“Admirer from outside of the cap table” is how I approached Kieran Snyder, Cofounder of Textio. That is to say, I DMed her on Twitter in 2016, impressed by the work she was doing but without a preexisting relationship. Thankfully the interest was mutual and we’ve had the chance to exchange thoughts here and there in the time since. When Kieran announced earlier this year she was stepping back from the CEO role at her company it seemed like a great time to (a) learn from her tenure and (b) promote her awesome weekly newsletter.

Hunter Walk: Textio, the startup you founded and CEO’ed until a few months ago, is almost 10 years old. When you think back to the founding vision/mission, what played out the way you expected and what was most different?

Kieran Snyder: We started off with a premise about AI and writing. Specifically, we thought that if you could use piles and piles of data to figure out which phrases, structures, and other linguistic characteristics performed the best with real audiences, then you could expose that data to help people write stuff that performed better. That premise turned out to be true, as we’re all seeing play out today with the rise of AI APIs that many applications are rushing to adopt.

Textio brought this vision to market with our first product, designed to help people write job posts that attracted diverse and qualified candidates. We had initially wanted to begin with a performance review product or a more conventional marketing product, but we started off in recruiting for very practical reasons: 1) We cared about the problem and 2) We were pretty sure we were going to win. We had some unfair competitive advantages in serving HR buyers: we knew them all. I was publishing quite a bit of original research about bias in workplace documents like performance reviews and job posts, a bunch of it went viral, and I got to know a lot of people who eventually became Textio customers.

However, we thought our initial focus on recruiting was a wedge – an important but small part of the enterprise communication market that we were going after. We way underestimated Textio’s stickiness within HR and how deeply HR execs would invest in Textio. We also underestimated how little we would want to fracture our go-to-market to explore other business writing use cases as the company grew. Today, you’d look at Textio and say it’s an HR Tech company. But that’s not really how we saw ourselves when we started.

HW: This was your first CEO run after a successful product management career. If 2024 Kieran could whisper something into the ear of 2014 Kieran, what would you have told her about the difference between leading a team vs leading a company?

KS: Being a CEO has a lot in common with being a product manager, except that the product you’re responsible for is the company itself. Because there are so many similarities, I often underestimated the differences, especially early on.

As a leader of teams within larger organizations, I was able to build phenomenal teams in terms of both delivery and culture. I did this in part by defining my team’s culture as being outside-the-norm; I worked hard to make sure that my team felt special in the context of the larger organization. My teams always had a recognizable identity and subculture. But when your team is the larger organization, you can’t use this strategy.

That being said, I’ve always been at my best as a leader when I embrace my passion for teaching and nurturing. It’s not an accident that I started out in academia or that I have coached kids’ sports for so many years. It’s also not an accident that, upon stepping back from being Textio’s CEO, I’ve built a sizable exec coaching practice working largely with early stage founders. I find it tremendously personally fulfilling to work with people to achieve their visions and see them grow.

As a CEO, I was at my best when I embraced this side of myself fully. I love leading workshops for the team on how to tell effective stories with data. I love working with customers to support them in meeting their goals. I even love leading a really meaningful performance feedback conversation. 

This work is my zone of excellence. If I could give 2014 Kieran one piece of advice from 2024 Kieran, it would be to center this much more explicitly in how I approached my job as CEO.

HW: nerd processor, your weekly newsletter, is great! A recent essay covered the ‘AI gold rush’ and as it related to startups operating in this area, very much ‘caution ahead’ in terms of building a sustainable, differentiated business. Are there specific paths/opportunities in AI that you believe startups are actually better qualified to take advantage of than incumbents?

KS: Isn’t this the trillion-dollar question? When it comes to AI applications, big companies have more competitive advantages than they ever have before, because they already own the workflows and data that can make AI features stickiest. Big companies are also rich with much more cash to fund AI investments and compute costs.

As I see it, startups have two major advantages. The first one is simply focus. Startups can build habits with customers in narrow wedge markets that might at first look too small for a big company to care about. However, the startups that will make this strategy work are not just building undifferentiated offerings by wrapping the same OpenAI APIs as everyone else. They are either building on some differentiated technology or more likely a differentiated data source – ideally a data source that is generated from within their own application, where the data is collected by default as people use their product. 

