28 minutes 20 seconds
🇬🇧 English
Speaker 1
00:00
To our next speaker, Kieran Snyder, who is the founder and CEO of Textio. So I actually started hearing about Textio last year from a number of YC alumni who use and love Textio. So they use Textio to analyze their job postings. So, Textio is used not only by startups, but also by large companies like Johnson & Johnson and CVS.
Speaker 1
00:27
And basically, they analyze data from job postings from around the world, and they can tell you which are going to be more successful. So Textio has been around for about 3 years and they have raised 29500000.0 dollars. Yes, which is amazing. And what I thought was really cool, a really cool detail about them is that about 50% of their team is female.
Speaker 1
00:53
And female or non-binary and their board is over half female as well. So before, yeah. I think that does deserve a round of applause. And so before Textio, Kieran held product leadership roles at Microsoft and Amazon.
Speaker 1
01:10
So I'm really excited to have you meet her today. And so welcome, Kieran.
Speaker 2
01:18
Hi. How many of you had heard of Textio before today? That'll help me level set. Oh, that's really exciting.
Speaker 2
01:25
So I'm Kieran. I am the co-founder and 1 of the founders and the CEO of Textio. We're gonna go back a little bit through my journey, getting to Textio and a little bit of our journey as a company. I'm gonna walk you through today, 5 really important decisions we made that have massively accelerated us.
Speaker 2
01:48
As Kat just mentioned, we're a Series B stage company. We've raised $29.5 million so far. And 3 years and 3 months ago, we didn't exist. OK?
Speaker 2
02:00
We were founded in late 2014. Thank you. So you can do this too. I'm going to show you.
Speaker 2
02:06
When you leave today, you'll be like, yes, I'm ready. So a little bit about me. Whoops. Too fast.
Speaker 2
02:15
There we go. Those are some old pictures of me. I grew up with a dad who is an engineer. He's 78 years old.
Speaker 2
02:25
He's still running his own company. He's totally my hero, right? He started it when he was 40. I started Textio when I was 40.
Speaker 2
02:33
So I've literally seen the entire journey of his small engineering company. My mom was a writer. And I grew up to be an engineer and a writer, and very much formative in what we do at Textio. So when I went to college, I double majored in linguistics and math, which was a really present combination that I didn't know would be present at the time.
Speaker 2
02:57
And then I continued on and I got a PhD in natural language processing. And I met someone else here today who is a graduate of the same program that I was in. So that's really exciting. But really, my whole life, in 1 way or another, from childhood all the way on up, was about the statistical measurement of language with math.
Speaker 2
03:17
I chose linguistics because I didn't want to pick between math and language. I've always really liked both. I've been coding since I was 3. I was really lucky to have a dad who was an engineer.
Speaker 2
03:28
So at my age, it was pretty unusual to have computers in your house when you were 3 because I'm in my 40s but we did. I won my first science fair project in fourth grade by writing an early dating application that in basic, for those of you who are old enough, where if you put in your name and answered some questions about you, it would say the name of the person in the class that you were gonna marry. Right. However, the people didn't know about if, then, and go to, so it was a really great coincidence.
Speaker 2
04:00
Whenever the boy I had a crush on put in his information. It said he was gonna marry me Teacher was a little suspicious, I think The other thing I did a lot as a kid was I played sports. I played sports forever I still coach my 3 daughters in basketball today. But sports and teamwork were a huge part of my upbringing, and it really gave me a major appreciation for teamwork and tenacity.
Speaker 2
04:27
So if you have daughters, get them into sports. It pays dividends forever. That's me as a little kid. That's me on the upper right in middle school.
Speaker 2
04:36
I've always had a flair for performance, with a mullet, by the way. Still was willing to do that. And that's me in grad school with, I don't know how much we had to drink at that party. But as I came out of graduate school, I didn't know what I wanted to do.
