Meet women at Opendoor leading the charge! Hear lightning talks on career leadership, design & strategy, data science, engineering & much more.
Like what you see here? Our mission-aligned Girl Geek X partners are hiring!
- See open jobs at Opendoor and check out open jobs at our trusted partner companies.
- Does your company want to sponsor a virtual Girl Geek X event in 2021? Talk to us!
Transcript of Opendoor Geek Dinner – Lightning Talks & Panel:
Angie Chang: I want to say hi to everyone. My name is Angie Chang, the founder of Girl Geek X…
Morgan Cole: And welcome to Girl Geek X Opendoor Dinner! I’d like to spend just a few minutes chatting with you all about how you can start or continue cultivating a successful career by prioritizing your own self-awareness…
Heather Natour: One of the things I love about Opendoor is seeing that demonstration of leadership every day with every single person I work with. And I’ve personally seen that leadership demonstrated by these particular panelists…
Annie Tang: …Really keeping in mind that execution matters. It’s not all about like coming up with cool ideas. We need to keep a high bar for what we do.
Maggie Moreno: There’s no shortage of good ideas, only people to make those ideas a reality…
Amy Yang: I’m going to give you a flavor of the type of problem and a project data scientists are working with at Opendoor. Specifically, this is a multiple hypothesis testing problem…
Sumedha Pramod: I’m going to start with an opener. We actually filled out a variety of customer experiences and dashboards, which spanned from educating the customer on buying and/or selling their home, all the way through to actually digitally closing on their home…
Angie Chang: Thank you all for sharing your insights and your journeys with us! Really enjoyed all the talks and the conversation about this is what leadership looks like. Cool. We are going live with our Opendoor Girl Geek Dinner. I want to say hi to everyone. I see people are joining us. Can you chat to us where you’re coming in from. I’m coming in from Berkeley, California. How about you, Sukrutha?
Sukrutha Bhadouria: Hi everyone. I’m Sukrutha. I’m dialed in from Yosemite this weekend.
Angie Chang: Cool. Awesome. Quick intros. My name is Angie Chang, the founder of Girl Geek X. When we started Girl Geek Dinners over a decade ago, I was the only female engineer at a startup, and I really just wanted to meet other women in tech.
Angie Chang: I started asking companies to host Girl Geek Dinners, so that we could go to different companies and hear from the women on stage about what they’re working on. And then also be able to meet other amazing people like yourselves, which we’ll be doing after the talks. I was able to meet people like Sukrutha. So Sukrutha, why don’t you just tell us a bit about you and what you’re up to?
Sukrutha Bhadouria: Yeah. I met Angie because I was looking for things to do outside of work and that’s how I ended up finding out about Girl Geek Dinners and Angie. Honestly, I think everybody is craving a network now more than ever, and this is part of why us doing it virtual makes it possible. We encourage you to have your respective company that you work at to sponsor a Girl Geek Dinner. We hope that at some point in the near future, we’ll be able to see you all in real life.
Sukrutha Bhadouria: By day, I also work at a large company, namely Salesforce, and we’re also transitioning back into the office. The way people are working right now is so different. We can work from anywhere, so we should also be able to network from anywhere. So yeah, I look forward to tonight’s content and the speakers are all amazing. Over to you, Angie.
Angie Chang: While we wait, we have a few minutes to say some things. I wanted to quickly say some things that we’ve done since we started. We have a virtual conference every International Women’s Day called Elevate and it’s March 8th. And it will be again, March 8th, in 2022. And all the talks are recorded and hosted at youtube.com/girlgeekx – you can actually find all those conference talks, all the Girl Geek Dinner events, and tonight’s talks. If you have to drop off, go make dinner, we totally understand that. Everything’s available on YouTube later in case you can’t stay for the entire hour or two.
Angie Chang: And pro tip, look at the playlists, because they’re actually categorized into different things like career journeys, management, engineering, machine learning, and you can dig into what you’re interested in and see what other girl geeks have spoken about over the years on those topics.
Angie Chang: We also have a podcast. So if you like to listen, like I do, we have two seasons, I believe, of podcast and we have another season coming out this summer, so stay tuned for that. We’ll have some new content coming out.
Angie Chang: And then we also are going to be contributing to our local community here in the San Francisco Bay Area and adopting a middle school/high school and really contributing to enriching and helping support the students there who are interested in STEM. So stay in the lookout for news on that.
Angie Chang: That’s another opportunity where we can see you and hopefully engage you with some students and get them inspired to stay in STEM. So a quick note, I want to kind of say, who is here tonight. I looked at our attendee list about half an hour ago, and I saw that we have about 45% of you, have over a decade of work experience.
Angie Chang: Often when I go to networking events, people always say everyone’s junior or they just got out of college or they’re looking for their first job out of a bootcamp. That might be true, that might not be true, but also at the same time, there’s a lot of really, I would call mid-career people, who are out there, and continue to come back to these events, so I really say thank you for coming back! And continuing to dig in and learn more about companies and the people that work at them.
Angie Chang: And I’m really excited that tonight we are going to be listening to the women at Opendoor. If you haven’t heard of Opendoor, it is a real estate startup company and I’m sure the women will be talking more about what they’ve been working on at Opendoor, so I’m going to turn it over to them, the experts.
Angie Chang: Our first speaker, the keynote speaker, is Morgan Cole. And Morgan Cole joined Opendoor in 2017, where she’s helped many, many customers transition to their dream homes and served as a people leader for sales and support. And she’s currently supporting learning development team as a senior trainer and curriculum specialist with an emphasis on instructional design, internal partner relations, and creative problem solving across multiple organizations. So when she isn’t navigating the world of L and D, you can find her spending time with her sour patch, pup, Arthur. So welcome Morgan.
Morgan Cole: Hello. Thank you so much for the warm welcome, Angie. I really appreciate that. I am going to go ahead and share my screen and then we’ll go ahead and get rolling. All right. Can you see that okay?
Angie Chang: Yes.
Morgan Cole: Very good. All right. Hello everyone. And welcome to Girl Geek X Opendoor Dinner! My name is Morgan Cole, and I’d like to spend just a few minutes chatting with you all about how you can start or continue cultivating a successful career by prioritizing your own self-awareness.
