How to Have Impact

The best minds of our generation are thinking about how to make people click ads. That sucks.‌‌ ‌‌‌
Jeffrey Hammerbacher

I disagree.

Many of us want to have a positive impact with our work. But what does that mean? Typically this is portrayed as “changing the world” or “improving people’s lives” or “solving important problems.” When we hear quotes like the above, it makes us feel that people who work on things like ad optimization aren’t spending their energy and talent on what’s best for the world.

There’s an implicit moral judgement being made about what is “good” and “bad” for the world; what’s valuable, and what’s not. Working on climate change is good, working on ads is bad. Public interest work is good, social media is bad.

This is where we go wrong. We think we can predict what will be valuable to society in the future. We can’t.

Why not? How does impact really come about? What is our role, as individuals, in changing the world?

The intention/impact gap

We know that the “road to hell is paved with good intentions” — i.e., there’s a gap between intention and impact. This follows a few common patterns. For example:

Good intentions → Negative impact

Many government policies fall in this category (the SF housing crisis is an example of well-intentioned policies from the 70's gone wrong).

Another common pattern (this one is harder to detect):

Good intentions → Positive first-order impact (the thing you’re measuring) → Negative second-order impact

Nonprofits often fall victim to this pattern. Clothing donations from the west to Africa are an example. The influx of cheap clothing has generated thriving local reseller businesses, which in turn has undermined local textile industries.

But the intention/impact gap reaches further than that. Sometimes the opposite is true:

Neutral intentions → Positive impact

Let’s take ad platforms as an example. On first glance, ad platforms like Facebook and Google just seem like ways to make money – surely it's not impactful work. I’m not so sure. What about tech innovations in ad engines that have been used broadly (machine learning, big data analysis, etc)? What about all the former employees who have become financially secure enough to start businesses and nonprofits? Or the concentration of talent that has led to awesome people meeting one another?

To take it a step further, consider all the direct-to-consumer businesses that are possible today. For example, I’d argue niche e-commerce businesses like the ones on Shopify are in large part enabled by modern ad platforms. These types of businesses can now reach audiences that are both much wider and more finely targeted than they could have pre-online ads. So in a sense, ads have been a democratizing force that have enabled more people to become entrepreneurs and build their businesses.

There’s a fundamental gap between intention and impact, because the world is complex. You don’t have control over the outcomes of your work, even with the best intentions. It’s also often difficult to measure impact. Second and third-order effects abound. Unintended consequences, both good and bad, show up without warning.

I’m not saying we shouldn’t have good intentions; of course we should. I’m suggesting we consider a different way of evaluating our work choices than just our intentions. Instead of telling ourselves, “I should do work that has an impact,” what if we ask ourselves: “How can I work in a way that contributes to the broader fabric of progress?”

Rather than thinking individualistically, can we think holistically?

The whole is greater than the sum of its parts

The current impact narrative is one of individuals that change the world through their vision and boldness — a hero’s journey. This narrative misses the forest for the trees. We're missing the bigger picture because we're overly focused on specific data points.

What's the bigger picture when it comes to impact?

Let's stick with the forest analogy. The bigger picture there is that a forest is more than just a collection of trees – it's a dynamic, robust ecosystem. How does that transformation happen?

We find some clues once we begin to see the forest as a complex system. The way the forest ecosystem emerges is through interactions between its diverse individual members (trees, animals, organisms, other vegetation, etc). An ecosystem isn't planned or designed. It emerges out of the collective behaviors and interactions of all the parts.

Once we start to view our work as one tree among many, it becomes clear we’re part of a greater collective, an incredible system of people and creativity and ideas that are interconnected over space and time. Impact emerges from this "ecosystem" – it's an indirect, secondary outcome.

This model explains why the traditional impact narrative is flawed. We look at the outcomes of a complex system and try to mirror them. We see an individual or an organization that executes well against a visionary mission, and we try to make our story look like that. What we don’t consider is that these individual stories of high impact are the result of countless other experiments (both successes and failures), randomness, timing and luck. In other words, these examples of high impact are themselves an emergent behavior, and wouldn’t be possible in isolation from the system and its history.

Let's reshape the impact narrative from being about heroic individuals into being about the story of the collective.

What is our role in the collective? It's hard to figure that out looking top-down. Complex systems aren’t amenable to analytical approaches like root cause analysis. Cause-and-effect doesn’t have quite the same meaning in a web of feedback loops and densely interconnected networks.