That leads me to the second advantage that startups have. They can design their products to be AI-native from the start, rather than having to bolt AI capabilities on as a layer on top of a foundation that was not designed for it. From the very start, they can design experiences and choose privacy policies that automatically collect proprietary data. Customers know what they’re getting when they first use the product and can make an informed choice.

Textio relied on both of these strategies for a few years before launching our first truly generative AI capabilities back in 2019.

HW: Textio is open to remote hires from a specific set of states which I found interesting. Is that administrative (those are places you already have to deal with taxes, payroll, etc) or was there another reason? How did Textio get smarter about remote teams over time?

KS: For the first several years of Textio’s life, our team was entirely co-located in Seattle. Jensen and I had colocation as religion, and our team was incredibly tight-knit. To this day, if you ask people who worked at Textio in 2018 what they most valued about their experience at the company, they will talk about the caliber of their coworkers and the energy they felt building alongside teammates in the office.

By the time the pandemic started, we had brought on a couple of sales people in New York and San Francisco, but we were still 95% colocated in Seattle. Like everyone else, we went distributed overnight in March 2020. When it became clear that we weren’t heading back to the office right away, lots of our team members wanted to relocate to be closer to loved ones. Most people wanted to go to a specific handful of other states, so we opened those states.

Since we were already doing business in those states, it made sense to start hiring new people there too. In the medium term, we ended up prioritizing a combination of places where current employees wanted to live and those where we could tap into a great local talent pool. And that’s still where Textio is today. We’ve talked about using a PEO to open up even broader geography at scale, but the work to do this hasn’t made the prioritization cut because we can already find so many qualified, diverse team members within our existing locations.

HW: I think of you first and foremost as a founder, not a ‘female founder.’ That said, as a woman you have a perspective on the ecosystem which I lack. Set aside the people who think our community is already a meritocracy – we’re not going to convince them – but is there something that even well-intentioned VCs do with female founded/led startups that is harmful or could be improved?

KS: The reason I have always liked using quantitative data to talk about industry bias is that data is a language that technologists already speak. 

Right or wrong, it’s easy for people to discount individual stories about bias in the industry. But if you can look at hundreds of thousands of written performance reviews across the Fortune 500 and see that Black people get 25% less written feedback at work than white people, or that women are 20 times more likely to be described as abrasive, or that the word ambitious is used to compliment men but punish women – well, that’s a lot harder to argue with. I love quantitative data based on real workplace documents because you don’t have to stretch to show what’s going on.

The best VCs are outstanding at analyzing data to find patterns and using those patterns to make decisions. But investors rarely use their analytical skills in the context of understanding their own interactions with founders. 

One example: As I’ve ramped up my coaching and advising work with founders, it’s striking how much more often the female CEOs I work with are pushed to sign away major decision rights compared to the male CEOs. Why does this happen? I doubt it’s because investors consciously trust the women less. It’s more likely because the women on average get fewer term sheets overall. When a founder has fewer options, investors typically push them to make more concessions on terms.

This is just one example of many that compounds over time, and could be studied with real data. Huh, maybe I’ll do that for a future nerd processor!

Thanks Kieran! Everyone should subscribe to her free newsletter.

URLs. Fresh hot URLs.

Air Canada Has to Honor a Refund Policy Its Chatbot Made Up [Ashley Belanger/Wired] – Air Canada’s chatbot gave a customer incorrect information about bereavement fare policies, and a court held the airline accountable. Unclear whether the chatbot was truly LLM powered – or by whom – but it’s a fun piece of case law now (at least in Canada) for GPT hallucinations.

The chatbot provided inaccurate information, encouraging Moffatt to book a flight immediately and then request a refund within 90 days. In reality, Air Canada’s policy explicitly stated that the airline will not provide refunds for bereavement travel after the flight is booked. Moffatt dutifully attempted to follow the chatbot’s advice and request a refund but was shocked that the request was rejected.