Morgan Cole: Now, before we dive in, I should share just one quick tidbit about myself. I am a words of affirmation girl. It is the love language that rivals all others, in my book. So that said, I am going to need your help just to make sure we’re all on the same page this evening. So if you all are ready to kick off this conversation, just take two seconds for me and go ahead and type the word yes, Y-E-S in the chat box at the bottom of your screen, just let me know you’re ready to rock and roll. Go ahead and type Y-E-S. Oh, they are trickling in. I like it. Very good. Very good. That was a test, and you all pass with flying colors. So let’s do it.
Morgan Cole: So a few years ago, Dr. Tasha Eurich, who is a best-selling author, she’s a psychologist and founder of the Eurich Group, her and her team, they conducted a study with nearly about 5,000 participants. And the purpose of this study was to better understand the meaning of self-awareness. Here we are. The research team’s findings, they were actually pretty astounding, they learned that there are actually two types of self-awareness. The first one is internal self-awareness and the second is external self-awareness. Now, the thing to note here is that these are not mutually exclusive, meaning it is possible to possess one or both types.
Morgan Cole: Now, before I dive into each type of awareness and the role that it plays in our professional and in our personal lives, I am curious to know your thoughts on this next question. So if you look on your screen here, you’ll see the question reads, how many people do you believe are actually self-aware to some degree? Do you think it’s A, 10 to 15% of people, B, 15 to 20%, C, 20 to 30%, or D, 30 to 40%? If you had to guess, how many people do you think are actually self-aware? Go ahead and type in A, B, C, or D in the chat box for me and let me know your thoughts.
Morgan Cole: Oh, a handful trickling. It’s a mixed bag. Very good. Thank you all so much for the responses. I appreciate that. So while you all are still continuing to put in your choices, I’ll tell you the not so fantastic news is that the average human believes they’re self-aware, but only 10 to 15% of those people actually fit the criteria.
Morgan Cole: The good news is that self-awareness is a learned behavior. What that means is that we can strive to inch just a little bit closer to fully understanding how we tick and what truly motivates us and how to dissect the depth of our perceptions of the world around us. And it also aids in stronger leadership competencies too.
Morgan Cole: Going back to the two types of self-awareness I spoke about a few seconds ago, let’s explore internal self-awareness first. Now internal self-awareness, it represents how clearly we see our own values and passions and aspirations, your thoughts, your feelings, your impact on others.
Morgan Cole: Studies have shown that this type of awareness correlates with higher job and relationship satisfaction, as well as just general happiness. Meanwhile, external self-awareness, it represents understanding other people’s perceptions of our value systems and our thoughts or our feelings, right?
Morgan Cole: Essentiall,y folks that drift toward external self-awareness, they typically understand how others view them and they’re more skilled at showing empathy and taking in other people’s perspectives as a result.
Morgan Cole: The takeaway here is that it’s most impactful to try to strike a balance between both subtypes of awareness, rather than over-indexing on one or the other, because here’s the truth, being crystal clear about who you innately are, your own behavior patterns, and what you need and want is form of leadership.
Morgan Cole: Self-awareness – it catapults your ability to clearly articulate your desires and ask for help in forging the appropriate path to get you there. This next slide, it’s a quick map that actually breaks down four self-awareness archetypes, which is basically how we present to the world based on the depth of internal or external self-awareness that we possess. I’m going to save you a bit of time, and I’m just going to give you a quick overview of these four categories. No need to read through each line.
Morgan Cole: The top left quadrant, it shows how high internal self-awareness and low external self-awareness is typically referred to as introspective. It depicts people who clearly understand themselves, but they rarely challenge their own views. And in some cases, this particular quadrant can limit their interpersonal relationships.
Morgan Cole: Now, the bottom left quadrant symbolizes seekers. These are folks with low internal and external awareness. And people that are currently navigating this quadrant, they might be in a state of self-discovery and may perhaps be a bit unsettled or in a state of flux in their personal or professional lives. Now going over to the bottom right quadrant, low internal and high external awareness, those are telltale signs of a people pleaser. This means that someone may be hyper-focused on how other people view them, sometimes to the detriment of their own personal or professional contentment.
Morgan Cole: And in many instances, folks that work through this category, they’re known to make decisions that are not always in service of their own success. And in some respects, it can be considered a self-saboteur. And last, but certainly not least, at the top, right quadrant, that depicts high internal and external awareness, which insinuates that a person is keenly aware about themselves or the external environment and they value candid feedback from other people. There was a study specifically from Gallup that show people in this particular category are generally proven to be good leaders in a plethora of environments because they intentionally seek out balance and inter and intra personal skills. So those are the four archetypes.
Morgan Cole: Now with those four archetypes in mind and in the spirit of bravery and transparency and leadership, I would love for you all to just take a few seconds, just to think about which one of those four categories you believe you are currently in, in your career. Go ahead and just write it down on a sticky note, or perhaps the note section on your phone or a piece of scratch paper, whatever you have nearby in your home. And this is only for your personal use. But take a look at those four categories and jot down which category you believe you’re currently in. Now, I wouldn’t ask you all to do anything that I wasn’t willing to do. So in my case, I’m actually going to share my self-awareness experience aloud. So full transparency, I toggle between being aware and the people pleaser. I do.
Morgan Cole: And I’ll share an example that clearly depicts this. Two of my former leaders at Opendoor, they taught me very early on in my career that in order to gain sustainable success, it was going to be my responsibility to always ask questions, to always raise my hand for help and to speak up and finish what I started. But here’s the only problem with that, I interpreted most of those tasks as signs of weakness, almost like a bird’s eye view of my professional inadequacies or inefficiencies.
Morgan Cole: And let me just tell you this, thank goodness for patient and nurturing leaders, because it took about two years for me to really get over this hump. And while it sounds a little half witted for me to say now, I truly believed asking questions or raising my hand for help would inadvertently highlight the things that I had not yet mastered. And my aha moment was, “That’s the point, Morgan. That’s the point.” I needed to acknowledge the things that I had not mastered because you can’t fix what’s hidden and you can’t practice what you refuse to acknowledge.
Morgan Cole: So when I finally came to terms with what was holding me back, myself, I held onto this quote from Thomas Edison that I love, and it reads, “Having a vision for what you want in life is not enough. Vision without execution is hallucination.” So while I could go on and on about this topic for days, I do know that time is of the essence, but I want to leave you all with a parting gift. So I challenge each of you to take about 30 seconds, and I want you to recall a book or an author or a podcast or perhaps maybe even an article that has been especially helpful in building your leadership skills or self-awareness or interpersonal skills, anything that has helped you in your career.
