Ramen profitable in 6 months

Ramen profitability means earning enough to sustain business growth while covering essential expenses. This actually doesn’t sound too exciting, so let me put it in other words: infinite runway. If I decide not to upgrade my lifestyle, in theory, I can live like that indefinitely. That sounds better!

Half a year ago, I quit my 6-figure designer job and became a solopreneur. It's fun, but now I make $3k/mo with my apps, which is on one hand buying me freedom, on the other hand it’s significantly less than I earned in my day job.

Breakdown from last month:

• ✨ Kerlig: $1.7k/mo

• 🌐 LocalCan: $1.3k/mo

While this is a huge milestone, it also means there is still a need for a lot of work to match the level of my past salary. And most likely no fancy vacation this year.

Here is a short Q&A from X / Twitter and LinkedIn threads after I posted it.

Q: What has been the hardest thing during these 6 months?

I think the hardest thing was finances, the mental shift to no monthly paycheck, and relying on savings. An important learning, I'd say, is to lower your ambitions regarding the app or product you plan to build, especially on your first product. Don’t get me wrong, quality is crucial, but perfectionism can hit you hard, and I’m sure your ambitions are too high already. You risk wasting time on a product nobody wants, or worse, giving up and never launching it. Instead, build a simple product that does one thing well.

Q: Any tips for someone starting out?

If it was easy everyone would have $10k/mo SaaS and lay on the beach in Bali. The truth is, it’s tricky and a lot of things can go wrong. But if you have the itch to do something on your own, then go and do it! It’s scientifically proven that the pain of regret is greater than the pain of failure. Go to X/Twitter and learn as much as you can from people few steps ahead of you in the journey. Build a product in which you believe, for a group of people you care about, and that you will be a hardcore user yourself.

Q: How about satisfaction and motivation to work?

It skyrocketed, and I love what I do. Sundays are now different. Instead of stressing out about the upcoming week, I can't wait for Monday!

Other updates

I’ve launched LocalCan blog

I defended a DDoS attack on Webhook.cool

That’s all!

Stay tuned and I’ll see you next time! 👋

- Jarek

When you're building a new app that enables people to do things that were previously impossible, your main focus should be on delivering that unique value. In this scenario, prioritizing design can actually slow down development and in turn even prevent you from being the first on the market, and losing all the first-mover advantages.

Let's consider this scenario:

1. Product A hits the market early with a poor but usable design. People start using it because it offers a new, valuable feature. It gains market share quickly and sets standards for its category.

2. Product B enters the market a little later with a slightly better design. Some users switch to Product B, but many have already invested for Product A, making the switch less appealing.

3. Product C finally arrives with a very polished design. However, it now faces the challenge of convincing users who have already invested in Products A, B or both (!), to make another switch.

While I believe that Product C can ultimately win as it provides the best experience, it will likely take a significant amount of time. Time is a precious resource, and taking too long can put you at risk of simply not surviving or giving up.

Examples of product A:

• AWS, the biggest player in the cloud provider space. I hear developers daily saying it’s confusing to use. But switching to a competitor comes at a high cost.

• Stripe set a high bar for design from the start, but without significant evolution over the years.

• Jira, Notion, ChatGPT, Zapier, Mailchimp…

Examples of product C:

• Linear, Pitch, Monday, Airtable, Cron (acq. by Notion), Lemon Squeezy

The key takeaway here is that when you're creating something entirely new, your primary focus should be on delivering that unique value to your users as quickly as possible, then iterate on improving design and making it more delightful. Design is important, but it probably shouldn't come at the expense of being first to market.

Of course, this doesn't mean you should neglect design entirely. Aim for a design that is functional and user-friendly, even if it's not perfect. Iterate and improve your design over time as you gather user feedback and solidify your position in the market.

Build in public updates

In March I’ve been focusing on Kerlig, as I aim to release it in April! I’ve added support for more AI models. The full list now includes: OpenAI (GPT-4, GPT-4 Turbo, GPT-3.5 Turbo), Claude (Opus, Sonnet, Haiku) and Gemini (Gemini Pro). I’ve also added support for formatted text, including code highlighting.

