My VC told me not to be cheap

LinkedIn broke my SaaS

In this episode: * The OG Kit LinkedIn saga: three days of debugging why Open Graph images stopped rendering, and the weirdly simple fix * Why paying for a tool like OG Kit is worth it (someone else deals with this stuff so you don't have to) * Tailwind's revamped partners page and the thinking behind turning it into a filterable directory * Phone.inc's infrastructure decision (build vs. buy), the team settling in, and the first TestFlight invites going out * The case for outbound sales in SaaS, and why Peter is weirdly excited to make cold calls --- Recorded outside in Niklas' garden, with his robot lawnmower making a threatening cameo in the background. Peter kicks off with the OG Kit story he'd been promising. A customer emailed to say they thought OG Kit was banned on LinkedIn: images rendered fine when hosted as static files, but not when generated through OG Kit. LinkedIn is a big chunk of the audience, so this was stressful. Peter walks through three days of debugging, testing different domains and URL structures, talking to Claude, trying a Cloudflare worker proxy, and looking into Cloudflare for SaaS (which would've eaten about 10% of OG Kit's revenue in per-domain fees). The actual fix turned out to be much simpler: LinkedIn's bot was choking on his robots.txt, and explicitly allowing the LinkedIn bot at the top of the file made everything work again, on the OG Kit domain directly. No custom domains needed. Huge relief, especially since the whole point of OG Kit is to stay small and simple. It's also a good argument for why paying for a tool like this is worth it: someone else absorbs the random platform breakage so you don't have to. On the Tailwind side, Peter shipped a revamp of the partners page (formerly the sponsors page). The big idea is to make it work more like a directory, so you can filter the ~70 companies by category. That's better for people browsing, and it also makes the sponsorships more valuable and easier to sell: being the third email company on a filtered page beats being the 71st logo in a giant cloud. The page now includes a testimonial section with companies speaking to specific parts of the program, like Shopify on supporting independent open source, CodeRabbit on tailored partnership opportunities, Resend on brand promotion, and Vercel on direct access to the team. Next up (not shipped yet): giving each company its own page on the high-traffic tailwindcss.com domain. Peter is also candid about churn, since a lot of companies signed up during the January news cycle and some are now rolling off, so the focus has shifted to making the program more valuable for the ones who stick around. Niklas gives updates on [Phone.inc](https://www.phone.inc/). Infrastructure is still the big open question, with deep negotiations on pricing, terms, and build vs. buy. He's managed to cut about 75% off the price from the main supplier they're talking to. The team is settling in after onboarding a new colleague pretty much every week, and the first TestFlight invites for the iPhone app go out Monday, with Android close behind. Katrine, the designer, has fully ditched Figma and is now working directly in Xcode with Claude Code, shipping pull requests, which has been a wild but exciting shift to watch. They also built a first Shopify integration, so when a call comes in you get a voice prompt with context about the caller (like "it's Peter, he placed an order five days ago"), which fits their ICP of small teams where the person answering the phone is also packing orders, not full-time support staff. The episode wraps with a good conversation about outbound sales. Niklas makes the case that hardcore cold calling is unfashionable in SaaS but genuinely works, pointing to companies like FlatPay, Trustpilot, and Firmafon back in the day. Even at Phone.inc's fairly low price points, the unit economics might support a small calling team. --- Still Early is sponsored by Phone.inc [https://www.phone.inc/]: a business phone number, welcome greeting, call routing, and opening hours, all from an app on your personal phone.