Morgan Cole: Go ahead and take a few seconds to just think of that one resource, or maybe there’s a few of them that you always go to when somebody asks you for a recommendation. And once you have that resource in mind, I would love it if all of you can go ahead and type it in the chat box below for me. And while you all are typing and putting in your resources, I’m going to share three books that have been especially impactful for me.
Morgan Cole: Now, the first one is called The Big Leap by Gay Hendricks. It’s really a good book about getting out of your own way. And it was really helpful for me about a year and a half ago. The second one is a person that we all know and love, Brene Brown’s Dare to Lead. It’s a classic. I would venture to say reading it once a year is always a good idea. I always get gems from Dare to Lead. And the last book that I loved is actually called Get Over It! by Iyanla Vanzant. It’s a really great self-help book that talks from a professional and a personal standpoint about removing yourself from yourself, so that you can present your best self when you are in a professional setting. So just to recap, my top three books, The Big Leap by Gay Hendricks, Dare to Lead by Brene Brown, and then Get Over It! by Iyanla Vanzant.
Morgan Cole: Now, let me see what you all have in the chat box here. I’ve got a couple coming in, The Power of Gentleness. Ooh, very good. Thinking Fast and Slow. Thank you all. Continue to go ahead and put them in as you see fit. And as books and articles come to mind, please continue to trickle them in for me. This is actually my first time reading about some of these titles. So thank you. Thank you to everybody who’s sharing so openly. This chat is chock-full of resources, and it’s an endless gift to each of us. There’s an African proverb that says, if you want to go fast, go alone, but if you want to go far, go together. And tonight’s conversation and your willingness to offer resources to cultivate your peers will continue to build an equitable bridge between different ethnicities, self identified genders, and neurological differences in this ever evolving world of business.
Morgan Cole: So take that sticky note or that note on your phone with your self-awareness type on it that I mentioned earlier, and the plethora of resources that you now have in this chat box, and I would love for all of you to use it as a first step to discovering how you can continue to evolve into the best version of yourself in business and in life. So thank you all so much for your open hearts and your listening ears and collaboration and huge thanks to Girl Geek team for cultivating a platform that acknowledges and celebrates women in tech. I am so, so appreciative.
Morgan Cole: For our next speaker, I would love to introduce the one and only Annie Tang. Annie is a Senior Design Manager for Seller at Opendoor where she works on drastically simplifying the home selling experience. And in her time at Opendoor, Annie has worked on designing various aspects of the Opendoor consumer experience like trade-ins, buying and mortgages, but outside of that, she loves to hang out with her sweet pup. So without further ado, Annie, I’m going to pass it off to you.
Annie Tang: All right, here we go. Sorry. I don’t use Zoom every day. Oh my gosh, Morgan. That was such an amazing and inspiring talk. This is actually not my first time hearing it from Morgan, but actually I always feel inspired the second time around too and it really got me thinking about the internal self-awareness piece, especially because I think as we think about self-awareness, it’s easy to think about the external piece, at least for me. And so this really got me thinking about the internal piece. With that, I’ll transition over to my talk for the day and that is about design and strategy at Opendoor. So like Morgan said, I’m the senior design manager for Seller. Seller is one of our teams here at Opendoor and I manage the team of designers that work on that experience for our customers.
Annie Tang: A little bit about me. I started out at architecture, so did not study UX design at all. And worked at a couple of various larger scale companies before I found my way to Opendoor. And the reason why I joined Opendoor about four and a half years ago was really I wanted to work on a very complex real-world problem. And at the time, it was a pretty small startup and I was really excited by the opportunities that it gave.
Annie Tang: But most importantly, I was really excited to design for online and offline experiences where I wasn’t really selling an app or an interface, but I was actually thinking through selling a service that included both the digital experience and a real world component to it, and that felt really exciting to me.
Annie Tang: And so today, I’m going to talk to you a little bit about design and before I get into it, I wanted to get a signal maybe in the chat, you can put it in the chat if any of you guys have been watching Mythic Quest, it’s been a favorite around our house lately. Angie definitely has. There’s a couple other people. I personally am loving it. My husband and I really love to watch this show. But one gripe, I will say, that I do have about this is that it really centers around this myth of the creative genius where you’ve got this creative director who over the course of the night comes up with these amazing visions for the new game and the art director and team just creates it and there’s no research or anything and makes for great TV!
Annie Tang: But unfortunately that is not really how design actually works in real life. And so over the course of designing and my design career, what I’ve realized is core for design and for product design actually, and creating products is that the genius-ness, the coming up with the ideas isn’t actually the hard part.
Annie Tang: And it’s a good thing and a bad thing. It means that even if you feel like you’re not an ideas person and you don’t have that, it’s not the end-all be-all to being a designer and also at the same time, being a designer doesn’t mean that you’re just the one coming up with ideas and other people execute.
Annie Tang: A lot of being a designer is validating the idea, effectively communicating across the company and figuring out how to build it out. And so I’m going to go over that little bit with a two-part agenda.
Annie Tang: I’ll talk through some of the principles that we have within design team at Opendoor that help guide us to make sure that we’re really being diligent about how we design for our customers. And then I’ll also walk through a case study for what we did for our Seller experience last year where we designed the experience end to end when COVID hit and really helped our customers figure out a way to sell their home faster and on their own time.
Annie Tang: Principle number one that I have with our team is we always start with research, we never forget the data. Every idea should be backed by reasons why people’s lives will be better with it. What we do is we spend a lot of time talking to our customers, our users, to discover problems and validate possible solutions before we even get into any ideating phase. It’s really important for us to really empathize with our customers and really understand what those needs are.
Annie Tang: A couple of ways that we do this, at a high level is qualitative research and quantitative research. On the qualitative level, that’s really talking to our customers, doing user interviews. An insight that we might get out of qualitative research could be something like I have here, which is, meet Linda. Linda’s looking to sell her home and buy a new one. She is completely overwhelmed with the process of coordinating two transactions to line up her move, and that’s the buy and sell transactions.
