Bayes Impact is a nonprofit startup under Y Combinator that utilizes data science to create new technological solutions for problems worldwide. Paul Duan, co-founder and president of Bayes Impact, joined Blueprint for a fireside chat on April 22. Here are some of the insights and experiences he shared!
I. Behind Bayes Impact
More about the organization — and the change it’s driving
Many global organizations today face major challenges with managing and collecting data. From the VA to healthcare transportation services, social impact workers bear a number of issues with limited resources and information.
Bayes Impact provides the manpower and data necessary to solve such problems —partnering with top data scientists and experts to offer new technological solutions.
Paul says that transforming the non-profit space is feasible, especially considering that most non-profits use outdated technologies. Many of the largest organizations in the sector, such as the Red Cross, are over a century old — making it a great place to create disruptive environments with high-growth, high-impact outcomes.
Over the last year, the startup has undertaken high-profile global projects with a wide range of partners, including the White House. From these projects, Paul has seen that the technological component of data (reading raw stats and understanding trends) is just the beginning.
He sees how using data science effectively can link together many different issues in the public sector and find problems that extend beyond a single industry (i.e. similarities between healthcare and criminal justice).
What’s next? The scalability of decision-making, so that the problems clients face are actually solved.
II. Needing Top Talent (and Being It)
A major takeaway from Bayes Impact’s growth this year and how it applies to students anywhere
Growth isn’t possible without top talent.
And although Duan says that finding dedicated employees hasn’t been a major issue for Bayes — people inherently seek more impactful, important work — he acknowledges the initial confusion for new graduates.
“It’s not really the choice between working for a non-profit or a for-profit. More like non-profit and no-profit,” he commented.
After graduating, many students enter corporate settings — given the tighter pay structures of the non-profit sector. In fact, the pay-off for working at a startup can take up to several years. On the outset, attracting talented graduates to social causes therefore seems challenging.
But according to Paul, 31% of students’ decision-making is about the mission. The data speaks to the power of non-profits, and that effectively creating system solutions and sharing a strong vision can powerfully attract top talent.
If a student truly wants to enter the social sector, he says, there’s no reason not to start right away.
Duan does admit that gaining prior experience, in his case, helped him accelerate Bayes Impact’s growth. Before starting the non-profit startup, he worked as a Data Scientist at Eventbrite, which enabled him to build a strong network of social partners, government decision-makers, and community influencers.
For non-profits, this kind of network is invaluable, as are the resources that come with it. Duan’s most basic piece of advice is thus this: begin connecting with the right resources and mentors (now).
“Apply for fellowships that give you $80k in free money as well as a pretty strong network. Surround yourself with people who are a bit older and have a more developed network base.”
III. Audience Q&A*
*These answers have been paraphrased and organized to provide the answer with better clarity and context.
Q: Most of the problems you’ve found with clients (from running Bayes Impact) are operational problems. Is there any incentive for companies to improve their processes other than their external competition?
This answer is split between government partners and non-profit ones.
A. Non-profits:
There is little to no competition between the organizations, and it is difficult to measure impact, costs, and expenses.
Some problems instead:
- There isn’t much incentive to improve processes (it’s much easier to keep on doing what has been done)
- Securing talent and finding catalysts for change within the organization are often difficult
- Solutions for such problems usually come down to money — and incentivizing an organization with funding
B. Government Partners:
Most government contractors are rife with conflicts of interest, mismatched incentives, and backroom deals.
Since these are difficult to change, the goal is instead to create more modern organizations that compete with them.
Just as Palantir has been trying to compete with Lockheed Martin, and have improved the cycle, we want to change the way things are done.
The main way to change the government problem is to join the entrepreneurial space and make the systemic change yourself, because if you don’t take action, no one else will.
Q: There are different routes to take a start-up, such as being a non-profit with a for-profit business model and vice versa. What advice do you have about forming and leading a start-up?
- As a non-profit, the main thing is that you don’t have any equity, which makes funding difficult.
Since we didn’t have any access to VC funding, most of the funding in the initial years was based on our projects, but not our growth. The notion of funding for growth in the non-profit world does not exist. In the non-profit world, you only have a handful of non-profits that really make it big as a result of this.
2. We have a business model where we charge our partners for service, with the idea that we can change the way we think about charities.
On the crowd-funding side, we have orgs like Watsi redistributing money from people in rich countries to mostly poor countries. This kind of model works, but it also makes sense to create a start-up mindset where instead of redistributing value, we create value.
If you want to create a non-profit, you need to have strong reasons why you want to do so.
In the past, we’ve thought about non-profits as not making money, and you choose to do so for the lifestyle. Since we have so many options now, such as hybrids, you have to really think about why you are choosing to be a non-profit.
In some cases, it makes sense to be a non-profit. In our case, we want to partner up with governments without any conflicts of interest and get our hands dirty in complicated systems, but you might not have to do that. Having the 501©3 status makes you accountable and makes you use impact as the greatest metric. In the end, it depends on if you care more about the money or the impact.
Thank you to Paul Duan for sharing your candid stories and advice!
Interested in learning more about Bayes Impact?
Make sure to visit www.bayesimpact.org (and read a Tech Crunch feature on them last year here)!
For more updates from Blueprint, follow us on Facebook and Twitter!
Thanks for reading, and have a fantastic summer!
