Back

Office Hours: I’m the co-founder and CTO of YC-backed startup Searchlight. I’m Anna Wang. AMA!Featured

I’m Anna Wang, the co-founder and CTO of Searchlight. We graduated from YC’s W19 batch and raised our Series A from Founders Fund and Accel. Fun fact: my co-founder Kerry is also my twin sister!

My focus at Searchlight is leading our product, engineering, data science, and design teams. Our software collects and connects data from candidate assessments, reference checks, and post-hire outcomes.

We are passionate about solving a really hard technology problem: creating an structured, objective standard for understanding people holistically and offering these insights in an actionable way. We know that understanding a person’s culture preferences, unique strengths, and motivations to a company and job is more predictive of employee success than just assessing technical skills. We build software to help companies like Coinbase, Zapier, and Udemy tap into these insights to make higher quality, more efficient hiring decisions.

Before Searchlight, I wore a few different hats and experienced life in software engineering at Uber, product management at Google, and management consulting at McKinsey.

I grew up outside of Los Angeles, California in a first-generation Chinese-American family. My parents are both scientists, and seeing them take such pride in their work helped me understand at an early age how important it is to find a professional career that connects to our purpose and passions.

Ask me anything, and yes, we’re hiring!

Thanks so much for joining us @annawangx!Elphas – please ask @annawangx your questions before Friday, October 14th. @annawangx may not have time to answer every questions, so emoji upvote your favorites 🔥👍🏾➕
Hi Anna. Wow - great idea! I like that you are taking a holistic approach. My question: What has the result been? What successes have you had with this approach in terms of reducing attrition or reducing time to hire?
Thanks for the question! We've worked with amazing companies to date and been able to deliver some exciting results.Udemy uses our behavioral data on Cultural Alignment and Power Skills to make great hires with a tighter interview process. They now have 40% faster time-to-fill and 20% higher first-year retention.Zapier uses Searchlight to hire more efficiently for values alignment, which increased the predictiveness of their interview process 3x (over just standard interviews) while enabling their recruiter team to extend offers 1 week faster.Searchlight helped Snapdocs improve Quality of Hire by hiring candidates that ramped to full-productivity 25% faster, and with higher hiring manager satisfaction.
Hi Anna, congratulations on the progress Searchlight has made! How did you come up with this idea? How can Searchlight determine these qualities of each applicant when many people do not put so much of themselves out there?
Hi! I shared the backstory for the idea behind Searchlight here :) https://elpha.com/posts/jbogx05n/office-hours-i-m-the-co-founder-and-cto-of-yc-backed-startup-searchlight-i-m-anna-wang-ama#gjr3wpyfSearchlight not only analyzes this data, but we also take responsibility for collecting it. We power our own candidate assessments, reference checks, new hire reflections, and manager reflections.By touching these different points in the candidate and employee journey, we're able to ensure a standard way of understanding a person's power skills, culture traits, and work motivations. The way that we design our survey experience is based on best practices in people science and question design so that we can learn a lot while still making the experience fun and engaging.
Hi Anna,This is incredible! What a needed solution to such a real problem. I am a first time founder and applied to the YC winter batch. My question for you is, what was your process from idea to conception? I am struggling with how to find the right next steps and getting bogged down in areas that don’t push us forward. So encouraging to see another female founder!
Finding the right startup idea was the hardest part of my journey. I find that some of the smartest founders struggle the most here, because there's no such thing as a perfect idea and really smart people are good at poking holes in ideas. Remember that if everyone thought it was a good idea, that startup would already exist and there'd be no market opportunity!So, the challenge of every founder starting out is finding the idea that they want to WILL into existence. Every idea evolves. So, when starting Searchlight, I focused on these factors:- personal connection to the problem (aka founder-market fit)- big market- ability to raise venture funding (correlated with the previous bullet)- ability to hire (aka "zeitgeist of the times") Ability to hire is the most important one. No startup can succeed on its own. So the idea has to captivate and encourage exceptional talent to come work with you!
Hi Anna, congrats on raising a successful round of funding! I love the holistic approach because the recruitment process is incredibly grueling and sometimes downright painful to go through as a candidate. 💜I'd love to understand your perspective on the following:1. What have been the challenges you've run into with leading your cross-functional teams? How did you/are you overcoming them? 2. Given your diverse work experiences, what lessons would you share with us about long-term career visioning and planning?3. What are the most interesting or unexpected customer problems you've discovered or observed while working on Searchlight? Thanks for your time and insights!
Hi Tiffany!1. Effective communication and alignment became a challenge as we transitioned to remote work and hired more cross-functional teammates. We are a hybrid team today, and making the most of our async and sync time is critical. One thing that we started doing is being more intentional about our meeting cadences, this resource is super helpful! https://coda.io/d/Rituals-of-Great-Teams_dJfqgxV3UDv/Meeting-design_su8Jf#_luMsF For example, we now do a weekly pod sync with backend engineers, our product manager, and our data scientist and that has been helpful for activating collaboration and pushing projects forward.On a personal note, I sometimes find it difficult to do so much context-switching! I'm trying to overcome this with better calendar management and more intentional goal-setting for each day by focusing on 1 function per day.And of course, the best way to overcome challenges with managing is to hire well and bring on exceptional people who can "manage up." When there's a hire that changes the possibilities of what you think you can do because they're just THAT good, it becomes easier to lead and trust your team. 2. I'm flattered to be asked this, but I haven't done much long-term career visioning and planning yet. At each crossroads to date, I've made my job decision based on what would allow me to take on the most challenges and learn the most. Building Searchlight has been a journey of ups-and-downs, but I am very focused on and energized by what I'm doing right now because there isn't another job that teaches me this much.3. As an engineer, I learned some tough lessons that customer problems aren't always related to the product and can't always be solved by better engineering. As Searchlight as moved into selling into bigger companies with enterprise contracts, I've realized the importance of stakeholder management and meeting customers where they are. So, the most interesting challenges to date has been around implementation and understand just how differently each company operates and co-creating a way forward to unlock Searchlight's value.