My VC told me not to be cheap

In this episode: * Peter's Tailwind ui.sh support inbox after inviting 30k people to early access * The first vlog landed well on click-through but had a watch-time problem (probably because the clip we pushed was also the first half-minute of the video) * OG Kit got hit by Meta's new scraper, which means another entry in Peter's Cloudflare block list (currently around 20 scrapers) * Niklas onboarding three new full-time hires and getting a much higher ambition push from the new London VC than he's used to from Danish ones * Telco infrastructure strategy musings: going one layer deeper in the supply chain makes phone numbers 27x cheaper and changes what commercial offerings are even possible --- Peter's week started a day late after a road trip to Germany with his oldest, which sounds nice but also means feeling behind for the rest of the week. Most of Peter's Tailwind work has been ui.sh support. They invited around 30,000 people to early access, so the inbox just doesn't really stop right now. OG Kit got hit by Meta's new scraper, which is how Peter usually finds out big companies are training models (Apple recently, now Meta). He has a Cloudflare setup a friend helped him build that lets him add new scrapers to a block list, currently around 20 of them. The big customers with hundreds of thousands of pages are actually useful here because they get scraped first, so the system gets hardened before normal users notice anything. Peter and Niklas recorded the second and third vlogs in Copenhagen (you can check those out on YouTube). The first one did well on click-through but watch time was lower than expected. Peter ran the YouTube analytics through Claude and ChatGPT and both pointed at the same thing: the clip they'd pushed hard on social before release was also the opening of the video, so people clicked, saw what they'd already seen, and bounced. Obvious in hindsight. The plan now is to release clips after the video, not before. Niklas's week has been onboarding new full-time hires, signing infrastructure contracts, and a first check-in with the new lead investor in London. Big takeaway: the ambition level from a London VC is genuinely different from what he's used to from Danish VCs. They pushed him to hire someone he wanted but couldn't quite justify yet (the answer being basically "if you'll need them in six months, hiring later will cost you more than the salary"). They also told him to take the most expensive serviced office option as long as it doesn't pull focus. The infrastructure piece is the fun part. Phone.inc currently builds on one provider that's great but premium (think AWS-tier pricing for telco). Going one layer deeper in the supply chain makes phone numbers around 27x cheaper and makes incoming call traffic free, but means building more in-house. That unlocks commercial models that don't otherwise work, like sending every new Danish company a free business number. The build-vs-buy decision is also a go-to-market decision. The episode wraps with Niklas asking Peter to start looking into paid acquisition experiments for Phone.inc. Peter has zero practical experience with paid ads but is interested, and the goal isn't to crack it, just to figure out where to even start. Phone.inc is sort of a prosumer product (selling to the business owner who just started their first company, not a procurement department), so the usual SaaS playbook probably doesn't apply. Instagram, TikTok, founder stories, visually compelling small businesses as mascots. They'll see what sticks. --- Still Early is sponsored by Phone.inc [https://www.phone.inc/]: a business phone number, welcome greeting, call routing, and opening hours, all from an app on your personal phone.

Building a phone company for the second time

In this episode: * What Peter and Niklas are each working on right now (Tailwind, OG Kit, Phone.inc) * Niklas hired his first employee and had to buy a laptop at Denmark's terrible excuse for an Apple Store * Phone.inc's second paying customer and the challenge of marrying old telco infrastructure with a modern sign-up experience * Peter's day in Copenhagen (coworking, Laravel meetup, the Royal Library, vibe coding a revenue dashboard) * The content plan for Still Early going forward (podcast and vlogs) Peter and Niklas are back for episode two, this time recording remotely from their respective towns. They use this one to properly introduce what they're each working on and catch everyone up on the week. They start with a quick intro to what they're each working on. Peter works with Tailwind Labs, where the team is building ui.sh [https://ui.sh/], which is basically trying to figure out how to get AI agents to become good at design. Peter also runs the Tailwind partner program and builds OG Kit on the side, a SaaS tool for dynamic open graph images that he describes as the perfect side project (tiny revenue, but super fun and near to his heart). Niklas gives the backstory on Phone.inc [https://www.phone.inc/]. He co-founded Firmafon back in 2009 (now called Relatel), which became the biggest SMB telco in Denmark. Around 10-15% of all Danish businesses use it today. But three things that didn't exist back then have completely changed: apps with built-in IP telephony, global telco infrastructure you don't have to build yourself, and AI that can actually access the contents of phone conversations. Phone.inc is basically Firmafon rebuilt from scratch as an app. The domain was just available, which is kind of unbelievable, and it was honestly the thing that made Peter get excited about the idea. The big update this week is that Niklas hired his first employee, a designer and product manager named Katrine who he used to manage at Trustpilot. It was his first day having a coworker in over a year, and the onboarding included a pretty rough trip to Denmark's closest thing to an Apple Store where they got the wrong laptop, couldn't get the phone they wanted, and walked out angry. Niklas has a message for Tim Cook. Phone.inc also got its second paying customer, someone from the waitlist who'd been banging down their door. The biggest challenge right now is negotiating with telephony infrastructure providers who want five days of compliance review before a customer can go live, which doesn't work when the goal is to have people up and running within a minute of signing up. Niklas compares the wow moment they're going for to Square's original product video, where suddenly your phone becomes a payment terminal. Peter's week included a trip to Copenhagen for the Laravel meetup (which he originally started back in 2014, now run by someone else). He stopped by Niklas' coworking space, bumped into a potential Phone.inc customer on the street, worked from his favorite spot at the Royal Library, helped send out 500+ UI.sh invites, and vibe coded some MRR tracking into Tailwind's internal dashboard. Got home past one, pretty tired. They also lay out the content plan for Still Early: podcast episodes every other week, vlogs on YouTube in between. They're getting help with editing and trying to keep everything as simple as possible. --- Still Early is sponsored by Phone.inc [https://www.phone.inc/]: a business phone number, welcome greeting, call routing, and opening hours, all from an app on your personal phone.

