work. After all, they're just here to analyze data and make recommendations for the business to grow KPIs — they aren't directly shipping anything to customers. Surely the ethical considerations of what to build are someone else's problem.
Raworth's proposal brings those concerns back to the data professionals. It's not enough to simply chart a path for the business to grow, you need to understand the quality of that growth, and the harms it may inflict along the way, as part of your recommendation.
I'm on a mission to improve corporate data culture for data professionals of all stripes, in Data Science, UX Research, Analytics, and beyond. If you’d like to join me, you can help by sharing this newsletter with the data culture drivers in your network.
Ethical principles to guide the data work itself
When it came to the core work of insights practitioners, I had a harder time finding strong examples. So I decided to switch up my approach, and ask the question from Finance’s point of view. Finance teams are similarly processing data to be more consumable by their stakeholders, but with greater regulation and higher stakes. It's not surprising, then, that there are numerous examples of codes of ethics for Finance professionals — especially in public records from the United States Security Exchange Commission.
One example from the SEC records comes from US electronics store HH Gregg. And I was surprised by how much resonated with the insights workflow:
1. Act with honesty and integrity, avoiding actual or apparent conflicts of interest in personal and professional relationships.
2. Provide constituents with information that is accurate, complete, objective, relevant, timely, and understandable.
3. Comply with rules and regulations of federal, state, provincial, and local governments, and other appropriate private and public regulatory agencies.
4. Act in good faith, responsibly, with due care, competence, and diligence, without misrepresenting material facts or allowing their independent judgment to be subordinated.
5. Protect and respect the confidentiality of information acquired in the course of their work except when authorized or otherwise legally obligated to disclose. Confidential information acquired in the course of their work is not used for personal advantage.
6. Share knowledge and maintain skills important and relevant to their constituents' needs.
7. Proactively promote ethical behavior as a responsible partner among peers in their work environment.
8. Achieve responsible use of and control over all assets and resources employed and entrusted to them.
9. Report known or suspected violations of this Code to a supervisor, a human resources representative, the Director of Internal Audit, and/or the Chairman of the Audit Committee.
10. Are accountable for adhering to this code.
I see a ton in there that's relevant to insights practitioners. Principle 4 hits on something that starts obvious — don’t misrepresent material facts — but then takes it a helpful step further, indicating that supporting independent judgment is also the responsibility of the insights practitioner. For teams often measured by impact via the extent of their influence, this principle can be a helpful counterbalance.
I especially like principles 6 and 7, encouraging practitioners to share knowledge, continue learning, and promote ethical behavior within and across their teams. Ethical challenges can sometimes feel isolating, so reinforcing the role of the group in maintaining the principles via their core data skills can counter this isolation.
Really, the only thing I find disagreeable in the code of conduct is who it's meant for: They explicitly say it only applies to the C-suite and financial professionals. Absolutely necessary, but I'd argue it’s insufficient. Even if insights teams don't have the same level of scrutiny and outside oversight as the Finance team, they are contributing to the same corporate knowledge base that Finance works from. These principles could go a long way in encouraging the same quality in data and insights handling, at all levels of the organization.
As I said in the outset, I don't yet have a fixed set of ethical principles for internally-facing data professionals. I think there are interesting components out there, and the optimal answer lies in combining a few different approaches into one tailor-made for this group.
If you're familiar with a set of principles for insights practitioners that's shareable, I'd love to hear about them! Please leave a comment and share what you can. I'm very much in brainstorming mode on this topic, and would love to have more inputs into the process. Otherwise, expect a return to this topic at some point in the future, once I’ve had a chance to crystallize the thinking a bit more.
By insights practitioners, I mean those data professionals who are primarily internally-facing, delivering data products and insights to steer product and business decisions. This group may comprise UX Research, Data Science, Analytics, and Market Research, to varying degrees depending on the business.
Great post as usual, Sam. These are a great start. I like the example you shared from Kate Raworth a lot. In the past, I've referred to the "Modeler's Hippocratic Oath" from the Financial Modelers Manifesto written after the 2008 crash. But its a little bit flowery, and focused more on the risks of opaque models. (https://www.uio.no/studier/emner/sv/oekonomi/ECON4135/h09/undervisningsmateriale/FinancialModelersManifesto.pdf, relevant excerpt below.)
The examples you shared provide better guardrails for the other work. A few years back I read Cenydd Bowles "Future Ethics", and while it didn't provide a "code", it did give some useful tools for ethical practice.
Excerpt of the Modeler's Hippocratic Oath:
"I will remember that I didn't make the world, and it doesn't satisfy my equations.
Though I will use models boldly to estimate value, I will not be overly impressed by mathematics.
I will never sacrifice reality for elegance without explaining why I have done so.
Nor will I give the people who use my model false comfort about its accuracy.
Instead, I will make explicit its assumptions and oversights.
I understand that my work may have enormous effects on society and the economy,
many of them beyond my comprehension"