Annie Tang: We’re talking to customers and really getting at what is really difficult for them in their process. Quantitative research is really about sizing this opportunity. So it could amount in something like this, like Linda, 70% of home sellers in the United States are simultaneously buying and selling a home at the same time. This basically tells us this is a real need. And actually, a ton of people are experiencing this need. And so that really centers around a problem that we can obsess over then that we can ideate upon. So the second principle that we really adhere to is first visualize the experience, not just the UI.
Annie Tang: I’ve worked with tons of designers over my career, designers junior and senior. I’ve seen so many folks, when we get a new prompt, or we get a new idea, we immediately hit the pixels and we design out a shiny experience, and it’s really, really amazing.
Annie Tang: We forget that actually the customer needs to be the star of the story. It’s not about the UI. And so what we really emphasize is when we start thinking about new ideas or solving problems, we think about the story, we think about the customer and how they experience the flow from end to end. And sometimes when we do storyboards like this, sometimes we do flow charts, but it’s really about putting the customer at the forefront of the story first and then the UI and the pixels fall to support that.
Annie Tang: Here’s an example of some work that we do when we put together flows and comps. You’ll see that we have the digital experience, but also first, we have images that support kind of telling a story of what happens in real life. We’ve got text updates and someone on the couch receiving them, we’ve got a walkthrough prep and imagining that a customer is cleaning their home and getting ready for a walkthrough before they log onto their mobile app to do that walk through. Really putting it situational with the designs is really core to how we try to think about designing new products.
Annie Tang: And then the last principle that we have is really that execution matters and we sweat the details. Again, design isn’t just about the idea and the strategy, equally important is the craft and the execution. Actually, a lot of times what I’ve seen is the final idea that gets executed might not be the most novel thing, but if we execute it really well and really diligent about it, and we track it, and we learn from it, that’s really what makes a product successful. What we really enforce is every detail, every pixel of experience matters, not just the strategy, also exactly how we execute it so that we’re delivering a high quality experience to our end customers.
Annie Tang: With those principles, I’m going to walk you guys through a tactical case study to kind of bring this to life and show you guys at a high level how we go about big design projects at Opendoor. What this case study is at a high level, I’m going to walk you through how we ended up designing a centralized dashboard for our Opendoor experience, where the design team created a vision to unify several parts of our experience, which guided a lot of the product roadmap throughout 2020.
Annie Tang: Starting with the problems and the research. What did we find? In 2019, throughout the course of the year, what we found is that our experience that Opendoor was just really disjointed. Customers were telling us that they were getting dead ends, that there was a lot of long wait times in between parts of the experience, and one thing that we give to our customers is they come to us and we give them an offer on their home.
Annie Tang: Starting with the problems and the research. So what did we find? So in 2019, throughout the course of the year, what we found is that our experience that Opendoor was just really disjointed. Customers were telling us that they were getting dead ends, that there was a lot of long wait times in between parts of the experience, and one thing that we give to our customers is they come to us and we give them an offer on their home. There was a really long wait time and they didn’t know what was coming up next. There was a lot of scheduling coordination going on. There’s a lot of things happening and a lot of dead ends and people didn’t really know what the next step was, and that was really impacting our customer experience.
Annie Tang: At the end of the 2019, what we did is we got together a group of cross-functional leaders, PMs, designers, engineers, and operators and we did a sprint over the course of a week. What we really wanted to do was figure out a strategy to solve those problems that we have identified from our research team and try to help to come up with a visualization of an end state, a future state that we want to aim towards by the end of the year that would solve all the identified problems that we had, and that would hopefully guide our work. And that way we can break it out into chunks that we can kind of slowly build towards over the course of the year.
Annie Tang: And the output of that sprint, what we created was a single narrative for that in-state. So we actually did, was we created a deck that included pieces like the images that we see on the right where we actually just laid out a story for our customer, Jill, who was looking to buy and sell, and we put together a couple screens, but really focusing on the story and how she feels and what she’s interacting with. We created this vision deck, and we actually tested a couple of these high-level concepts with potential customers.
Annie Tang: What we aimed to do is we really want to show strong concepts that were new to the experience that we could refine at a later time, but it was really about thinking about this new experience for our customer (Jill) that would solve all of her needs by the end of the year.
Annie Tang: A couple of the big ideas that came out of this was the idea for the Opendoor dashboard where we would centralize all of these disjointed experiences into one place where our customers can come back to see what’s next.
Annie Tang: But a couple of different ideas too, was like this idea of an instant offer. Previously we were having customers wait 24, 48 hours. What if we can give them an instant offer? As they were telling us information about their home, we can update their offer live.
Annie Tang: What if we could give them clear milestones? We always internally call it pizza tracker kind of like the Domino’s pizza tracker, but what if we could make it super clear like that for every stage of this buying and selling process on their dashboard?
Annie Tang: Another big pain point was that customers were having to schedule these inspections and figure out how to line up having people come to their homes. What if we leverage technology and help customers do self guided inspections where they could just upload a couple of photos of their home and we could do the inspection without having to actually go into their home?
Annie Tang: All of these ideas culminated into this vision deck and what was really cool that came out of it, is once we had an aligned vision where we were wanting to go towards for the end of the year, we could then formulate a roadmap and piece off different projects that each team would then take to work towards that vision. And we could get really tactical and figure out what the right way to execute towards it would be.
Annie Tang: And the great thing about this is that the sub teams then had a more or less unified idea of where we wanted to head towards and build towards for the year. So these are just some screens about how we sweated the details. We started from each project then, went through various rounds of research and low fidelity mocks on the left side to high fidelity mocks in the right side. Ultimately, whittling down to one experience that we ultimately shipped. I have Q and A on here, but we’re actually going to save Q and A until the very end, but that’s it on an example of how we design at Opendoor, both on a principle level and also in terms of case study.
Annie Tang: And with that, I’m going to introduce Amy. Amy is a Senior Data Scientist on the advanced analytics team at Opendoor. Her responsibilities include defining and leading pragmatic, casual learning practice at a company level using advanced experimentation and decision science techniques. So welcome Amy.
Amy Yang: Hi everybody. First, thank you, Annie, for the great showcase of all the cool design work your team have been working on. I’m a senior data scientist at Opendoor. I’m going to give you a flavor of the type of problem and project data scientists are working with at Opendoor. Specifically, this is a multiple-hypothesis testing problem. As you all know, this multiple-hypothesis testing problem means when we do more statistical tests, we’re more likely to make a type one error, which is a false positive discovery.
