I want to make useful things that make life easier — but I’m just one person.

It’s hard to do it alone. Especially in a field like tech, where specialization is so important. I’m not a product designer. I don’t know how to oversee user testing or make data-visualizations of growth projections!

So I’m entertaining the developer mindset of “I could build something for that” and making tools that fill in the gaps. They don’t have to do the job for me automatically but could they make it easier for me, extend my reach.

Product ideation, validation, budgeting, presentations, marketing, testing — this list is overwhelming. Basically everything about making something for some audience that isn’t development (the part I’m actually good at) is something you can build a whole career on, but for now I just need a passable score in each category.

Over the past few months, I’ve been building a bunch of tools for myself so I can simplify my product design/development process. In this writeup I’d like to look at the core concepts behind each of these product development steps and figure out some way to build some metaphorical mech suit to robo-assist my dumb human brain into building better, more useful things.

Better Ideas

Social Media Scanner

The foundation of a useful thing is that it solves some problem. I want to find someone struggling and make them a solution. So how could we use technology to find people with problems? That’s right, people complain on social media constantly.

Let’s build a social media scanning tool which fetches, sorts, and delivers any consumer complaints, feature requests, or recommendation requests. If someone has a problem, I can find it on the scanner. Pick the source, Twitter, Reddit, Bluesky, etc. and pull in a bunch of posts.

I put together a list of notes and requirements for exactly what counted as an insight, and I followed this list pretty strictly. It’s a lot to include here but things like: “Identifies an actionable idea” “Avoid vague complaints, nostalgic or resistant attitudes” and of course ethically “Does not directly identify a product idea or state intention to create the product” because we respect IP, and there are a million problems to solve so why steal one.

I gathered data for a while whenever I looked at social media, when I saw a relevant post I’d add it to my dataset. I used it to create embeddings for a natural language processing algorithm, with a few tricks to fine-tune results and clear the noise. It fetches posts, sanitizes out filler and stop-words, compares them against good and bad examples and returns a score. Here’s a basic example of what the scanner returns after about 6 weeks of gathering data, ~900 positive examples and ~8000 negative examples:

Not a bad starting point, my head’s already spinning with what we could build to solve this, including some method I’ve never had a reason to learn like “How do I build a VSCode plugin that does this?”

The seed is planted, now we can build it, right? Okay no we can’t but I like the energy.

Semantic Outlining

This part of the process would always be a huge mess for me, I’d just keep notes about the project and eventually I’d put some sort of plan together, often skipping some fundamental steps. The first thing I knew I always skipped was if my idea even sounded good on paper. Like I could imagine it, sure, but when someone asks me why its useful my argument would fall apart.

In the toolkit I built, I added a mad-lib which essentially says “Product Name is a Product Type that helps Target Audience to Product Goal by Key Feature which leads to Product Benefit/Mission.”

Maybe that’s a little information dense, let’s look at an example based on the post we got from the scanner:

It takes 2 minutes max to think through these prompts and give a reasonable response, and by the end of it I have a pretty strong sense if this product has any potential. What is it? Who is it for? What does it do? How does it do it? Why does it need to exist? All of those questions are addressed in a short statement.

With this made I built some other ideation tools and filled out a tab of my dashboard with them, here’s what that looks like at the time of writing this:

Inspiration Block

This displays any associated social media posts (the one we saw above). This also offers us a quick link to the post, the author’s account, and any discussion underneath this post to reference at any time.

Feature Definition

This helps to figure out exactly what our app needs to do to accomplish it’s goal, and exactly what the key feature will be. I’ve decided here that I could build a few interfaces for the same core logic, and defined these all as features.

With that, we already have a strong base for understanding what our product will be.

Validating Ideas

Before we rush into development, we can validate your idea with our target audience so we know we aren’t wasting your time for 3 active users.

Survey Builder

One thing I never tried to tackle as a dev was getting feedback. Of course it’s important and I knew it but it’s just been a challenge to get people to even take the time to take a look, let alone take a survey. As one person with a slim budget, I couldn’t pay testers or offer any monetary incentive for taking a survey. Or could I?

