Yana Welinder: From Lawyer to Product Lead, Founder, Investor & Community Builder
A conversation about her startup, Kraftful, product management, LLMs & AI, community building and more. (#58)
Yana Welinder is an experienced product manager, founder, investor and community builder. She puts out a lot of great content. I was very happy when Yana agreed to this interview. 😃 I learned a ton.
Yana has an interesting and diverse background. Currently, Yana is the founder of Kraftful, a co-pilot for product teams. Previously, she was Head of Product at IFFFT and worked at Wikimedia Foundation. Like many product managers she didn’t start in that field.
Most interviews like this would be podcasts. I enjoy written interviews because I find the depth of answers is better; interviewees get to really think about their responses, and interviewers get to do follow-ups with new questions. You may disagree (which is fine!) and perhaps one day I’ll do a Focused Chaos Podcast, but today—we’re going old school…
First a little background…
Yana has been a product lead at various tech companies, everything from PM #2 at a fast growing unicorn to Head of Product responsible for teams serving many millions of users. She’s appeared on Masters of Scale, Product School Pod, Products That Count, One Knight in Product, Product Tank, led Women in Product SF, and written PM blog posts used by thousands of PMs. Suffice it to say, she knows her stuff. And her startup, Kraftful, is a company dedicated to helping product teams (more on that later).
As I mentioned, Yana didn’t start out in the product field or entrepreneurship. She grew up in Sweden after going there as a child refugee.
“As a child, I was most interested in art and design. My mother and grandmother are both architects, so I got early exposure to it. But I ended up studying law, possibly because it seemed like a more practical career being an immigrant. I went to college in London, spent a gap year in Dublin, and moved to the US for law school. After dabbling in various legal fields – tech law, constitutional law, and even environmental law – I discovered my true calling in product management.”
Now, onto the interview…
You founded Kraftful in 2019 before the AI craze hit with the launch of ChatGPT. What did Kraftful look like then compared to today?
We started from a very different product and pivoted a few times along the way. The only constant is that our mission was always to help product managers build better products.
Kraftful’s first product was a standardized user experience for IoT products that improved based on analytics. You can think of it as Shopify but for IoT. For example, our UX framework powered the smart thermostat apps for some of the largest HVAC brands in North America, winning industry awards alongside Google Nest. We pivoted to a standalone analytics solution with GPT powered feedback analysis in 2021, but it was much more rudimentary because LLMs were not great at summarization until later. Part of our reason for pivoting was that the IoT industry was going through issues that limited the speed of our growth.
Witnessing the LLM evolution and getting more demand for our GPT powered feature, we pivoted again shortly before ChatGPT, which ended up being an incredible wave to ride.
For more on pivoting, check out How to Pivot the Right Way.
Where did the initial inspiration come from for Kraftful? What sort of validation did you do before building?
I was fortunate to be one of the first alpha users of GPT-3, and summarization was the first use case I experimented with. It wasn’t great back then, but there was a spark of potential. As LLMs got better at summarization, our team launched the first GPT-powered feature in 2021.
The following year, the models got even better at summarization, prompting me to spend more time validating whether we should double down on this use case. Around that time, I attended Masters of Scale Summit – an invite only conference where I met lots of founders and product execs. I’d casually bring up the idea without specifically asking for feedback and observed the reactions. My MVPitch was roughly: “I’m Yana, founder of Kraftful, we summarize user feedback with AI to get lists of feature requests and complaints.” And, wow, did it resonate. Every convo turned into a mini therapy session about user feedback nightmares. "Our CS reports are biased" or "We've got hours of user interviews, but who's got time to watch?" And all these stories reminded me of my own pain points as a PM. It was eye-opening how this solution could really help product teams build better products.
Learn how to identify and validate meaningful problems to solve.
Do you think LLMs get so good and easy to use that companies build summarization capabilities on their own? And how can PMs be sure that the problems that need solving are really shining through? We often hear, “people don’t know what they want” so prioritizing feature requests from feedback is…risky? A mistake? Or perfectly fine as an approach?
Doing summarization properly requires a complicated architecture to collect data from all your sources daily and analyze it in different ways that make sense for each type of feedback. Teams can build it internally, but won't, just like they don't build their own Amplitude replacement.