Amy Yang: Here is an illustration where we test 20 kinds of colors of different colors of jellybeans and how they show a correlation with acne. One out of the 20 tests will show a significant correlation, even though it’s purely out of random noise. That’s telling us whenever we do statistical tests, we want to control the overall type one error rate in case there is cumulative inflation by… The more tests we do, the more likely we are going to see a significant result.
Amy Yang: At Opendoor, the problem came up frequently specifically for product improvement, sometimes we want to track multiple outcomes of interest, not just one. When we want to evaluate the effect of certain product change, long-term or short-term outcomes, we are encountering this problem. Or if we want to dissect the data set into multiple subgroups and then do a statistical test within each subgroup, we are encountering the same problem.
Amy Yang: Sometimes we want to run an experiment with multiple treatment groups, not just one case which is one control group., where we have multiple treatments, we want to test each one compared to the other one, which one give us a [inaudible]. So other scenario we’ll generate this multiple hypothesis testing problem. If you look at the literature, you will find out statistically, there are ways to control this type of type one error rate inflation by adjusting your p-value.
Amy Yang: For example, some common measure you will see are Bonferroni corrections or a Holm’s method, false discovery rate control. Those are all thoroughly researched statistical methodology to control this problem. However, practically we have some objections when using this type of statistical method.
Amy Yang: People will say p-value adjustments are actually pretty arbitrary by the number of tests we are going to consider. How do you decide what’s the correct number of tests we are doing to adjust? Should we adjust tests we have done in the past? Should we adjust tests have been done in other teams but not specifically our teams.
Amy Yang: This number becomes very arbitrary and sometimes people will use this arbitrary concept to falsely adjust the total number of tests. The other objection is when we reduce the type one error, we are inherently increasing our type two error rate, which means we don’t have enough power to detect a significant true result which also means we are going to increase our total sample set.
Amy Yang: To solve this controversial problem as a data scientist, we come up with a very practical recommendation and strategies not only for researcher and the data scientists, but also for leaders and stakeholders who are going to review and read those report.
Amy Yang: For example, we would recommend not just focus on interpreting the p-value part, but also focus on the true magnitude or the effect size of the finding from the data. Also, we want to pay attention to the quality of the study and the data set.
Amy Yang: Focus on more from the design and the data quality side of the report and the study, not just purely based on the p-value from the study. For researcher and the data scientist, we come up with a set of statistical methodology recommendation, for example, when handling correlated outcome or metrics.
Amy Yang: We have this index method that can utilize the correlated information from multiple outcomes. Try to aggregate the common information and reduce the type one error rate, or we have this Bayesian multilevel modeling method. Completely move away from the frequent test p-value based decision making process and move to a more Bayesian probabilistic recommendation system.
Amy Yang: Practically, we also give recommendations. One of my favorite recommendation is rewrite the error rate into family-wise error rate control system based on theoretical related test groups. Let’s say you have two set of tests. Three tests, all measuring user satisfaction. You may have different metrics, but you are going to run three tests, they are all follow the user satisfaction category. You have another set of tests which are testing the total error rate or page load speed, which can fall into the safety metrics category. In this case, instead of submitting the total error rate across all the tests equally, you can divide them by the two family. So each family can share their own overall error rate.
Amy Yang: Now that’s just one project data scientists are working on at Opendoor. I also want to use this opportunity to introduce some of the other projects data scientists work with at Opendoor. Why is Opendoor investing so much on data quality and data rigorous and that data science role? It’s because Opendoor business is really unique and it’s very complex.
Amy Yang: If you think about housing transaction, it’s very important, and maybe you only do one or two housing transactions in your whole life. It’s very complex, the transaction process. Second challenge is our data is super sparse. Due to it’s a rare event, we don’t have repetitive interaction with a customer. Sometimes we only serve the customer once or twice in their whole lifetime. It brings a lot of analytical and statistical challenge. We don’t have the luxury of the e-commerce or internet type of traffic.
Amy Yang: A lot of the decision we are making need to depend on statistical influence and statistical expertise at Opendoor. The last point is optimization. We want to produce the best user customer experience with constrained amount of time and constrained cost. We want to work within limited costs and trying to optimize within the constraint and the produce the best customer experience.
Amy Yang: I’m going to share with you some other typical projects our data scientist team work with, for example, in the buyer team, we study the local housing demand and the price elasticity, and we feed that information to our resale pricing team in order to better or more accurately price the resell price for the home we acquired.
Amy Yang: In the seller team, we try to target specific seller group and provide more customized seller experience by serving based on seller input. Their characteristic provide a different type of unique service. So that’s our optimization model our data scientist team work on. From pricing side, we want to understand how to combat risk, adverse selection, and the competition and build other factor and the macro information into our pricing model. So that’s an overall introduction for the data science team and some of the cool project we’re working on.
Amy Yang: I’m going to introduce our next speaker, which is Maggie. Maggie is a Senior Software Engineer on the sales and the support team at Opendoor. Maggie has been a part of a wide range of projects at Opendoor, and recently she co-designed and is currently implementing a new role-based access control system for the company internal tooling. Outside of work, Maggie is a competitive swing dancer. Welcome Maggie.
Maggie Moreno: Thank you, Amy. Hi. I’m Maggie. And I’m going to talk about role-based access control at Opendoor. In this talk I’ll go over what role-based access control is generally, what the design goals were for Opendoor’s RBAC system, technical design of our RBAC system, and some challenges and recommendations from our experience. If you aren’t familiar with the term role-based access control, I can almost guarantee you are familiar with the concept. At a high level, in an RBAC system what a user can and cannot access is based on their assigned role. There are three main data entities, users, roles, and permissions. Let’s take GitHub as an example.
Angie Chang: Hey, Maggie. Real quick. Can you share your slides? I don’t think we see them.
Maggie Moreno: Oh, yeah. You know, it goes to show all the preparation in the world…
Angie Chang: Perfect.
Maggie Moreno: Let’s take GitHub as an example. Can you guys see my slides now?
Angie Chang: Yes.
Maggie Moreno: Excellent. Thank you. GitHub. I am part of the backend infra team, which means that my role is an administrator on the web Repo, which means that I can access admin features like merging pull requests without all the checks passing.