The best founders don’t care if an investor is “founder friendly” [Harry Glaser/Modelbit] – Repeat founder Harry Glaser wants a few things from his VCs but “founder friendly” ain’t one of them [disclosure: we’re investors in his ML deployment startup ¯\_(ツ)_/¯ ].

Founders want to win so badly they put themselves in an all-in position. The traits an investor can demonstrate that make them irresistible to founders are: (1) A track record of winning, (2) a deeply-held, thoughtfully-justified belief that this business will be a winner, and (3) an unbroken history of high-integrity behavior when the chips are down.

An old-school Las Vegas bookie takes on a new era of sports betting [Danny Funt/WaPo] – Always love a ‘these guys are still doing it the way it used to be’ piece. South Point, one of the last family run hotels in Las Vegas, does 3/4 of its sports betting business IRL and still takes more action than any other institution in the state.

Asked about some national sportsbooks profiting as much as $30 for every $100 wagered while the South Point holds about $5 for every $100, Gaughan told a story from early in his career. Many riverboat passengers, he noticed, spent much of the trip sitting around glumly after going bust on slots. Gaughan lowered slot hold percentages so customers only ran out of money as their boat was docking.

The death (again) of the internet as we know it [Noah Smith/Noahpinion] – Feeds, enshittification, AI slop and disinformation is making the consumer web a mess. But maybe the modern internet needs to die? Or at least transform itself.

So the internet as we know it — social media sites and the Web — is becoming a generally worse place to hang out. Wading through oceans of advertisements, algorithmic randomness antisemitic Russian bots, Tiktok-poisoned shouters, AI slop, and deepfakes is just not a fun way to spend anyone’s precious limited lifetime.

Better, perhaps, to simply withdraw from the public internet — to spend one’s time chatting directly with friends and in small groups, having fun offline, and maybe watching TV or reading a book or a Substack. That sort of human interaction worked fine before the internet, and it will probably work just fine today. Which is why despite the fact that all of the trends I mentioned are negative, I’m optimistic about the future of our digital communications and our media diet. Maybe someday historians will look back on the era when we lived our lives on social networking sites as a brief anomaly.

An Investor’s Perspective on the Value of Regular Check-Ins [Charles Hudson/Precursor Ventures] – Going beyond the one:many email update for why quick but ongoing conversations can be mutually helpful.

As an investor, I don’t know, ex-ante, which of the 10-12 meetings per year with a given company will be impactful. My experience is that neither I nor the company can evaluate the value of any of those meetings in real time; the value is often only known after the fact with the benefit of hindsight. Usually, 2 or 3 meetings each year will address decisions material to the company’s outcome. The other meetings allow us to get to know each other and (hopefully) build mutual trust and a shared understanding of the problem the founders are tackling.

The good stuff on the Internet really does occur in the comments!

1. I shared a post by Ben Werdmuller where he details that a company he used to work for had a recent repricing/restructuring of their stock to better reflect the current state of the company and provide incentives for the team going forward.
2. Ben’s general POV is that stock options are like lottery tickets and that in most cases it’s a losing proposition for the average employee to exercise them – in fact, he’s never bought his options at any company. He recommends just saving your salary instead.
3. As you can guess I disagreed (stock options are an important part of startup incentives and over the course of a career, can be quite valuable for employees), but with three reminders/caveats
   • Don’t behave as if they’re worth anything until they actually are
   • Don’t over-extend yourself to exercise them in scenarios which put your financial well-being at risk
   • Remember founders and investors will often find ways to protect themselves that employees/common stock cannot

I thought all that would happen is a few ‘likes’ and a comment or two. But then a CEO jumped in to point something important out…. while it technically might have been the recent recap which officially lowered the value of previous stock, it really was only an effect, not a cause. [You can read between the lines in this whole discussion to connect the dots on what company they’re both talking about – out of respect for neither of them specifically naming it, I’ll abstain as well].

Tony’s (the CEO) response was an unlock for me in how I explain the ups and downs of private stock to the average person.

For much of a startup’s life new FUNDING VALUATIONS are LEADING indications of POTENTIAL. They are what someone is willing to pay for shares today based on what they believe the company CAN DO in the FUTURE.