Speaker 2
04:54
I spent a year just writing and ran out of money. And I had a friend at Microsoft at the time who said, I really think you'd like working in technology. You should work in software. I was like, no, I hate all those people.
Speaker 2
05:09
I went to grad school with all those people. Not interested. He's like, okay, but you said you needed a job. You're out of money.
Speaker 2
05:15
Good point. And long story short, he was right. I am absolutely a technology and software person. So I was at Microsoft for about 8 and a half years and it was a really good experience for me.
Speaker 2
05:30
I was hired as a computational linguist building spell checkers and grammar checkers. By the time I left the company, my culminating project was leading Bing's native integration into Windows across thousands of people. So I really learned how to build software. Always with a product and engineering focus.
Speaker 2
05:46
So when it came to Textio, I understood product, I understood natural language processing and I didn't understand that much about starting a business. My co-founder and I, who I will talk about in a minute, were both technical co-founders. So everything to do with sales and marketing and business operations and fundraising was all really, really new to me. I started writing a lot when I decided to leave big companies.
Speaker 2
06:15
Didn't know what I wanted to do. Didn't know I wanted to start a company of my own. But a bunch of these got some press. I had started blogging quite a bit.
Speaker 2
06:25
And I started writing a lot about statistical measurement of bias. So I did a study when I was at Amazon. I don't know if I was a very good employee at Amazon because I mostly sat in meetings logging how many times men and women interrupted each other and then I published it in Slate. I don't know whether they were that happy.
Speaker 2
06:45
I published a piece right as we were getting started that in Fortune on bias in performance reviews and technology that was really fortuitous because it went hyper viral and it ended up formative in Textio's history. I blogged, I had little comic strips of me and my daughters that I drew on the blog. That blog is still there, but now most of my blogging is at Textio. But I was really writing a lot.
Speaker 2
07:11
So again, always an engineer, always a writer. When Jensen and I decided to start Textio, we had a really clear vision, okay? And I'll talk a little bit about his background in a bit. From everything we had both done, we had this idea.
Speaker 2
07:33
It had nothing to do with talent. It had nothing to do with hiring. It had nothing to do with bias. Still doesn't.
Speaker 2
07:39
But if any time you wrote anything at all, just think about this for a minute. You're writing a pitch to an investor, you're writing a sales email, you're writing a job description, which is probably how you know of us if you know of us. If you knew ahead of time who was going to respond, that would be super powerful. If any time you were writing, you knew exactly which aspects of your language were getting 1 group to engage while another walked away, then you could change your approach.
Speaker 2
08:12
You could get a different result before you published it, not after you published it. And so this thing we've come to call augmented writing, we didn't have that language 3 years ago, really means writing supported by masses and masses of data where your engine can find the patterns. Which words, which formats, which constructions get 1 group to engage while another walks away. That was really our founding mission.
Speaker 2
08:37
The very first thing we built before we had ever had funding was a Kickstarter predictor. And Kickstarter was a fantastic place to prototype our technology, because all the projects are public domain, and all the outcomes are known. So you can look at the successful projects and the unsuccessful projects, and you have a really good data set. And so that on the left is the first text to UI.
Speaker 2
09:05
Jensen, our CTO, drew that on a napkin when we were in San Francisco. And you can see the first prototype we built looks actually a lot like that. And About 8 weeks into the company, we had this prototype working. We got to 97% predictive in minute 1 of a project as to whether it was going to make money.
Speaker 2
09:23
A little depressing. The quality of your idea has no bearing on whether you make money on Kickstarter. None at all. As a longtime product person, that was actually a little demoralizing.
Speaker 2
09:35
But the fonts you use, how long it is, where you place your video content, how you caption your video content, Many of these things are highly predictive of success. So that was really interesting, and it felt like we were onto something really big, which is the way you communicate can sometimes even be more important than what you communicate in getting to a certain outcome. So 9 weeks into Textio, we raised our seed round with this. This was never going to be a business.