Maggie Moreno: For Opendoor’s role-based access control system, we have some very specific design goals. As a result of going public in 2020, we needed to comply with the Financial Operations Act, widely known as SOX. SOX compliance can mean different things for different companies, but the most pertinent part of SOX compliance for Opendoor was showing that only authorized employees were allowed to perform financially sensitive actions.
Maggie Moreno: Other goals for the system included easy use of maintenance for engineers, straightforward management for IT, and peace of mind for security. Security was a bigger concern than usual for our new RBAC system because we would be trusting the system with protecting financially sensitive actions from both internal and external users. We wanted employees to follow a clear process to change role and permission assignments.
Maggie Moreno: In our design, we assigned roles to users in Okta, an industry standard authentication and authorization tool, bringing a lot of advantages for security and IT. We already have moved towards using Okta for authentication for our internal tooling and this change felt like a natural extension of prior work.
Maggie Moreno: We store our permission to role mapping in a separate service config, which means that changes were handled through GitHub. It also means that for every API we call a role to permission service Gatekeeper to check the users roles. API end points are tied with permissions directly in the code, which is easy for engineers to implement and maintain. And once we get the users roles from Gatekeeper, we check whether they match the permissions for that API end point.
Maggie Moreno: Our biggest challenge for this project we encountered in the design phase. We had a big challenge understanding what SOX compliance would mean for Opendoor and what actions we should take to limit access first. Like many companies, Opendoor has a few different deployment environments for our internal tooling, and one of our biggest questions was whether we could limit our RBAC gating to one of these deployment environments. After further exploration, we discovered that that was not the case and this significantly changed our design requirements.
Maggie Moreno: Additionally, this project stress tested how major engineering design decisions are made at Opendoor. Getting alignment on the critical design decisions was especially difficult given the lack of clarity on the scope of the project. If you happen to find yourself in a similar situation in the future, we recommend that you clarify and align on your RBAC project goals before you start the design process. We also recommend involving cross-team stakeholders early in the project and communicate to engineering management early and often if the project needs more resources. Thanks.
Maggie Moreno: The next speaker is Sumedha. Sumedha is a Senior Software Engineer at Opendoor on the seller core experience team. She builds out various tools and interfaces to help customers find a home selling experience which best suits their needs, whether that is listing or selling their home using Opendoor services. These range from dashboards to see their home value to a digital closing experience.
Maggie Moreno: She also manages and maintains Opendoor’s design systems and React UI component libraries, which power the Opendoor site and admin tools. Outside of work, Sumedha is an avid baker and always trying to find ways to fit more plants into her environment. Welcome Sumedha.
Sumedha Pramod: Hi. Thanks Maggie. Sorry for taking over a little bit early. Role-based access control and SOX compliance are so important now that we’re a public company. It’s really great to see what went on behind the scenes to make that all happen. Hi. I’m Sumedha.
Sumedha Pramod: On the Seller team at Opendoor, we actually filled out a variety of customer experiences and dashboards, which spanned from educating the customer on buying and/or selling their home all the way through actually digitally closing on their home.
Sumedha Pramod: As you can imagine, the UI gets increasingly more complex as the functionality gets more and more important. This means that customers need to be 100% sure that they’re trusting their home and their money with Opendoor and that the site that they’re on, and that they’re signing contracts on is legit. Delivering on these expectations while continuing to add new features requires pretty thorough testing and it also helps us as a company build trust, brand, and consistency across all of our customer experiences.
Sumedha Pramod: With different UIs, it’s not as straightforward to catch a lot of UI changes and things like colors, font sizes, mobile responsiveness, and all those little nitty gritty things aren’t as easy to catch. Especially when there’s a ton of engineers working on the same UI, it’s pretty easy to miss if one person’s change unintentionally impacts everyone’s experience. I spend a lot of time personally staring at my team’s customer experiences, so I tend to notice some pretty nitpicky changes, but that doesn’t mean that an engineer from another team has the UI memorized the same way. This is amplified even more with design systems and shared component libraries, because a lot of these components that are changing are used across tons of different experiences.
Sumedha Pramod: At Opendoor, we have a bunch of micro front ends and a single change to a shared component like the button here on the left-hand side can have an unintended side effect on pretty much every experience at Opendoor without people really realizing it. UI regression tests are one of the ways that we found to actually help mitigate a lot of these unintentional UI side effects.
Sumedha Pramod: It allows us to compare screenshots of the UI or visual snapshots against a predefined baseline. It really helps us catch a lot of these really nit picky things. And it also helps us QA the user experience in a way a lot easier. And it helps us QA against design expectations. Additionally, we can also develop a lot of these components in a completely isolated environment without any kind of network calls. We don’t have to worry about data loading or any of that, we can really focus just on the UI and the pixel perfect stuff.
Sumedha Pramod: At Opendoor, we use a combination of two frameworks called Percy and Storybook, which are two different open source tools that enable the development and documentation of UI components and also automating a lot of that UI screenshot testing. In this example, you can see that we have a dashboard being rendered, but what we can actually turn this pretty simple unit test into something that renders a component.
Sumedha Pramod: If you want, you can have it mock out some data calls, do all that stuff, and you can actually render that and test it against a specific baseline. UI regression testing or UI visual screenshots don’t actually replace your standard unit tests or smoke tests or integration tests. Since those actually validate the expected behavior and experiences, this is really just to focus on the nitpicky UI things, so things like the CSS changes that I mentioned earlier.
Sumedha Pramod: Also you have all those other tests to validate behaviors such as whether or not a [inaudible] pops up when a user clicks a button or typing something in an input field will enable a button somewhere else. And this isn’t just useful for engineering, it’s been extremely helpful when QAing new features and designs. It’s also greatly reduced the amount of back and forth as we’re launching new features with design where we’re like, “This is a little bit off, this pixel is a little bit off.” As we’re developing, we can send these over and we can make sure that any future changes isn’t actually breaking that experience.
Sumedha Pramod: All of these things combined allow us to develop really beautiful and seamless experiences for customers and really make one of the most expensive and biggest transactions in people’s lives a little less scary.