I had an idea for the logo update and gave it a shot! How do you like it?

Pre-order Kerlig now to pay the reduced price of only $19 $39 🤑
With every pre-order comes a gift! Abstract Wallpapers set in 6K. 🎁

At the same time, I'm a bit stressed out about spending too little time on LocalCan, but it's in good shape to sustain the busy period of developing Kerlig 🤞.

That’s all!

Stay tuned and I’ll see you next time! 👋

- Jarek

Product-Founder-Market Fit test

Here are the questions with my example answers.

1. Does this product solve a problem I have?
   Yes, I encounter this problem a couple of times every day.

2. Am I passionate about this market/industry/tech?
   Absolutely! It has been a long time since I was this excited about something, like now with AI.

3. Do I have the right skills to build it?
   Yes, I was a Product Designer for almost a decade, and I've already built a few apps myself.

4. Do I have any unique insights or ideas for solving this problem?
   Yes, so far no other app I've seen can solve this problem quite as well. Plus, I will ensure it has the best UX possible. AI apps are still in the early days, and the UX has not matured.

Only if your answer is "yes" to all of them, continue to the next step: Validation.

Build in public updates:

1. Kerlig™ my new product

Kerlig.com is an in-context AI writing assistant for macOS. It is now available for pre-order (100+ already) with an expected release in April. No subscriptions. Use your own OpenAI API key and pay for what you use.

See launch post on X / Twitter

2. Webhook.cool hits 1M webhooks 🥳

And a few other wins in February:

• Ranks at #4 on Google for “webhook tester”

• Rocketseat made a video about it

3. I’ve reached 4k followers on X / Twitter 🥳

Last 3 months:

That’s all!

Stay tuned and I’ll see you next time! 👋

- Jarek

A few days ago, I got a leg injury that led to having stitches ...without anesthesia, ouch! That moment made me realize I can endure more pain and tough moments than I thought. I'm a designer who can code, and I'm f***ing unstoppable! There are other options too, like building a new product or freelancing part-time, while growing LocalCan.

So, my advice: believe in yourself, you can figure it out!

I’ve launched Webhook.cool 😎

Exactly one month ago, I’ve launched a free online webhook tester webhook.cool. The main idea was to give back to the community and promote LocalCan. The numbers are quite impressive, 45k webhooks sent, almost 7k visitors, and my favs:

• 16% of LocalCan traffic is coming from Webhook.cool

• 9th place on Google under “webhook tester” keyword 🤌

I’ve launched it on Hacker News and Product Hunt, with the latter being rather disappointing. The launch on DevHunt is scheduled for March.

Custom domains

LocalCan will soon allow Public URLs to be created on a domain you own. It improves branding and creates a more professional look. Unlike Public URLs on LocalCan.dev domain, these never expire. This feature is now in the final testing phase and will roll out in the next few days as a free update for existing customers. And with a likely price increase for new purchases …just letting you know ;)

Also…

• Collaborated with autonomous.ai and recorded Ship more! video and took some nice photos

• Launched LocalCan on BetaList

• Extended Public URL persistence from 1 day to 7 days

• Gave a controversial UX tip that gained 220k views

That’s all!

Stay tuned and I’ll see you next time! 👋

- Jarek

November revenue

While I don't want to brag too much, I'm particularly proud of reaching this milestone, especially considering I was working a 9-5 job just a month ago! 🤯 Now, with so much time unlocked to do what I want, productivity has skyrocketed!

Successful Black Friday campaign

I was initially skeptical about offering any kind of discounts, but in the end, I decided to give it a try. The benefits were obvious: reaching tons of new potential users, generating free backlinks by featuring LocalCan on promotional websites, and, of course, boosting sales!

