Plans for 2024

As I look ahead to 2024, I plan on working on just one pet project. My day job is pretty full on, and I also want to make the most of my son growing up through his infant years. I still find indie hacking a joy, a privilege and a source of infinite lessons to be learnt, so I don't plan on quitting any time soon, however, I've learnt some lessons.

1. Stop assuming you know how to sell something you've never sold or created before. The art & science of selling evolves just the same as creating something to sell. You don't know how to sell it until you've sold it, so everything at this point is your best guess. My wife sells more than me with zero sales training. She knows her audience amazes me every week.
2. Stop building in isolation. Share the idea in a demo video, ask for validation from a wider field and go beyond your comfortable circle of contacts. I try and remember the child I was when knocking on people's door asking if I could mow their lawns. That child rocked and was a master of experimentation to make a quid!
3. Don't be afraid to share and ask questions. If you know your subject matter and have put time in to understand your audience, then you might be your own worst enemy. When you're building something, it's all too scary to ask a question because it can make you look unconfident, unsure, unskilled and uncertain. The good news is, you're selling to humans, and (almost) everyone is like that. Do your homework, try and understand your target audience and when you ask questions, you might be surprised what comes back.
4. But what if someone steals my idea? That my friend is called instant validation. Well done, you might just have a workable idea. Nothing is ever new, and the chances are, someone else is already working on your idea. You have nothing to lose. If someone else launches in competition, then good. An adversary that validated your idea, is a great adversary indeed.

Justifying Expenditure

Indie hacking costs money and despite efforts to keep those costs low, sometimes justifying that cost is hard to swallow. This year I accidentally left three servers on for a month on AWS, which cost me just shy of £800. It doesn't matter how many cloud billing horror stories are read; I made a mistake that ate a significant chunk of my budgeted pocket money.

Indie hacking without a profitable project is just a time and cost sink. This year I've figured out a pattern of essential services that allow me to hack around with my ideas on a monthly fixed cost. Let's start with the Tech Stack, which is still Golang, React & NextJS based.

AWS | Fly.io | Vercel | ngrok | Termly | Mailersend | Stripe | NATS | Clerk.dev | TailwindCSS

AWS provides me a bare minimum set of services, being DynamoDB, Route53 and S3. These provide the NoSQL system of record, easy and simple DNS and simple yet powerful object storage. Overtime, I'll migrate away from S3 to NATS, but for now, my re-usable code snippets are wired into S3, so I'll stay there for a short while. Sometimes I'll use Amplify for React hosting and Lambdas for periodic jobs like mass deletions on a DDB table or emailing out offers.

Fly.io is where I run production Go written API services. It's simple to deploy, and when they're not having issues, it's smooth and simple.

Vercel runs my NextJS front-ends and if I need plain React, then I'll use AWS Amplify.

ngrok provides my tunnelling endpoints for dev services. That means whatever machine I'm developing on with whatever connectivity, I can try my apps on my mobile devices and other household equipment prior to pushing them to prod. I've been a long-term user of ngrok and will no doubt continue for a long time to come yet.

Termly is the service I use for managing my terms of service and privacy policies. I'm not a lawyer, and whilst I might know some of the UK laws, I do not know half as much as I should about global law. Termly takes all that pain and makes it go away. I get documents that I make available to would be customers and hand over a monthly fee to Termly for doing the hard work.

Mailersend provides my transactional email. It's simple and makes sending programmatic email a no brainer. With their selection of SDKs and graphic mail builder, I can do away with my previous workflows and the likes of SMTP services.

Stripe is still my current choice of payment provider, despite trying out Paddle and others. Stripe provides simple recurring services, one-off payments and payment links. It doesn't handle global tax like Paddle does, but I've come a long way understanding its idiosyncrasies and how to do things like account management and deal with limited usage-based consumption of a SaaS service on a fixed plan by introducing top-ups and the like. Stripe's dispute charge is eye watering however at $20!!!

NATS is my secret weapon in all of this. NATS provides asynchronous handling of my east-to-west events, UI updates, email queues, API high availability and more. Without it, my software development processes would be messier, and I couldn't just throw my services where I have the ability to put them, like on Fly.io or my crappy dev server farm, which comprises of some eBay sourced mini Dell desktops and i7 Mac Minis that run Ubuntu!

Clerk.dev provides user authentication and is hard wired into my UIs and APIs so that I don't have to roll my own auth layer. I've found Clerk to be easier to work with than Auth0 and more flexible for the kinds of projects I build. I could move to another platform, but Clerk provides me everything I need to bootstrap an idea and get rolling at no additional cost than the base level. It costs more above 10,000 MAU, but if I get to that point, I have other problems to solve!

TailwindCSS - need I say more? It's simple, my brain likes it and it's easy to get something that looks half decent working. I'll continue with it in 2024.

These services cost me about £120 a month in total and it's enough to run existing projects and experiment with other ideas.

The final thought for the stack is that I'm trying to limit the components I use, so I can move quickly thanks to familiarity and speed and keep abreast of the good stuff that comes out of active communities. It might not be what the cool kids are doing, but it works for me!