Speaker 2
10:06
It was never going to be like, we're ready to go out the door with a go-to-market plan of helping Kickstarter entrepreneurs. There aren't enough of them to make that a billion dollar business. But it proved that we knew what we were doing. And it worked.
Speaker 2
10:18
And so imagine you're in a room with a bunch of investors, and they all, by the way, are familiar with Kickstarter. They all have a favorite campaign. And you're like, so look, it works. And they're like, no, no, go find Robot Turtles.
Speaker 2
10:28
How does that 1 work? What would you have said about that? No, no, go find the potato salad guy. What would it have said about that?
Speaker 2
10:35
So you have to do this over and over again. But it worked. That was our first office in another co-working space that I will not name here. The Riveter didn't exist.
Speaker 2
10:46
And on the right up there, those are the monitors we bought for our team when we got our seed round of $1.5 million. We made our team work without monitors. We worked without a lot of things. It was a huge day for us.
Speaker 2
11:00
Like, yes, we bought monitors. 3 years later, we just moved into a 25,000 square foot space downtown. And we're like, ah, oh, man, this is big. But that is what we looked like in our beginning, which I'm sure for a lot of you who are just getting started is a lot what you look like, too.
Speaker 2
11:18
So I'm going to walk through quickly 5 decisions that either I made or we made together that I think have accelerated our growth. And hopefully, there's something useful in it for you. The first 1 may or may not be repeatable, but I started the company with Jensen Harris, who is our CTO. He's also my husband.
Speaker 2
11:40
So you may not be able to find that as a repeatable pattern. So That's our life, right? Up in the left is our professional headshot, which by the way, took a long time to get. All of our pictures look a little like family holiday photos.
Speaker 2
11:55
And those are our 3 daughters. So we have 3 daughters who are third grade, second grade, first grade. Yeah, we're not bored. Jensen's background is in all ways a compliment to mine.
Speaker 2
12:08
So while I have been deep in language and statistical modeling and building software forever, Jensen was a musician. He went to music school, but also, like me, had been coding since he was 3. Made his way through college building Shareware, which again, like 20 years ago, that was a thing. And really developed an industry-wide reputation around user experience.
Speaker 2
12:34
He invented the first UI for email ever, right, in the form of Outlook. Led the design and implementation of Office Ribbon commanding interface and then Windows for the whole world. So where I had this experience, you know, really understanding which, how do you use NLP machine learning to find patterns in language? I was really the back end sensibility of the company.
Speaker 2
12:56
He knew how to build a user experience truly for a billion people. We didn't have a billion people. We didn't have a billion people, we didn't have a million people, we still don't have a billion people, but you've got to aspire there, you've got to build it so it's usable by people who are not linguists, not engineers, not data scientists. The fact that we've done this together wouldn't work for every family, but for us it is an incredible foundation because there's no 1 I trust more, right?
Speaker 2
13:23
So if he picks something up, I know it's good. And if it's not, he's going to tell me and vice versa. And we have very different strengths. So that's been a decision.
Speaker 2
13:33
By the way, didn't make it easier to raise money the first time around. There's certainly some people who didn't want to invest in us because we had the family relationship that we have. Didn't make a difference in the A or the B. So just remember that.
Speaker 2
13:45
If you get over the hurdle, people look at what you can do together and you have investors kind of in your camp. So whether you're co-founding with your spouse or not, I really like what Kristen said in the last panel, co-founding is an amazing choice. It's solo founding. I look at what Amy Nelson is doing here with 3 babies under 4, and she's a solo founder.
Speaker 2
14:05
And like, hats off to you. That is hard. Co-founder gives you someone to rely on. And even if they're not your spouse, they're a little bit like your spouse.
Speaker 2
14:14
And you really need to have that trust. So that's kind of the first decision that really, I think, was how we got started that really accelerated us. OK. The second important decision was the decision to start with job posts.