What is also important for a successful campaign is increasing the sense of urgency even further by limiting number of available coupons:

Thanks Daniel Nguyen for compiling a list of all Black Friday sites and Tony Dinh for his special list on GitHub!

Positioning as Ngrok alternative

Positioning LocalCan as a Ngrok alternative has instantly made the value proposition crystal clear to many people - without changing the product at all. I'm already ranking quite high for the 'ngrok alternative' keyword (see tweet) for which I've created a dedicated landing page.

Pricing experiments

I've gradually raised prices over last 6 months, effectively increasing it from $19 → $24 → $29 → $39 for one license, as I was adding new features. Guess what? No complaints, and sales went up!

LocalCan 2.1 release

In version 2.1, I've added a tool that allows inspecting web traffic from all your domains. It will make testing webhooks a delight!

See the demo:

What’s next: webhook.cool 😎

Building free tools is a powerful marketing strategy that I'm going to push forward. The idea is simple — make webhook testing frictionless with a free online tool. Then, let people know they can do even more with LocalCan as they use the free tool.

I have 2 domains for this purpose, not sure which one to use: webhook.cool or iWebhook.com – which one do you like more?

Gem of the month 💎

Earl Nightingale, a self-help pioneer from the 60s. Watch: Why do you go to work?

That’s all!

Stay tuned and I’ll see you next time! 👋

- Jarek

Magic skill set

A solo founder bootstrapping a business should be the default mode. But the reality is, most people lack the skills, expertise, confidence, or a mix of these needed to do it, which is totally understandable – it’s a lot, and it’s tough. But then, starting hiring and raising money makes everything 100x more complex, dilutes focus …and ownership.

This is the minimum skill set to build a software products. It’s easier than ever to pick up a complementary skill and become unstoppable! What’s stopping you - which skills you’d like to learn? Leave a comment!

Thoughts on roles

Companies created artificial buckets of specializations, e.g., frontend developer, marketing designer, customer support, and so on. This is convenient because it’s easy to replace a person of a given role with someone else, which means companies consist of modular elements – roles and teams. Modularity is good for systems, but is it good for individuals? Not always. Each person specialized in a role can’t do anything substantial alone! You’re dependent on the system - the company. Designers don’t know how to code, backend engineers don’t know how to design, product managers can’t do either coding nor design. These are strong reasons why they won’t start their own companies or build their own products, but will rather stick to their jobs.

LocalCan 2.0

I've released LocalCan version 2.0, which introduces persistent Public URLs (subdomains of localcan.dev), a refreshed UI, and an enhanced tray menu. Existing users can enjoy this update for free!

To make persistent Public URLs happen, I’ve built a custom backend service which was quite challenging. It uses a neat technology called SSH tunneling, something which big companies like Cloudflare or ngrok use as well. I can now control various aspect of the URL e.g. expiry date, subdomain name, transfer limits, bandwidth.

Public URLs do create a monthly cost on my end, while LocalCan's license remains a one-time payment. However, I see this as an experiment. I'm closely monitoring how actively this feature is being used, how many users are utilizing it, and how much they appreciate it. After a few weeks of learning, I'll be able to decide on the next steps. This could involve offering it as a paid add-on, limiting data transfer, or making no changes at all. We'll see, and I'm feeling optimistic!

What’s coming up

There are few interesting features coming out like incoming traffic inspection to help testing webhooks. But November will be mostly about marketing. I’ve noticed a decrease in purchases after the all the buzz from multiple launches in the summer is getting more quiet. So, I need to invest in longer term marketing strategy: SEO. First I’ll release LocalCan Blog with a few posts as a content marketing effort. If you have any wish for pretty much any topic related to local development, please leave a comment!

That’s all!

Stay tuned and I’ll see you next time! 👋

- Jarek

After many launches in the past few months that generated a few peaks in orders, it looks like revenue has stabilized at $1K/mo. My goal is to reach $2K/mo early next year! How do I do that? I plan to make more marketing efforts, including short-form videos and a blog.