Main Project for 2024

ClientRecordCards.com (CRC) is the first project that will go live in the new year and will likely be my only live production project for 2024. It is a simple GDPR compliant CRM system with a responsive, mobile friendly UI for tiny one person businesses specialising in beauty treatments, massage, personal training, plumbing, DIY, gardening and other one-to-one services. It's already about 95% complete and my wife's business uses its predecessor, which I created back in 2019.

So far, my facts are thus:

Fact: Small one person businesses and sole traders manage just fine with a shared calendar and want control of their life.

Result: Managing bookings isn’t something I need to worry about and thus, do not need to build anything.

---

Fact: No one cares too much about email and SMS verifications. My wife has a cancellation policy which states that full repayment is required if less than 24-hour notice is provided. That policy has never been invoked.

Result: I don’t need to send email and SMS verifications for anything nor deal with my customer's non-paying clients.

---

Fact: Target customers have verified they are happy to pay a small amount per month to have the nightmare of managing client records taken away from them. Printing bits of paper off is a liability and information is strewn across emails, SMS messages, Facebook messages and probably on a mirror written in lipstick.

Result: Great! For PLG purposes, I have a free tier for tiny businesses with 20 clients or less. Anything more, they can upgrade and pay a small monthly fee, or the price of a good cup of coffee. 20 clients doesn't seem much, but it's enough for those dealing with friends and family, students and those building up their business. If they invest into CRC early, then the chances are they'll stay with it and surpass the free tier limits.

---

Fact: Target customers seem happy with the ability to invite people to create their own record with an email link that’s valid for three days or so. Creating a customer and health record on the first appointment can be stressful and make appointments overrun. The previous version had a locking mechanism, which made the app session private to that one customer. It’s been confirmed that whilst a nice idea, it was rarely used anyway and the unlock “forgot code” feature was used as much as the lock feature.

Result: I’ll create an invite customer button to my client’s dashboard which will send an email containing a link.

---

Fact: Easy visual identification of records for validity is great for speed. Red blobs around record identifiers show the record is out of date. This is important especially for beauty therapists that deal with the risk of allergens, directly affecting their insurance and the health of the clientele.

Result: Show visual markers around dates (red/amber/green) for six month validity.

---

Fact: Target customers have no idea how many clients they have and knowing the numbers is often a smile maker.

Result:  The dashboard will show the current number of clients and growth month-on-month.

---

This application however has some risks and mainly Stripe centric. With my entry level price of £5 per month and upper price of £9 a month, chargeback pricing is more expensive than my customer monthly value. That means I need at least a 4:1 paying customers to chargeback ratio. Without that, I’m going to exposed to large risk. It's something to keep an eye on. My target is 10:1 and anything lower, will mean a rethink over what I'm providing and why the chargebacks are happening. It's normal to receive things like "I can't even login" style comments and let's face it, the competition will be the bulk of the problem in so much as trying to sink the project. Two mitigations are in place for that. One is to lean into my terms and conditions. There will be a stipulation that if the user has upgraded after signing up to the free tier, cancellation will be possible with repayment to the value of a single month via a refund initiated by ClientRecordCards.com. Any attempt to initiate a chargeback on your credit card will result in a lifetime ban, permanent destruction of data and an entry on the wall of shame. Yes, there will be a wall of shame. The bank will still nail me to the wall, but I can dispute it based on free tier usage and upgrade from free tier and with terms and conditions the user is somewhat forced into accepting etc. The second mitigation will be to prevent the upgrade of an account until the maximum number of free records has been created. That way a measure of usage can be observed prior to triggering a chargeback and any sniff of "not being able to login" style nonsense can be prevented.

Summary

As I go into 2024, some of my weekend and evening hours will be spent on this little pet project and I’ll report the progress as it launches and the challenges I’m facing or have dealt with. This project is gloriously simple and extremely targeted at a problem my wife and her business social circle face every day. Something I think will pay for the odd boy’s toy here and there! Also, my wife knows her market well and has offered to share links and advise people based on her experience. As the whole thing was created for her and has received her input all the way through, it should meet the market need nicely.

You’ll notice a clear avoidance of LLMs or AI with this project as well and that’s because I got burnt a little bit in 2023 with a project called ThoughtZen, which was an AI panic attack and anxiety WhatsApp chat bot. Whilst the service worked and had tremendous interest (1m + page visits and lots of inbound emails), I could not overcome the concerns over privacy and shelved the project, with many would be customers holding off due to unknown regulatory issues. Second to that, I had lots of page visitors talking to the customer support chatbot as if it was the actual ThoughtZen chatbot, which was just bizarre. There was even a message on the customer support chatbot that said "This is customer support only". It didn't help.

So that's it from me for 2023 and now you know my plans for 2024.

What have you been working on and do you have any plans for 2024? I’d love to hear from you! Maybe even join me for a new podcast episode of Built.fm in 2024!

Have a great Christmas, holiday period, festive time and a fantastic 2024.

Much love,

Dave @built.fm

• Share
• Tweet