Speaker 2
14:29
So as I mentioned, we got started. We built a Kickstarter predictor. It has nothing to do with talent or bias. But remember, I was publishing quite a bit about, really, HR documents and documents about people and bias.
Speaker 2
14:43
We knew and Still know that in 6, 7 years from now when we IPO, we want you writing everything with Textio. We want you writing every single thing at work with Textio. It is a bold ambition. We are going to take down Microsoft Word and Google Docs.
Speaker 2
14:57
And we can do it, because we've had the opportunity to run those products. You've got to start small as a startup. If you're making something that's really, really predictive, right, that actually when you type in a word processor, Kat talked about TXEO's analyzing job posts, it's actually more for writing job posts, you need to give guidance that specifically works for the kind of writing that is at hand. So we realized early we had a lot of opportunity, because I had a lot of relationships and talent in HR, because of work I was publishing, to do something really good for that audience.
Speaker 2
15:31
They had a problem. They're spending a lot of time and money recruiting and making recruiting content. They're not happy with how it performs, both from a time to hire perspective and certainly a diversity and inclusion perspective. They had data about the problem.
Speaker 2
15:46
So they have masses of job posts and they keep track of how those job posts actually do with real people. We had a technology. And so the decision to start with job posts was really essential, so focusing that broad vision in on a place where we could deliver value for companies. So you could actually prove that the platform works.
Speaker 2
16:10
We had unique relationships to get us started. So if we had started out in sales and marketing, which by the way, every investor that we first talked to asked us why we weren't starting there, because there's a lot of money in sales and marketing, we would have been choked out. As the head of an HR tech product, I get to go meet with CEOs of Fortune 100 companies, because they care about that problem. And it's a very underserved population technologically.
Speaker 2
16:35
If we had built something in sales and marketing to start, I would have been just another sales enablement CEO. So find the place that is uniquely you that You can win. We started here because we thought we could win. Deliver value.
Speaker 2
16:49
And at the same time, it's had this stealth benefit, which is, by the way, in all of our companies, so you think about a company like Apple or Cisco or Johnson & Johnson, because we're mostly an enterprise product, everybody in that company is using Textio because every job function hires. And so when we're ready with products for those other domains, they know us. In fact, 1 of our challenges is people try to give us data for other domains ahead of our desire to be in those domains. Because remember, you're a startup.
Speaker 2
17:16
We're still only 65 people. You've got to focus. You've got to do something you can do uniquely well and win. So for us, this was a really good place to begin.
Speaker 2
17:24
The series B we raised last year is the recognition that the platform works, and we're ready to start attaching it to other kinds of writing. And then, of course, there's the wonderful benefit that actually you can solve some really significant bias problems for companies, too, which is really important to us, as Kat mentioned. Our company's half women, our exec team's half women, our board is more than half women. We live this.
Speaker 2
17:44
Okay. Let's get through. The third really big decision we made is around rebranding. So in 2000, whoops, this is a skittish clicker.
Speaker 2
17:55
In 2014, that's what our logo looked like. And in 2016, we rebranded. And it really is a very stark contrast before and after. The 1 on the right looks like an enterprise product.
Speaker 2
18:08
It wasn't just about the look and feel though, it's also how we talked about the product. So our director of brand experience, when he arrived at the company, observed that our team, when we were selling the product, was talking too much about how easy it was to use. And it is easy to use, right? That's Jensen's contribution in history.
Speaker 2
18:27
But no 1 buys something because it's easy to use. No 1 gives you a quarter million dollars for something because it's easy to use. Like a skateboard. Well, a skateboard is a bad example.
Speaker 2
18:35
Not easy to use. Like, these shoes are easy to use. You buy it because of the value it provides to you. So he really made the observation that we were pitching the company and the product wrong.
Speaker 2
18:47
We were pitching about ease of use, which is a strength of ours. But actually, we should pitch the fact that when companies use the product, they fill their jobs 2 weeks faster. Or they get 23% more women into the pipeline for engineering and technology roles. That's why people buy the product.