Public URLs release

At the beginning of September I’ve released the most requested feature Public URLs! You can share your work with clients or stakeholders, test webhooks, OAuth, and more - all with just one click by creating a Public URL.

I'm already working on improving Public URLs - it will be a custom service that will allow the use of localcan.dev subdomains for public URLs. They will be persistent, short, and based on project names. The release is estimated for around Oct/Nov!

I'm thinking of two variants:

1. Subdomain of localcan.dev based on project name + 4 random numbers. If you stop using it (e.g. your Mac is turned off) it will be reserved for 1 day (or more?)

2. Premium add-on: Any subdomain of your choice on localcan.app. It will be reserved for 1 year and will be a paid feature. I'm still not sure about the exact price, but it will likely fall between $9 and $29 per year. This feature is expected to release a little later, around Dec/Jan.

There are other features like server logs and requests replays that I'm keen on adding next.

Let me know if you have any feedback, questions or ideas 😉

First 100 trials – learnings

It took 1 month to cross 100 trials and the results are quite surprising!

57% of users don't even try the trial version – they purchase LocalCan™ immediately. Moreover, half of those who do try the trial end up buying it before the trial ends. 🤯

My take is that after introducing trials, I've broadened the top of the funnel and reduced refunds to 0 at the same time. Additionally, it confirmed that many people see immediate value in the app!

That’s all!

Stay tuned and make sure not to miss the October update! 🤗

See you next time! 👋

- Jarek

• 100 subscribers
  in this newsletter since I started it 2 weeks ago – thank you! ❤️

• $4.5K revenue
  since I’ve launched LocalCan in May

• 25 trials
  in 5 days after introducing trials in LocalCan

Strong community is the best thing that can happen to your product

This tweet generated a bigger spike in sales than Product Hunt + Hacker News – why? I was not acting as a salesperson; in fact, I don’t think I even played a role here at all. Instead, a couple of people (thank you! ❤️) simply recommended LocalCan as a solution for the problem stated in the tweet.

Trial version - hopes and fears

I've added the ability to download LocalCan and use it for 7 days for free – no credit cards or email address needed. Already, 25 trials have started!

On one hand, I'm confident there's a large group of people, myself included, who prefer trying a product before purchasing. On the other hand, I'm concerned that they might not be convinced enough to buy, potentially leading to a drop in sales 🙀. 🙀

I treat Trial version as an experiment and I’ll report back in September!

I’ve specifically set up an analytics event that will tell me if the app was in Trial or is still in Trial when entering new license key (so just after a purchase)

How do I keep refund rate below 1%

There is 30-day refund policy and no try-before-buy option (well, except Trials introduced 5 days ago)

Here is what I do:

1. 9 out of 10 issues people face are around setting up their apps behind a reverse proxy, not about LocalCan itself. I assist them in solving these issues individually. I actually recreate their stack on my MacBook, whether it's Laravel, Vapor, Remix, Storybook, etc. After solving the problem, I report back with a solution.

2. Then, I also add a troubleshooting guide to the Docs. This way, when a similar issue arises, I can simply send them the link to the Docs.

3. I update the app regularly, as other issues are missing features or bugs

That’s all!

August went supa fast!

See you next time! 👋

- Jarek

What are large language models (LLMs)?

There is a concept of “next word prediction” that you likely use if you have predictive text turned on your iPhone. Large language models (LLMs) are basically highly sophisticated models that leverage this concept and are trained on internet data (think blogs, forums, StackOverflow, and more)

In this issue I’ve focused only on text-to-text models, however there are many others e.g. text-to-image, text-to-video, text-to-speech, speech-to-text to name a few.

There two most most popular LLMs are GPT 3.5/4 by OpenAI and Llama by Meta (Llama is open-source). You can interact with GPT using ChatGPT UI or OpenAI API. For Llama you can use Huggins Face which is like GitHub, but for AI models – see Llama demo or use the model hosted there using code. You can also download the Llama 2 model and run it offline (size vary from 4 – 129 GB) e.g. by using llama.cpp and just 7 lines of code.