The useful thing that we’re building has value, it’s not built yet but it is speculatively real. What if we offer some “credit” incentive (early access, free product, x months of subscription) for users to take this survey.

We’re also in a very unique position that all of our ideas start on social media, so we can always track down the author of the original post and ask them some questions about what they want. Once you get that feedback, you can build an even better survey with incentives for people to give you even more feedback. Just like that we’ve gotten the engine to turn over and we’re getting helpful information for no up front cost.

One-Click Landing Page

Another useful thing to have at our disposal while we’re talking to all these people is someplace to send them when they ask about beta, or a waitlist, or want to be notified when the product is built. My secret for this is that all my landing pages come from the same tool and just collect to an email list that I’ve designated for each project. I’ve even got default copy for each of these cases, so I can just click an option instead of having to think.

With that, we can send out a few surveys, post a link to our landing page and wait for the market to confirm or deny our idea.

Is My Idea Viable?

Let’s say people like what we’ve come up with and have signed up for the waitlist, but can this product even survive? We need to estimate what it takes to run this tool, and budget that against what people are willing to pay.

Cashflow Summary

First, we need to define costs and revenue for each feature we added earlier. Costs and incomes can come in all shapes and sizes, and there are options to fit as many archetypes as possible. Costs can even be tiered, to account for user and usage limits on different compute providers.

I’ve built a summary tool that will automatically calculate our net with a given number of users, so we have a target to aim for. Features can even be toggled on and off so we can see the impact in realtime, and use this to prioritize features in an MVP.

Once our features are audited, the Cash Flow Summary automatically calculates monthly, yearly, one-time, and first-year revenue based on user count. This user count can be a target, a dream, a baseline, or whatever else. For this cash flow model, if we had 4400 users/mo consistently from day one, we could make money our first year.

We can also toggle features on and off, to see what the impact of the feature is on our cashflow and prioritize features for an eventual MVP.

However, services don’t start with a starter pack of 4000 users, so we need some way to model our costs along with user growth. Enter The Chart (I don’t have a better name yet).

Growth and ROI Charting

We can simulate linear growth or exponential with a growth rate percentage. This will calculate our net for the total duration, and tell us how long we’ll have to seed running costs. It also calculates the total running costs until ROI, and adds that to up front one-time costs to come up with an Investment Needed. In this case, ours is $9760.00 which covers development and 22 months of hosting the webpage. For a website only running on ad revenue, that’s decently offset (if not a little optimistic about $0.10/user).

This considers all cost and income intervals, including tiered, and adjusts the costs, income, and net for each month accordingly. (in this example, we never even left the first tier of our cost, since we were already in the green under 5000 users/mo.

This for me is the final decision on whether a product *can* be built. If people want it already, it’s financially viable, and I can reasonably build it and maintain it by myself (or bring a team in later) I’d start that same day.

With that, we can safely make our decision on if the project gets a green light, and move on to development, or start again from the scanner.

Access these tools

This toolkit has been useful to me and I’m using some tool from it every day, and working on other tools to help pick up the slack in other areas like presentations, testing, marketing. I want to make these tools available to other people, especially since this can enable better tools with more feedback, and if there is a small fee or advertising to cover costs then there’s opportunity to make tools that wouldn’t be possible to run for free.

I’ve made Cocoon, a dashboard for these tools where you can manage a project and use all of these tools on it in tandem.

Find an idea on social media
Build the outline
Define and prioritize features
Survey your audience
Build a waitlist
Calculate costs and revenue
Simulate growth to get funding and timeline estimates
If it works, get to building more confidently

In a time where lots of developers are struggling to find full time work and have dedicated themselves to solo and small-team projects, I hope it can be useful to someone else.

If you have some feedback or want to talk about accessing this app before I give it a proper release, give me a shout on Twitter or Bluesky, or sign up for the beta :)

More to read:

I finally tried Next.js

How To Create Useful Things By Yourself

Snippp v0.4.0 — Devlog

Snippp v0.3.0 — Devlog

Snippp.io — or How I Made My Organization Tool Into a Service

Snippets for Art

Intro to Polar Shapes

LUTs in Generative Art and Graphics Programming