Speaker 2
19:02
And so all of the rebranding and the way we talk about the product have contributed to an acceleration for us, because people read it differently. You buy it when you sell the value. So that was an important decision. The other thing I learned as somebody with a product background is that the product isn't only what you build.
Speaker 2
19:24
It's also how you package and sell it. So when we started out, The very first thing we did was we had an invite-only beta. We ran that for 12 weeks. Once we got 5 months into the company, we had built this beta for job descriptions.
Speaker 2
19:38
We had early prospective customers who were sharing data with us. We figured that out. The beta went well, and so we decided to commercialize. We didn't know anything about how to commercialize.
Speaker 2
19:51
So we did what a SaaS company might do is we're like, here go on our website you can buy a seat of Textio for $59 a month or something for yourself, which is really cheap for the value. People loved it because it was really cheap. And they went and they bought the product. And that seemed really exciting until we tried to get back into those bigger mid-marketer enterprise sales where they're like, well, we could just buy like 3 or 4 seats and we're good.
Speaker 2
20:20
Our team can share them. That's not good for building a scalable business. So we shifted the model. And today, the way we sell the product is with a flat fee according to the number of jobs you have open at the time you subscribe.
Speaker 2
20:36
So if you have 1,000 jobs open, you pay a flat fee for 12 or 24 months and you add as many people as you want. In fact, you should add everybody because the product will work better for you. As a machine learning company, that's super important to us because we're learning from more writers in the platform. And of course, in terms of our sustainable revenue, that's more important to us because we shift the value away from the individual writer to the value of the platform.
Speaker 2
20:59
What's it doing for your company? I can't even tell you, Again, I had no sales background at all. Our first 6 people were all technical at the company. I wrote a lot of the original back end of the product.
Speaker 2
21:10
But I was the only 1 who wanted to talk to people, which is how I ended up being the CEO. And that's certainly how I ended up being the first salesperson. So I ran sales, customer engagement, basically for a year before we hired our next person. And you can't imagine my shock the first day we received a paper check in the mail for $35,000 from my first enterprise sale.
Speaker 2
21:33
I was like, what do I even do with a paper check? It was 2015. I didn't even know how to get it into our bank account because it never occurred to me people would pay that way. It was like sitting around the office like a hot potato for a few days.
Speaker 2
21:47
We were all like, oh my god, someone's gonna lose it. It's gonna get, what are we gonna do? Most people don't pay with paper checks. Although that customer on their third year of renewal still pays with paper checks.
Speaker 2
21:57
Luckily is no longer my job to take care of the paper check. It's not where I excel. But this was a really important decision for us. Changing the way we licensed the product, you can see what happened in our revenue graph.
Speaker 2
22:15
You can see the before and after. You can see that we were going along, and then you can see, whoop, we went right up. And our customers were happier too, because it made sense to them. They got it.
Speaker 2
22:25
They didn't want to have to limit usage. That's not good. You want your whole recruiting and talent and hiring team to do it. So pricing it in a way that made sense to them, and by the way, allows them to not worry about who they're giving access to, really worked for us.
Speaker 2
22:40
So think about when you first are licensing your product. Remember that the way you price it and license it is just as big a part of the product itself as the actual thing you're making. Really, really important. We also figured out how to make it a better value proposition for our customers to share more and more of their outcomes data with us the longer their customers.
Speaker 2
23:06
So they have a lot of incentive to share data already because the product works better for them and if they do, right, so they don't have to just use everybody else's aggregate model. Today, Textio's job engine is about 350 million jobs with outcomes attached. So we're adding about 10 million a month. How do you do that?
Speaker 2
23:24
You do that with your customers. The customers are our deep partners in this. And this was a really important piece here, which is that they provide their data, the platform learns, the guidance in the platform gets better for them as they write, they use it, we all measure together again. It creates this virtuous cycle.
