Introducing HerPhase

This is a personal monthly reflection on building in consumer health, looking back on what I've been working on and forward to what's next. I'm sending this to F&F who support my entrepreneurial journey, to share what I’m up to, create more accountability, and to increase my surface area of luck. More context here.

PS: I encourage you to give me honest feedback and challenge my work, don’t lie to be nice :) Just hit reply to this email or text me on WA.

March's biggest news is that I’ve launched the waitlist for HerPhase and onboarded the first beta users.

What is HerPhase?

HerPhase is an agentic cycle companion that tracks her cycle and keeps her partner in the loop.

HerPhase lives on WhatsApp for two main reasons: to meet users where they already are* and to focus fully on functionality before anything else (ie capabilities instead of user interface, App Store admin, etc.).

HerPhase replaces apps like Flo Health or Clue for cycle tracking while improving the partner sharing feature by offering more personalized recommendations and education, without requiring the partner to download an additional app.

*except the US

Who is HerPhase for?

HerPhase currently targets women with a natural cycle who’d like to keep their partner in the loop. Think: What phase is she in? What does that mean? What are her needs and how can I support her?

That said, if you’re that woman, I’d love for you to try it and share valuable feedback. If you’re that woman without a partner, you can still provide me feedback on the cycle tracking features. And if you’re a man, I’m sure you’ll leave a good impression if you propose using HerPhase with your female partner (or just share HerPhase with your best female friend).

For the time being, the process to be able to start using HerPhase is:

Step 1 is more or less the survey from the last newsletter, but with auto-categorization of the results. Currently, I’m filling the waitlist slowly and just from my network. The reason for this is to stay closer to the user and enable faster and deeper feedback loops, which will hopefully allow me to improve the basic functionalities of HerPhase before adding new features.

To this end, for step 2, I’m jumping on a short video call with every new user to watch them onboard in real time. It may sound silly, given that it’s basically looking at users using WhatsApp, but I was myself surprised to see how telling the popular advice of ‘talk to your users’ can be. One example: a user kept scrolling up after sending her last message, signalling that she was checking whether HerPhase had sent a response already. Nothing she said out loud, nor would I be able to see this on screenshots or time logs.

Building on WhatsApp also comes with a downside/risk: platform dependencies. From the few interactions I’ve had with Meta Business: it’s a surprisingly bad tool. And it comes with real limitations, mainly around proactive outreaches outside of a 24h window since the user’s last message. Nothing that I cannot deal with for now, but something to be aware of. Just like Meta’s crackdown on general-purpose AI agents on WhatsApp (lucky to be in Europe, I guess 😅).

Lastly, I’ve definitely had people (including myself) wonder why I’d be the right person to build a female cycle tracking companion. I was therefore somewhat satisfied to read that Flo Health (the #1 cycle tracking app in the world with 400M+ downloads, great story btw) was built by two men (Clue, the #2 app, was built by a 75% male founding team).

With a working prototype in my pocket and first beta users onboarded, the focus for April is three-layered, with a clear priority around distribution.

When I say distribution, I generally refer to it as everything related to reaching the user by “distributing” the product, which can mean different things depending on the business, audience, stage and geo.

In my case, I'm revisiting a project I flirted with last year, something I referred to as "Lovable for distribution". It's the same idea that led to my freelance gig building AI workflows for marketing teams. This time, I'm not trying to make money off it. Instead, if it works, I'm keeping it to myself as my unfair advantage. In other words, build a growth engine that helps me run Marketing/Growth on mostly autopilot.

This will be a complex undertaking that could easily stand as its own business - see companies like Epiminds from my Stockholm Startup Tour building in that space. Given the scope, I’m prioritizing a staged rollout, focusing first on organic socials to align with my current growth phase. To give you a first idea: it might have to do with farming and fruits, but more on this next month!

The main rationale why I’m starting to expand into organic social is to find more beta users, specifically “strangers” with zero stake in my success, who won’t sugarcoat any feedback, compared to my current and more obvious acquisition channel, people from my network. Besides that, it comes with many other good reasons:

The second layer is to continuously stay as close as possible to all beta users, running tight feedback loops, incorporating feedback, and fixing bugs on the fly.

And the third layer is to keep working on and improving my “system architecture” (basically how I work with Claude Code - my number one interface) to optimize output quality, speed, and token usage across my overall setup.

I think there are many thoughts I could try to squeeze into this section, but I’ll try to spread them over the next edition(s) so you can actually finish reading in a decent amount of time.

On founding

There’s so much content, wisdom, and opinions on building startups out there that it can be both distracting and disillusioning. I do not doubt that starting one’s own startup is very risky and more often than not, it doesn’t work out. But I also believe that it’s more nuanced than just that; besides, there are also plenty of good reasons to give it a real shot.

On this, I found Daniel’s (EWOR founder and fellow SGS and START alumnus) perspective interesting, who argues that entrepreneurship is actually the most rational career choice. He argues that people massively overestimate the risk of entrepreneurship and underestimate how many attempts they get. He explains the math as follows:

When you start out, you have around a 40+ year career. A serious startup attempt takes 2-3 years before knowing if it works or not, which gets you around 20 shots at building a company and you only need one to work.

If every attempt has only a 20% chance of success (and 80% chance of failure), that sounds bad. But probability calculation gives failing all 20 tries a probability of just 1.15% (0.8^20) - or 12% if you’re more pessimistic (0.9^20).

He goes on to compare the hard ceiling a traditional career in employment has (trading time for money linearly with limited outcome) to the asymmetric upside a startup can have.

Lastly, he makes the point that the catch is that the math only works if you keep trying and learn from each failure. Only over a lifetime, the repeated lottery with improving odds and asymmetric upside is a better bet than the safe linear path.

In my opinion, his thesis is very theoretical - it’s from a math perspective after all. Besides the immense willpower required to pull through, it doesn’t account for a scenario where you can actually afford to try a 20th time, let alone fundraise after 19 failures (?). But overall, especially in a tech inflection point like today’s, it’s likely truer than not that ‘the safe path is less safe than we think, and the risky path less risky than we think’ (I believe this is a quote that I picked up somewhere).

On building

Having said that, I’m very aware that I’ve already invested my fair share of time into this journey with yet little (measurable) outcome. I’m far from being able to live off it, but I also wouldn’t put myself through all of the anxiety and stress if I weren’t 100% convinced that it’s both the right path for me personally (fulfillment, growth, etc.) and for what I plan to build up around me long-term (family, stability, etc.).

By far the biggest outcome of this journey was becoming AI-native. I'm no CS grad or PhD - and talking to ETH folks, I still feel behind (which is fine for what I'm building). But talking to peers at events like START Summit last month, I’m confident that I’m in a good position to eventually bank on it by gradually building a compounding, AI-first system around me that helps me build better and faster - especially in my current position as a solo founder.

Using AI day in, day out definitely shows me that pure software moats are thinning incredibly fast (although I still believe that we’re very early in consumer AI - probably more on this next time), which I guess is a good reminder to wrap up here and get back to building moats now.

As always, if you've made it this far, you're genuinely a super-supporter (or my very curious mom), THANK YOU! To have an even bigger impact on my life:

Less ask and more recommendation: Please follow my Daily AI Digest (WA Channel) to stay on top of the most important AI news out there. You won’t hear it first there, but you won’t get it briefer anywhere else!

It might sound too technical/abstract at times, but that will fade as you become more AI-native yourself 🙏