Thanks for taking the time to write up such thoughtful responses! 1. Ooh, I do recall Shishir's podcast conversation with Lenny Rachitsky on the topic of the rituals of great teams - thanks for sharing the Coda doc with me again! I agree it's so important to regularly have "health checks" on meeting health to make sure the relevant folks are having the necessary conversations without being bogged down by stale or useless meetings.Totally get you on the calendar management. Time-blocking and grouping higher-affinity work together helps a lot with keeping context-switching under control for me, too. That's such an interesting point re: wanting to be managed up. I never thought about it that way from a people leader's POV before - thanks for sharing! 2. You sound like you trust and follow your intuition. I think optimizing for learning and being challenged are great variables to focus on. :D 3. Managing expectations across the board is key, hehe. And I love what you said about "meeting customers where they are." It reminds me of one of the most important lessons I was taught as an Agile Coach - "meet people where they're at." It's so easy to treat different folks as an aggregate or average, but like you said, there are nuances and flavours between different B2C or B2B users/customers. We need to learn what will make the biggest impact in their personal and work lives, then structure our product and business models against that to best fit with their mental models.
Hi Anna! Congrats on your impressive career – love the product that you're building. I'd love to hear what it's like to run a company with your twin sister, that must be quite a unique experience!
It's quite an experience :) Every co-founder relationship is complex. The foundations of a great co-founder relationship are trust, candor, care, and accountability. What's amazing about building a company with my twin sister is that we already had these qualities in our relationship before we started a company together. It makes a lot of things easier when I know that I can rely on my co-founder completely.I'm also grateful that Kerry and I's skillsets are complementary. She leads GTM, and I focus on EPD. Beyond that, our Searchlight profiles will tell you that we have very different working styles so we're able to see the same situation differently and find the right answer in between. She is a visionary leader and there's no one else I'd rather go through this journey with.
Hi Anna! Congratulations on the amazing progress and funding — I'm so excited to see how you develop as a company.I'm a freelance writer in the HR tech space, and my (main!) question is around the intersection of tech and the human part of hiring.We've seen orgs (like Amazon) go with algorithm-based hiring in the past, and get things horribly wrong. On the other hand, research shows that humans tend to over-estimate their ability to choose a great candidate and go on bias — especially when they're not trained to interview properly. How do you tread this line between the need for hiring to be a fundamentally human process, and reconciling these data points to get a rounded view of a candidate that builds decision-making confidence — especially in a challenging talent market?And one bonus question: What's been your most frustrating hiring experience — either as a candidate or as someone making the hire?
Great questions! Love the bonus question, so I'll start there first.The most frustrating candidate experience I had is the reason why I started Searchlight. My twin sister and I were both applying to jobs, and we look the same in person and on paper. We had similar resumes: same school, same prior employers, but we have very different skillsets and working styles. Unfortunately, the companies we applied to pushed us through the same generic recruiting process. It felt so dehumanizing! We didn't feel understood, and we had no confidence that we would end up on teams and roles that would be the right fit for us.We started Searchlight to help candidates feel understood, and to help companies make hiring decisions that will build amazing teams.--The question around the intersection of technology and human decision making is such a big one! I agree that hiring is a human process, and I know that technology like Searchlight’s can assist us in making better decisions.Just some reasons why humans make bad hiring decisions sometimes: inexperience, unconscious biases, too much information, not enough of the right information. In a challenging talent environment with extra pressure, human decision-making gets worse.To successfully tread the line, technology should focus on removing the barriers to making good decisions.Here are 3 ways that we focus our efforts at Searchlight:1. Efficiency. We know that a soft skills mismatch is the main reason why new hires fail, so we bring this essential information into the hiring process without manual work. Our product and engineering team prioritizes a fast, online experience that is supportive and engaging to candidates, while creating structured, quantified data.2. Effectiveness. Our software provides greater visibility on how hiring decisions really play out post-hire. Most recruiters and hiring managers don't have a feedback loop when they hire, so Searchlight helps them proactively surface the positive indicators (and risks) to employee success. This helps decision makers focus on the most important things that actually connects to outcomes.3. Explainability & Equity. We vet the way that we collect the information and our algorithms to have no adverse impact towards any group, which supports more equitable hiring decisions. Also, a lot of AI out there is "black box" and this leads to low accountability. We strive to show the reason why our AI provides the recommendation that it does.I can't emphasize enough that I believe tech should supplement and support the humans in the hiring process! When humans and technology can work together, it's better for businesses and the candidates who are seeking the right job.
@annawangx How do you deal with when your algorithm misbehaves? Do you have an internal mechanism for dealing with incorporating new data from users?I’m actually working on an AI ethics white paper related to this topic (funded by a grant from Notre Dame) and would love to interview you for it, if you’re up for it.
Hi! Yes we have internal mechanisms in place that continually audit our algorithms to ensure no adverse impact. This ensures that as new data comes in, we aren't seeing our algorithms impact different subgroups differently.We also proactively vet our assessment methodology before we go-live. Every trait that we assess for is verified to be job-relevant and our assessments are checked such that no demographic group is less likely to have a job-relevant skill.
@annawangx Very cool to hear you have auditing practices in place, especially since regulations are changing on AI in HR (like the NYC bias audit law coming into effect Jan 2023). Are your assessment methodologies and auditing data publicly available or for internal use only? I'd love to read more about it!
thank you so much for these thoughtful answers — I really appreciate the detail and I've really loved learning about all you're doing at Searchlight :)