Speaker 2
23:41
And if they're data partners, they pay less to use the platform because they're helping us make it. They're part of our family. That's been a huge advantage for us. Finally, the press has been really kind to us since we started.
Speaker 2
23:58
I think because we got out of the gate early, showing some impact for diversity and inclusion. The press was really excited. I was writing quite a bit, so I had some connections. But 1 thing we learned over our first couple of years is that not all good coverage is actually good for your business.
Speaker 2
24:17
So this is what the first, some sampling of what like the first year of Textio Press looked like. These are great brands, right? You have the New York Times writing about you, Bloomberg writing about you, Fortune writing about you. They're all writing about Textio, which is awesome, in a really narrow way.
Speaker 2
24:36
They're not telling the story of augmented writing. They're talking about bias. And we care about bias. We collaborated with them on a number of these stories.
Speaker 2
24:45
Some of them are based on blogs that we did, but we realized about a year into the company that if we're going to deliver on the bigger vision, we need to change the narrative. We want these stories. This is important to us. But we don't want to be pigeonholed as a bias play.
Speaker 2
25:00
That's not what we are. We're writing play. So we worked really, really hard over the following year or 2, and Marissa, who's our director of columns, is here today and really led a lot of this work to change the narrative. These are a couple of examples of pieces we've had that have broken pretty big.
Speaker 2
25:20
We started talking about other aspects of hiring language. We started talking about what does it matter, what does it mean if you have grammar errors in your job post? Is it going to change their performance? By the way, it doesn't.
Speaker 2
25:30
I don't know if that's good or bad, but it doesn't. We have piles of data about the public job posts of every company in the world. And Textio knows the patterns that are culturally distinctive in those environments. So I know there's a lot of Amazonians here.
Speaker 2
25:44
But Amazon jobs are, let's see if we can see it, 33 times more likely to contain wickedly. And it's not all wickedly prime. But when you see things like, you know, maniacal, wickedly, fast-paced environment show up in job posts, it's not an accident if a bunch of people in your company are using the same language. It's your culture.
Speaker 2
26:07
So we've started telling bigger stories about writing and culture, and that's been really important for us. When I got to be on Bloomberg TV last year, I was talking about augmented writing, what it means for the future of digital writing. And yes, we absolutely talked about the fact that when Expedia's jobs are gender neutral, they fill 3 weeks faster. That's awesome, right?
Speaker 2
26:29
Not an accident. You reach the whole population. Of course, you find the qualified person more quickly. But people are interested in us now for the bigger story.
Speaker 2
26:37
So we started small, and we've been broadening as we go. And you can see, just in the last 6 months, since we really brought this term big, people are talking about augmented writing as the category. And it's the category we're creating. So this is fantastic.
Speaker 2
26:53
This is what we wanted. So hopefully there's something in here you can learn from. If I were Recapping, I only have a minute left. Pick that co-founder very, very thoughtfully.
Speaker 2
27:07
If they're not your spouse, they're going to feel like your spouse. In my case, I'm lucky, because he's actually my spouse and he's awesome. Start small. Pick a problem that you can not only make traction on, but you're confident you have unique skills to win.
Speaker 2
27:23
Because you know what, if you don't win early, you don't win. You got to pick a place, you know, that you can win. And then be really mindful of the fact that your product is also how you license it, not just what you build. And the stories you tell about yourself change how people see you, what they're willing to pay for you.
Speaker 2
27:43
And aim big. Don't let anybody put you in a kind of a small box. If you start out in a small box because you're trying to make progress, that's a choice you're making. It's not a choice that investors get to make on your behalf.
Speaker 2
27:53
It's not a choice that the ecosystem gets to make on your behalf. So thank you.
Speaker 1
28:03
That was awesome. Thank you. Thank you.
Speaker 1
28:19
You
Omnivision Solutions Ltd