DOWN ROUNDS and RECAPS are LAGGING indications of PERFORMANCE. They are what someone is willing to pay for shares today based upon what the company HAS DONE in the PAST.

Obviously there’s a little nuance here because in the former, each successive round builds on what’s been accomplished already, and in the latter there’s usually still some premium to current enterprise value based on what a stock holder imagines they’d be able to get down the road if the company improves performance.

There are a lot of down rounds and recaps in the market right now. And a bunch of private companies still holding on to valuations that they have not yet grown into. When those expectations are preventing a company from moving forward productively they should be cleaned up. I’m usually on the side of even doing it proactively (versus only as a last resort) because it adds clarity in my mind. But it’s an emotional issue for many who hold on to the legacy mark and see any decreases as money lost.

Updated: Ben responded to my response 🙂

More stuff for you to enjoy….

What I Learned Selling My Company [Harry Glaser/then: Periscope Data, now: Modelbit] – Harry sold Periscope Data for $130m and is back building again with Modelbit, an ML engineering platform. Here he provides actionable advice for founders who are building long-lasting companies but know M&A might be the eventual, and successful, outcome.

This is about the bread-and-butter, $50M-$500M acquisitions of mid/late-stage startups who probably took the offers because they had serious doubts about whether they could go the distance. 

The enterprise version of the great re-bundling & vendor consolidation trends [Pat Kinsel/Proof] – Entrepreneur Pat Kinsel noting that many B2B software/tools are at risk of being displaced in their customer relationships by ‘good enough’ versions provided by another existing vendor. Why? Not just the cost savings potential of bundling but also…

risk and InfoSec are still HUGE drivers for vendor consolidation. We are seeing this everywhere. Sophisticated customers want less risk and using fewer vendors is a way to achieve this. Not only for its direct sake, but because each vendor relies on their own set of vendors so it is a geometric problem for the enterprise buyer. And let’s face it, startups don’t always pick the best vendors. We’re seeing this articulated by customers as “4th party risk,” which they very much want to control.

Fit [Molly Graham/Google, Facebook, Quip] – Molly passes along advice for assessing when friction (or even failure) in a job role isn’t about your skill and functional expertise but about more nebulous ‘fit.’ Some of her framework is about how to understand in advance of taking a position, whether the ‘fit’ circumstances are right

Often when we’re searching for a job or being recruited for a job, we’re too focused on the company side — Do they want me? Can I pretzel myself into the space they’ve created for this role? 

But the most important questions are where the two sides come together, particularly as you get more senior. Here are three questions I often ask when I or one of my friends are considering a role: 

1) Are they hiring YOU? 

2) Is what they want what you are great at?

3) Does what they need match up with what you want to do?

When it might actually be a good idea to ask to be ‘layered’ under a new manager [Camilo Fonseca/Business Insider] – From an interview that Lenny Rachitsky did with Noam Lovinsky [YouTube, Thumbtack, Grammarly], Insider excerpted a piece of advice that goes against a conventional wisdom of career advancement.

How to Build a Better Motivational Speaker [Tad Friend/New Yorker] – Where do we start? Read this if you’re curious about the economics of the high end self-improvement industry. Read this if you’re trying to understand what the customers of these personalities are hoping to achieve. Read this if you want to know how Jesse Itzler went from white college rapper to Marquis Jet cofounder to this guy:

His home is an incubator for optimization. Itzler recently told an audience, “I said to my brother about my son, ‘He’s a good swimmer, but he doesn’t really have that eye of the tiger,’ and my brother said, ‘That’s O.K., as long as he’s happy.’ ” There were murmurs of approval. “And I’m, like, ‘No! He’d be happy playing Fortnite and eating Häagen-Dazs every night. We want him to live up to his potential.’ ”

Oh and by the way, he’s married to Sarah Blakely of Spanx fame.

My favorite links goes multimedia this time with two podcasts, among the other articles.

Jelly Roll: The Popcast (Deluxe) Interview [Jon Caramanica and Joe Coscarelli/New York Times] – The guy with the face tattoos from the Super Bowl Uber Eats commercial. I’d known he was also a rising music star but not his backstory. In this podcast he’s confident, humble, thankful, curious, funny, competitive – just basically a great chat between folks who care about the music. Must listen for founders IMO.