So how can we figure out how to best contribute to the system that generates impact?

How impact is generated

Taking a page from evolutionary theory and natural selection, the way to find things that work well in a complex, dynamic system is through massive trial and error. A simplified version of the process:

  1. Experiment with a lot of new things.
  2. Determine which of those things works best by testing their suitability to the environment.
  3. The things that work survive and reproduce.

How does this apply to generating impact? First, what do I mean by "experiment"? Broadly an experiment is:

  • Coming up with an idea.
  • Realizing the idea (i.e., making it into a thing).
  • Testing the idea by seeing if it works in the real world.

An experiment can take the shape of a startup, side project, or open source work. I’m not just talking about tech here but any kind of creation in uncertain environments: nonprofit work, art, pure science research, writing. Also, there's nothing here that specifies scope – small experiments can turn out to be just as valuable as big ones.

Every experiment is a risk (we’re not sure beforehand if it’s going to work or not), but we want to choose ones with asymmetric risk profiles: limited downside and much higher upside.

This evolutionary process happens both at the scale of the individual and at the larger scales of communities, cities, countries, and the world. At an individual scale, we don’t know if an idea will work. At larger scales, aggregated over many individuals, the things that work ("suitable to the environment") will spread ("reproduce"), generate more experiments ("mutate"), and every now and then something will emerge that makes people's lives better.

The most important part of trying out ideas is the learning that happens at both the individual and system levels whenever you realize an idea and expose it to the real world. The benefit might not be immediately obvious; sometimes old ideas become relevant again decades (or centuries) later.

Notice this process depends on what came before; future experiments depend on what we learned from past ones. And the more experiments we have, both in quantity and diversity, the better chance that impact will emerge at larger scales. So the more of us that try (and the more times we each try) the better.

You don't need to "go big" to have impact. You just have to go.

Answerable questions

On discussing this essay, a friend asked me, “Optimizing ads vs. working on a crazy carbon capture system, which do you really think has more impact?”

My answer was that we can’t possibly know which is going to have more impact. I consider these equivalent choices in terms of potential value to society – before you try them.

Some aspects of the impact generation process are in your control, and some aren’t. The world (i.e. the "environment") isn’t in your control. Neither are the outcomes you get from trying things – remember the intention/impact gap. However, generating ideas, making things, taking risks, and the people you choose to work with are.

So how do you make impactful choices? Ask yourself answerable questions instead of unanswerable ones.

Counterintuitively, "Which choice will have more impact?" is an unanswerable question. It's unpredictable and out of your control. (Of course, excluding things that are outright evil or spammy or criminal.)

Answerable questions are ones like:

  • “How can I take more risks?”
  • “What do I personally want to try?”
  • “Who do I want to try it with?”
  • “Does my experiment work in the real world?”
  • "How can I support others in their experiments?"


In The Theory of Moral Sentiments, Adam Smith wrote about how social norms come about:

It is thus that the general rules of morality are formed. They are ultimately founded upon experience of what, in particular instances, our moral faculties, our natural sense of merit and propriety, approve, or disapprove of. We do not originally approve or condemn particular actions; because, upon examination, they appear to be agreeable or inconsistent with a certain general rule. The general rule, on the contrary, is formed, by finding from experience, that all actions of a certain kind, or circumstanced in a certain manner, are approved or disapproved of.‌‌
- The Theory of Moral Sentiments (1759)

In a sense, Smith is talking about citizenship. What role should we play as members of society? How can we best contribute to improving our social norms and making the world better? Smith's answer is that those "general rules" emerge through our individual actions and responses to the actions of others. And so the way to improve society (i.e. change the rules) is to make different choices as an individual.

How can we have impact as individuals? I hope to relieve you (and myself) of the need to save the world. Instead, focus on being a citizen of the system that generates impact: ask yourself answerable questions and experiment.

As Steve Jobs eloquently said:

You can’t connect the dots looking forward; you can only connect them looking backward. So you have to trust that the dots will somehow connect in your future.

Jobs was talking about our personal paths through life; how you never know in what ways the choices you made in the past influence your future. I think this idea is just as applicable to impact. Can we trust that when the world of the future looks back on our collective actions, the dots will somehow connect?

(Thank you to everyone who reviewed drafts of this essay, including many folks at South Park Commons.)

Show Comments