We make it easy to get a good overview of what users request and dig deeper to understand requests in context. This frees time for PMs to figure out how to actually solve the underlying problem (or if there's something that is even more important from a product strategy perspective). I tweeted about this recently because it came up in a demo of a new feature.
Would you say your core value proposition is saving time?
I'd say the core value proposition is to get a better overview of what users really need.
Most PMs spend very little time listening to users because they have lots of competing priorities and it's less obvious day-to-day if they spend less time gathering feedback (compared to say backlog grooming).
And depending on your product, it may actually be impossible. For example, there were not enough hours in the day for my product team at IFTTT to read every app store review, every subreddit post, every support ticket, every user interview transcript, every sales call report, etc.
Inevitably, PMs miss some context. We help them get a full picture - in minutes.
How do you approach go-to-market / customer acquisition? On your website you highlight a number of big companies (Google, Dropbox, Netflix, etc.)—are they the ICP?
Our ideal customers are product teams at mid to large scale companies – basically teams with enough user feedback that they can’t review it manually. We have a PLG model, where individual PMs can sign up for free and upgrade to a paid plan to analyze more data. They tend to invite more team members before speaking with our sales team to get an enterprise plan.
What was the fundraising experience like, and when did you go through it?
I’ve done three fundraises: 2019/20, 2021, and 2023. They’ve all been challenging in their own ways.
The last round was opportunistic when a few of our users (CPOs and Heads of Product) reached out to angel invest after our MVP launch. I first considered using that opportunity to raise a larger round, but quickly realized that as a solo founder I couldn’t step away from product building at such a critical time. So I wrapped up the round with Kraftful users and two of our existing VCs in 2 weeks and went back to building.
The first round was before the very first pivot. But I had many investors reinvest in subsequent rounds across the pivots. Some of my investors expressly focus on founders rather than ideas (e.g. Precursor Ventures).
As a testament to our team’s progress, our 2023 round was at a significantly higher valuation than our 2021 round. 💪
On your website you list competitors: “compare with Productboard, Aha, etc.” Why? Is there an SEO play there?
The search term “Kraftful” has some of the highest SEO in the user research space due to organic traffic, mostly from X. The competitor lists on our site are some recent attempts to improve SEO for other search terms. So far, the traffic we get from them is negligible compared to our direct traffic and people searching for “Kraftful.”
The ChatGPT Store just launched. And Kraftful launched Kraftful GPT. I’m curious to get your thoughts on the store and its future potential.
I believe there’s a significant chance that web apps will be replaced by apps within AI experiences, possibly ChatGPT. In that scenario, Kraftful might become accessible through our GPT.
Kraftful GPT in its current form has much more limited functionality because it’s designed with ChatGPT’s constraints in mind. One insight I had from working at IFTTT is the challenge of getting users to authenticate within another product. So for the full Kraftful experience to work seamlessly within ChatGPT, both platforms would need to evolve significantly.
For now, we’re strategically positioning ourselves to be ready if the platform proves to be a suitable environment for our experience. The fact that this drives traffic to kraftful.com is a nice bonus.
You’re also an investor. What do you look for? And please share more about Pioneer Fund.
As an angel investor, I've focused on YC companies (I previously went through YC) with exceptional founders (e.g. Flutterflow, Elpha). As a founder, I have less time to evaluate startups, so I joined Pioneer Fund as a Venture Partner to invest alongside other YC alumni that I trust. Within the Pioneer Fund, I now mostly invest as part of a designated AI fund. It’s a unique model where a group of YC alumni evaluate new YC companies and then pool funds and other resources to help those companies be successful.
Tell me about VC Backed Moms and GenAI Founders.
VC Backed Moms and GenAI Founders are two founder communities that I started with a few friends and have now grown to two active Slack channels with a few hundred founders each.
We started VCBMs during the first part of COVID lockdown, when it seemed crazy to try building a startup with a small child at home without any child care. It was completely uncharted territory, but I also knew that there were lots of founders out there figuring out the same problem. VCBMs has been incredibly helpful for my founder journey and I ended up hiring ex-founders from the community as our VP of Engineering and Ops lead.
When I started building with LLMs, I found that I similarly wanted a community of founders in that area. I tweeted a link to a Google form and a few hundred founders signed up within 24 hours for GenAI Founders. The group has lots of helpful discussion about evaluating different models for scale, how to talk about your moat when building on foundational models, etc.