Has Gratuity Culture Reached a Tipping Point? [Zach Helfand/New Yorker] – The PoS spins around and you see a 25% minimum suggested tip box, for something that just a few years ago was a generally accepted ‘non gratuity’ transaction. Here we learn the history behind tipping, the psychologies at play, and where the breaking point might be. Even if you don’t click through, here’s the fact you should know, its potential origin:

By the seventeenth century, visitors to aristocratic estates were expected to pay “vails” to the staff. This might have lowered payroll for the estate itself. At least one aristocrat helped himself to some of this new income stream; he threw frequent parties to increase revenues. The system spread. English coffeehouses were said to set out urns inscribed with “To Insure Promptitude.” Customers tossed in coins. Eventually, the inscription was shortened to “tip.” 

Stop Trying to Replicate Silicon Valley [Chris Neumann/Panache Ventures] – While the title might sounds like a Bay Area VC telling all other geos they just can’t compete with the OGs, it’s actually a Canadian investor trying to direct local energy into more productive strategies than low-res carbon copies. Chris cites more innovative strategies such as governments helping their native startups sojourn to the Valley for stints, bringing back relationships and learnings.

Here’s the thing: if governments really want to accelerate their tech ecosystems, they should be encouraging their founders to travel to Silicon Valley in order to learn from and work with the best. Sure, a few might stay. But the vast majority won’t for a wide variety of reasons. And guess what? Those who do stay will learn a ton while they’re in the U.S. And a good number of them will one day repatriate home and bring back with them the knowledge and experience they gained. And for those who choose not to return, where do you think they’re going to open their first remote office…?

The reality is most US cities shouldn’t waste money on cut rate startup incubators or similar, but work their asses off to get large tech companies to locate offices locally, even if it’s just starting with QA and other entry level roles.

The Huberman-ization of America [Rex Woodbury/Daybreak VC] – Rex analyzes the popularity of neuroscientist Andrew Huberman and builds a startup investment framework based on society’s growing interested in wellness. He breaks it into three categories (Performance; Aesthetic; Health) and gives examples of companies selling into these trends. As well as areas that are less covered right now.

Side note, I didn’t realize how popular Huberman’s podcast is!

Guest host Hank Green makes Nilay Patel explain why websites have a future [Nilay Patel/Decoder] – I’ve know Hank Green for a while now due to our YouTube connections. He’s a sharp guy who, along with his brother, are some of my guideposts for what makes a healthy internet. In this interview he switches rather effortlessly from guest to host, interviewing The Verge’s Nilay Patel on Nilay’s own pod [more podcaster should allow this reversal from time to time]. I loved this section in particular:

One of the wildest moments of this conversation for me was when I made a comment that I thought was just a universally believed truth about the post-platform internet: that people these days prefer individuals to brands. And then Nilay told me, “No, that’s wrong. It’s not people who are doing that; it’s the systems that deliver content to people” — a distinction that I’m going to be thinking about for a long, long time. 

All The Carcinogens We Cannot See [Siddhartha Mukherjee/NewYorker] – Read it for the science and/or the symbolism. Article covers the role of agents which aren’t considered carcinogenic but which end up promoting cancer based on the inflammatory response of our immune system (such as air pollution as a precursor to lung cancer). For example:

In the experiment, two researchers working at Oxford, Isaac Berenblum and Philippe Shubik, assembled a group of mice, clipped a patch of hair on each rodent’s back, and painted the patches with DMBA, a cancer-linked chemical that was found in coal tar. Yet only one animal in thirty-eight developed a malignant lesion. When the researchers added some slicks of croton oil to the same area, the results were startlingly different. (Croton oil, a blistering, inflammatory liquid extracted from the seeds of an Asian tree, was used as an emetic and as a skin-sloughing exfoliant.) Now malignant tumors bloomed, appearing on more than half the mice. The sequence mattered. Reverse the schedule of application—croton oil first, tar after—and there were no tumors.

Enjoy!