Why are we starting a podcast?

In this episode: * How Peter and Niklas met and why their previous podcast attempts didn't work out * What Niklas is building with Phone.inc [https://www.phone.inc/] and why distribution is the biggest challenge right now * Founder-led marketing and content strategy in the age of AI * Why vlog-style video is probably the future but still feels super cringe If you prefer video, subscribe to our channel on YouTube: https://www.youtube.com/watch?v=Hz2TxxEkW9k [https://www.youtube.com/watch?v=Hz2TxxEkW9k] Peter and Niklas have been trying to start a podcast for years. They met back in 2019 when Niklas was listening to Peter's podcast Out of Beta and asked if he wanted to give feedback on his startup Legal Monster. Peter came by the office, they hit it off, and Peter literally ended up moving into Niklas' office. They've tried starting a podcast together multiple times since then — including a Danish-language show called "Store Planer" that produced a few genuinely great episodes before the production effort killed it. This time around, Niklas is building Phone.inc [https://www.phone.inc/], a modern business phone system for small businesses and startups. He's got a first paying customer (who basically wouldn't stop banging on the door until he got access) and over 120 companies on the waitlist. But unlike Peter's work with Tailwind Labs, where the team is sitting on massive built-in audiences, Phone.inc is starting from zero on distribution. Peter is helping out as an advisor and small investor, and one of the things they agreed on pretty quickly is that founder-led marketing is the move right now. So part of Peter's role is helping Niklas figure out the content side — editing, promoting, keeping things moving week to week. Every week they just talk about whatever's top of mind — Phone.inc, Tailwind, Peter's stuff with OG Kit, the startup journey in general. No heavy research, no rigid scripts. Just two founders checking in every week and figuring out the format as they go. There's also an interesting tension around content strategy in the AI era. Phone.inc is still in semi-stealth mode — not ready for self-serve, carefully controlling who gets access. But a podcast generates thousands of words per episode, and those transcripts eventually get consumed by language models as they're trained. So it's a way to build awareness with both humans and AI simultaneously, without prematurely blasting out marketing copy on the website. They also talk about vlog-style content and how that's probably what actually works for discovery on YouTube. Niklas had a recent day full of potential content (office tour, lawyer visit, VC skybox event at a football match) and he had his camera in his backpack the entire time. Never pulled it out once. The cringe barrier is real, but they're working on it. Still Early is sponsored by Phone.inc [https://www.phone.inc/]: a business phone number, welcome greeting, call routing, and opening hours — all from an app on your personal phone.

Still Early Teaser

The first episode of Still Early is dropping on April 23nd. Subscribe now to get notified!