How do you feel women are treated as founders today from colleagues, investors, boards, talent (that they’d recruit), etc.?
As a female founder you definitely have to prove yourself more, particularly with investors. This is less of a problem for hiring or interacting with other founders – fellow builders usually see vision and get behind it quickly.
There's also a silver lining – underdog status can be a powerful motivator. 😉
How did you get into product management initially? What’s your advice for people trying to get into product management?
I started my career as a lawyer, but ended up doing law as if I were a PM – focusing on crafting the best solutions to specific problems rather than just applying legal frameworks – and ultimately realized my true calling.
I did a short stint as a law professor, focusing on AI policy and published one of the first pieces on computer vision in Harvard Journal of Law & Tech and a chapter in the Cambridge Handbook on Consumer Privacy. Then I started doing research on how UX can replace legal policies, quit my 2-year tenure early, and went to apply that research to Wikipedia. At the Wikimedia Foundation, I led legal and policy teams. I also ended up leading a few product initiatives, including product strategy for reader experience and the rollout of HTTPS for all of Wikipedia. Encrypted Wikipedia access ended up being a pretty big deal considering government attempts to censor free knowledge. That’s when I realized that I should really work in product and I became a PM.
I found that it was hard to look for my passion when I was pretty successful in a career that I wasn’t actually excited about. The further I got, the harder it became to make a hard career pivot. If I were to do it again, I’d try to figure out how to pivot earlier. I was so much happier in product management.
For folks who already know that they want to get into product management, my advice would be to try to get PM experience before applying for your first product role – either in a different role or by starting your own project. Product management is all about practical experience, so it’s hard to get in through some sort of theoretical learning.
What are your thoughts on the hubbub surrounding the future of product management; specifically related to Brian Chesky’s comments, Snap letting PMs go, etc.? Is this the end of product managers?
The PM role apocalypse? I think it’s a red herring. I bet Snap took advantage of these conversations to explain a regular layoff, i.e. an opportunity to eliminate poor performers.
Brian Chesky’s comments are something that I’ve frequently heard from design leaders. Brian is a designer, who happens to be the founder CEO, so he has the opportunity to shape his company based on the role he wants for designers.
Product management will go through a big change, but it’ll have nothing to do with the current debate. With AI handling a lot of the process work that currently falls on PMs, each company will have fewer PMs and they will need to be better at product strategy, design, and positioning.
That doesn’t necessarily mean fewer PM roles because AI will also enable many more companies to solve problems that aren’t being addressed today because companies will need less funding to get started. Each of those companies will need PMs. At least until we have super intelligence…
You’re very active on X. How do you feel that your tweet, “Why did the product manager cross the road?” was one of your most popular. 🙂 And did you come up with that or ask AI to?
X is one of our biggest user acquisition channels, so I try to post regularly. Many of my posts are written with ChatGPT. I usually come up with an idea for a post and have ChatGPT elaborate on it or revise my posts. Unfortunately, LLMs are not great at humor yet. So that particular tweet about PMs crossing the road is my own dad joke or rather mom joke.
Biggest misconception about product management?
PMs aren't just feature jugglers. All the focus on prioritization makes it sound like product management is all about taking features and bugs from a list and figuring out how to prioritize them using RICE, MoSCoW, or any other prioritization framework. The real job is trying to figure out user needs and how to best solve those needs. It's about insight and impact.
Common mistake(s) first time product managers make?
Newbie PMs often get caught up in visibility tasks instead of diving deep into problem-solving, even if it's under the radar.
Things like stakeholder management and backlog grooming are most visible to colleagues. But unless you’re stepping into a role where someone else has already defined the problem and the product, those tasks won’t make the product successful. You need to allocate a lot of your time to understanding the problem, even if no one at the company can tell how much of that work you’re doing until it hopefully pays off with the users.
#1 skill product managers need?
PMs need to be clear thinkers and communicators. This means being able to take in lots of information, clearly identifying the problem, figuring out the solution (or first steps toward a solution), and concisely communicating it to the team and ultimately users.
I’ve been reflecting on how you need the exact same skill for building with LLMs because the quality of the output greatly depends on the clarity of your prompt.