Sumedha Pramod: Now, I’ll turn it over to Heather for our next segment. Heather brings 22 years of engineering experience to Opendoor, having led teams at Lyft, Capital One and Blackboard. At Opendoor, she leads the engineering organization focused on the core product experience for home sellers along with growth initiatives and retail partnerships. Heather lives in Berkeley, California with her husband and two boys. Welcome Heather.
Heather Natour: Thank you, Sumedha. And thank you so much for presenting on the Storybook and Percy testing. I’ve personally seen the impact of that on our quality and productivity. And I’m really excited to host this next session which will be a Q & A with our panelists.
Heather Natour: I’d like to invite everyone back to come back on the screen. And while they’re doing that, I wanted to talk about leadership in this Q & A, and as an engineering leader, I believe we should be creating opportunities for leadership at all levels, whether you’re an intern or a staff engineer.
Heather Natour: And one of the things I love about Opendoor is seeing that demonstration of leadership every day with every single person I work with. And I’ve personally seen that leadership demonstrated by these particular panelists.
Heather Natour: I’d love to ask each of you first, what do each of you believe has contributed to your ability to demonstrate leadership at Opendoor? I think we have everybody on now. So Morgan, maybe we’ll start with you.
Morgan Cole: Sure. Thank you, Heather. I appreciate the question. There’s two things that come to mind for me. The first piece I would say is, it sounds pretty simple, but I have practiced the art of assuming good intent at all costs in every scenario. Because I think sometimes in leadership, it can be quite easy to become a little bit defensive because you want to do well and you want to show up correctly, and so I think if you operate from a perspective of no matter what’s thrown at me, I’m going to assume that this was thrown at me with good intention, it will help calm that defensiveness so that you can respond in an appropriate manner. So that’s the first piece.
Morgan Cole: The second piece I would say is the team that I’m on specifically has done a really good job of teaching me how to lead collaboratively. I think it’s super important that whenever you are leading, whether it’s a new project, whether it’s a team, whether you’re just building new relationships with other partners throughout your business, it’s vital, it’s paramount that you don’t look at yourself as the single source of truth, but rather you work in conjunction with the parties that are involved to make sure that you all are leading in the same direction. So the two that I would say is assuming good intent and leading collaboratively.
Heather Natour: Yeah. That’s so true and really insightful, Morgan. I really appreciate your thoughts. How about you, Annie?
Annie Tang: Hey guys. I am unable to start my video, but I’m here. If whoever’s hosting could start my video for me, that’d be great. If not, no worries. Oh, here we go. All right. I’m back. This is such a great question. I think that my ability to demonstrate leadership at Opendoor has really been stemmed from as a designer and as a design leader related to what I was talking about in my talk, really helping everyone at the organization really obsess over problems, over ideas, and really… One of the key values that we have at Opendoor now is to start and end with the customer, and I think a lot of being about a designer is building that empathy and really obsessing over our customers. And that ultimately gets you to obsess over problems rather than solutions, so I’d say that’s one aspect.
Annie Tang: And the other aspect is really keeping in mind that execution matters. It’s not all about coming up with cool ideas and cool visions and stuff. That at the end of the day, we need to keep a high bar for what we do. And so execution really amounts to making sure that teams are working really collaboratively, that designers and PMs and engineers are working collaboratively and are in sync, can we do these workshops to make sure that people are in sync? So I think the dual sense of just making sure that we’re obsessing over customers and problems, and also making sure that we’re really executing to high level is kind of my leadership style.
Heather Natour: Yeah. That’s really great. And I agree, I’ve seen so much of that at Opendoor, the collaboration, especially collaboration that you facilitated. And it’s certainly a differentiator, I think, how much Opendoor obsesses over their customers. So that’s great. Maggie, what are your thoughts on the topic?
Maggie Moreno: Yeah, I feel like one of the best things about working at Opendoor is that there’s no shortage of good ideas, only people to make those ideas a reality. Opendoor’s business has a lot of different facets and there are opportunities everywhere for taking on more responsibility. So for me personally at Opendoor, I just feel like getting more leadership opportunities and developing myself as a leader has mostly just been a matter of raising my hand.
Heather Natour: Yeah, that’s very true. And I think the business complexity at Opendoor is really interesting and absolutely creates those opportunities. I love that. Amy, how have you experienced leadership at Opendoor?
Amy Yang: Hi. Not only the complexity of the type of problem Opendoor trying to solve, but also I would say, I admire this whole industry is still very new and young, like a startup feel when you join Opendoor compared to… I used to work at a more mature, larger technical company. I can definitely feel the culture difference here. Not everything is perfect is set up already. You don’t see maybe a perfect data engineering team prepare a perfect data that can be consumed by data science team. A lot of times you need to do the heavy lifting and see where the gap is, where the problem is, not just complaining for the missing pieces, but actually propose a solution and just do it. So that naturally you creates a gap create an opportunity for emerging new leader, especially for me coming from an IC now transition into a leadership role.
Amy Yang: The other thing I want to mention is this one team, one dream culture that is core value for Opendoor. So when you see area for improvement, even not within your immediate team, maybe it’s a cross-functional team, but you can contribute. There is nobody will stop you and say, “No, just do your own team’s work.” There’s always collaboration opportunity and you can just extend your influence outside of your immediate team and service area.
Heather Natour: Yeah, absolutely. I think that’s spot on and the sense of ownership that everyone has and doing that as a team, it’s so much more powerful, it becomes real multiplier. How about you, Sumedha, how do you feel the Opendoor experience has allowed you to demonstrate your leadership?
Sumedha Pramod: Okay, sorry about that. Honestly, one of the best ways I’ve been able to actually demonstrate leadership and grow is every week, every other week we have these kind of architectural meetings, and it’s really allowed me to not just have ownership of the code and the surface areas that my team operates in, but really expand that beyond that. And gives us a lot of opportunities to propose and facilitate a lot of discussions that have impact beyond just your specific team. And so really getting to establish that level of ownership at a much, much, much broader level and really expand your impact across the company. Yeah, that’s been really one of the most unique and interesting ways that I’ve been able to really develop leadership here.
Heather Natour: Yeah. I’ve seen that there’s a lot of people and we really want to include everybody in that process, and I think it’s really elevated a lot of amazing ideas, much more long-term thinking and has really pushed the organization to the next level. That’s a great example. Clearly each of you have developed deep domain expertise, and so on top of that, each of you have considered the direction in which you grow and whether it’s moving to people management or focusing on deepening your multiplying impact.