Embeddings

Now, the fun part! While LLMs are highly generalized, what if you want to give them specific knowledge e.g. about your organisation? You wont fit the the whole knowledge base of your company into a prompt. And this is exactly what embeddings allow to overcome.

To create embedding means to convert a piece of text into a vector (array of numbers) which will hold the semantic meaning of this text. This allows to do similarity search, recommendation, classification etc.

Example: how chat with PDF works under the hood?

1. Create embeddings from a PDF by first dividing it into chunks. OpenAI’s text-embedding-ada-002 embedding model has max 8191 input tokens (around 3000 pages PDF).
   Check out OpenAI Tokenizer to visualize how word count translates to tokens.

2. (Optional) For very large PDFs, save embeddings in a vector database like Pinecone or Postgres with pgvector extension.

3. When the user asks a question, create embedding from it as well. Now you can easily perform a semantic search to find parts of the PDF that are the most relevant (similar) to what user has asked.

4. Create a prompt to ask LLM (e.g. GPT-3.5) that consists of context - search results from step 3. and the question user has asked.

5. Done!

6. (Optional) Add a memory to the conversation, so that chat remembers past messages. This is simply including “past messages” as a context to each subsequent question asked by the user.

Damon Chen grew his “chat with PDF” app - PDF.ai to $8,333 MRR in 2 months read more…

Chaining

In the "chat with PDF" example, there is a sequence of instructions like preparing the data or interacting with the model. Majority of AI product have it and they are called chains. For this reason LangChain was created, it’s a framework that helps you create and play with chains easily by using components which will keep your app modular. For example your could quickly replace one model or component with the other.

Example LangChain use cases: QA over Documents, Code Understanding, Summarization and many more.

Product ideas

I remember Sam Altman saying in this interview that the biggest business opportunities and value (and yes, money) lie in the “middle layer”, which he describes as base-model e.g. GPT-3.5, tuned to specific industries e.g. medicine. So, companies can create their own, private chats that have specific knowledge built-in, then possibly share it with others and charge for it.

Another point to make is that many AI product don’t have moat. That’s because most of them rely on publicly available models, instead of training their own. In other words it’s easy to copy them. So the competitive advantage often comes down to whoever is the first to build the product, or has better distribution channels (existing customers, followers etc.)

Idea 1: Company knowledge-base chat

In many organisations the knowledge is scattered around, so being able to chat will all sources (think Notion, Confluence, Slack, Google calendar) would be great. For example, a user may ask:

• “What is the process to do X”, and the answer could be:

• “Here is the Notion link explaining the process, but recently Greg and Sam discussed it on #process-x on Slack and they have a meeting about it scheduled for tomorrow, reach out to them for more info”.

Other question: “Summarize what happened in team XYZ during my time off“.
Allow hosting it on-premise, so data never leaves company servers.

Idea 2: Browser extension: Code explainer

Go to GitHub, select a piece of code and let the extension explain it in plain English. Allow to ask additional questions e.g. about specific line of code. Additionally, it could translate it to a different programming language. Great for learning!

Idea 3: Wonderful web browser

A browser that opens any website, especially blogs or news, and shows them in a good looking, reader-like layout, without ads and summarized. 🤤

Idea 4: Monitor your brand

Let’s say you own a well known company or your are a well know person. This app will search for mentions of this brand everyday in Google and various social media platforms. Then will show frequency and sentiment of those mentions.

Code implementation

Python is the go-to language for anything related to AI, machine learning and data science. But I don’t know Python 🙈 So, I’m happy that JavaScript (and TypeScript) is supported by OpenAI API, Hugging Face API and LangChain. This is awesome because it’s now possible to build the whole app: backend, frontend and the website with JS/TS only, yay!

That’s all!

Would you like to read about other AI models e.g. for image generation next time?

Share your thoughts and feedback in the comments.

See you next time! 👋

- Jarek