Heather Natour: Were there specific things you considered in order to decide which direction to take your career? And maybe we’ll start with Maggie.
Maggie Moreno: Thanks, Heather. I have been thinking a lot about going into people management. So I’m working on that transition right now. About a year ago, I got the opportunity to be a tech lead for a team, and I found that performing the leadership role was a lot more rewarding than being an IC. So I’ve really been taking on pursuing that and getting involved with the crafting of our transition role and starting to craft those documents.
Heather Natour: That’s great. And I think your selflessness that you demonstrate every day is really a huge impact to the rest of the team. And so it’s really great to see you moving into people management. Amy, how about you, you mentioned you were recently transitioning more from IC to leadership, what were some specific things you considered?
Amy Yang: Yeah, I think I’ve summarized the decision making process. I consider two factors. One is what my strength is, two, what my passion is about, three, is what’s the company goal is. I think the perfect position is how the three factors can align the best. Sometimes what you want to do doesn’t really align with the bigger picture of what the company want to go. Either the long-term goal, I haven’t seen sometimes, especially technical, a very deep technical person, they just want to utilize specific technology or tooling but that doesn’t necessarily solve the immediate business problem. That is a misalignment. I think the perfect position are when everybody would try to evaluate what’s the best role is. Do you see opportunity or can the next position that help you align the three factor better?
Amy Yang: For me, I joined Opendoor relatively recent, end of last year. I joined as a senior IC. Now I’m in the tech lead position. I enjoy my current position. It gives me both the freedom to do some project roadmap planning management, and also stay close to the technology while I’m still learning about the business, but eventually, based how I feel, how I evaluate the three factor alignment, I will make my decision for the next step.
Heather Natour: Yeah, that’s great. It’s so great that you’ve gotten these opportunities so quickly and yet, you can always change your mind and feel supported in how you grow here, I think that’s great. Sumedha, how have you thought about your career direction?
Sumedha Pramod: For me, it came down to as simple as I really just like writing code, as nerdy as that sounds. Writing code and really spending a lot of time diving into our customer experiences, whether that’s from a product design or even engineering standpoint. Obviously, as I grow more as a senior IC, it’s definitely less about doing a lot of the nitty gritty code myself, but really figuring out how I can better enable those around me to accomplish whether it’s technical goals or OKRs or things like that. And then also, how can I set some technical standards around best practices while really starting to think a lot more big picture about our systems and those are the challenges that really excite me and made me really want to go down the route of becoming a more senior IC for sure.
Heather Natour: That’s great. It’s hard to debate not liking to code. Annie, you’ve provided leadership at Opendoor for quite a while, how have you thought about your career direction?
Annie Tang: Yeah, so I started off at Opendoor long time ago as a senior designer as a senior IC, and I worked on much of the end to end experience and various part of the experience. And then, about two and a half years ago, I moved to management. And I think one guiding principle that I really had about my career has always just been to optimize for growth because I am happiest when I am learning and challenged. And so I was honestly really happy being an IC. I loved designing, I learned a lot, especially going from designing, just digital experiences at previous gigs to Opendoor, where I was really challenged and designing these online and offline experiences. And as Opendoor grew, I got the opportunity to manage.
Annie Tang: And when there was the opportunity, I was actually really excited and I actually told my manager that I wanted this opportunity. And I think that’s one thing that I would advise everyone too, is I actually don’t believe there is a certain stereotype for ICs, like if you are this way, you’re a great IC or if this way, you’re a great manager. If you’re curious and you want to grow into one or the other, make it known to your manager. It might not be that immediately you can do it, but that’s one thing I always tell, especially females. If there isn’t a right or wrong answer, if you’re curious about something, just to say that, because that’s ultimately how I ended up in management. I was really curious about it. I saw an opportunity, I told my manager and he helped me work my way through it.
Annie Tang: And as our company grew, I got to scale our team and I have really found a lot of enjoyment in supporting my team and not being the hands-on IC. I also haven’t ruled out though that maybe in the future or next gig, I want to be an IC again. And so I believe that there’s a lot of fluidity to this.
Heather Natour: Yeah. That’s excellent advice. I wholeheartedly agree. I have similar experiences. Morgan, you inspired us with just your self-awareness talk. How do you think personally about your own career direction?
Morgan Cole: Sure. Well, let me just tell you Heather, my career has been nothing short of a jungle gym. I have been a senior manager, an entry-level IC, a senior IC, a lead, back to management, back to IC. All of that’s happened in the span of about six years, and in different sectors too, whether it’s sales, whether it’s marketing and advertising, leadership and development, or learning and development, rather.
Morgan Cole: My decision-making process generally speaking, is with regard to the skills that I like to nurture. That’s basically how I base my decisions. So the trajectory of my career, it’s based on the skillsets and the acumen that I hope to cultivate at a specific time. And whatever those skillsets are, I am going to look for a position or role in which it will specifically help me get to that next step, and I’m less concerned about what the title is and more concerned about what the end result could potentially be.
Heather Natour: Yeah, that’s great advice. It’s sometimes hard to take the time to even think about that, but I think it’s really important. I appreciate all of the panelists’ thoughtful answers and that’s a wrap for the presentation portion. I also wanted to thank the Girl Geek team for all you do for the community.
Heather Natour: Here’s my LinkedIn – feel free to connect with me and reference this event if you want to hear more about Opendoor. I’ll invite Angie back on the screen to share the next portion of the breakout sessions.
Angie Chang: Awesome. Thank you all for sharing your insights and your journeys with us. Really enjoyed all the talks and the conversation about this is what leadership looks like. We will be sharing some Opendoor jobs via email. Keep an eye out for that.. that’s actually an event survey, and also links to the Opendoor jobs!
Here is a list of recommended resources – crowdsourced by attendees to help each other evolve into the best versions of ourselves in work and life!
Short on cash? Here’s a helpful hint: It’s easy to find sites with used copies if you Google the book title. In the results, there’s even a section to see the library closest to you that has the book!
Like what you see here? Our mission-aligned Girl Geek X partners are hiring!
- See open jobs at Opendoor and check out open jobs at our trusted partner companies.
- Does your company want to sponsor a virtual Girl Geek X event in 2021? Talk to us!