The largest website build we have attempted.

Thanks for stopping by the blog.

I know most of you probably came here to look at guns, night vision or whatever questionable financial decision I recently added to the website. Instead, today you are getting a behind-the-scenes look at what it has taken to build this thing.

This version has been the largest undertaking we have attempted to date. It is more complicated, more capable and considerably more work than any of the previous designs.

I am also more confident in this build than any version we have had before.

I like a challenge. That is probably obvious at this point.

However, if you own a business, have the available capital and are considering building a complicated e-commerce website yourself, I would strongly suggest hiring a professional.

There is a reason web developers charge what they charge.

This shit takes forever.

This is more than a new coat of paint

This was never supposed to be another basic website with a new homepage, a few product pages and a phone number at the bottom. The goal is to create a real operating system for Frontiersmen Firearms.

Every one of those items sounds simple until you connect it to everything else. Then you change one field and somehow the shopping cart forgets how shipping works.

Welcome to web development.

Building a cart for a gun shop is “fun.”

A normal shopping cart has a fairly straightforward job.

Put item in cart. Pay for item. Ship item.

A firearms shopping cart has more responsibilities.

• Identify whether the product is physically at Frontiersmen or sitting at a distributor.
• Apply the correct shipping rules for each inventory source.
• Recognize when a firearm must be shipped to another FFL.
• Allow the customer to select the receiving FFL.
• Handle a mixed order containing shop inventory, distributor inventory, firearms and ordinary accessories.
• Keep inventory-source labels visible on product pages, collections and inside the cart.

Then there are account requirements, identity verification, payment choices, transfers and all the other little details that become very large details when money and regulated products are involved.

It has taken a considerable amount of time to make the cart behave like an actual system instead of a decorative button that occasionally remembers what it is supposed to do.

The 60,000-SKU pain in the ass.

Distributor integration was supposed to make this easier.

That was adorable.

The feed is not giving me a clean, dependable pricing model for MAP and MSRP. The fields are not consistently usable, and the information cannot simply be dumped onto the website and assumed to be correct.

Getting either number wrong can create problems with manufacturers, customers or both. That means each product we add must be checked against information from the actual manufacturer.

One product at a time.

Talk about a pain in the ass.

Having access to 60,000 products is not useful if I cannot trust the data enough to publish them responsibly.

I could dump all 60,000 SKUs onto the website tomorrow and make the store look enormous. I would also create a giant mess filled with questionable prices, inconsistent descriptions and products nobody actually wants.

I am not doing that.

We are going to curate the inventory, verify the information and build something usable.

If you want specific ammunition, magazines, optics, firearms, night vision accessories or other products, tell me. Your answers will help determine what gets verified and added first.

The bulletin board is going to matter.

Out of everything being added to the website, the bulletin board may be my favorite feature.

It gives us a place to highlight local businesses, organizations, useful services, community resources, partners and people doing good work.

It is not intended to become another random directory filled with paid listings nobody reads. I want it to become something useful.

Frontiersmen has never existed in a vacuum. We work with local businesses, collectors, dealers, agencies, attorneys, families handling estates and people with specialized knowledge.

The bulletin board gives us a place to connect some of those people. We are going to use it much more in the near future.

Customer accounts and direct messaging

I have also been testing a direct-message function through customer account panels.

The basic idea is that a customer can log into an account and send a message directly to Frontiersmen.

Still, the messaging system could become a useful way to maintain good service as we scale. It gives customers another place to ask about an order, product, transfer or service without trying to remember where the conversation started.

Was it a text? Facebook? Email? A phone call while I was driving? Something shouted across the shop six months ago?

Putting more communication inside the customer account could make it easier for everyone to keep track of what is happening.

I will not be available every second of every day, but I want customers to have a reliable way to communicate directly with Frontiersmen.

Things are going to get crazy for us. I have stuff in the works. That is why we are putting in the work now.

Two AIs argued until the integration worked.

Yes, a lot of this website has been vibe coded.

For anyone unfamiliar with the term, vibe coding is when you describe what you want to an AI, let it write code and then spend several hours trying to understand why fixing one problem created three new ones.

At one point, I literally had Google Chrome open with Hostinger AI coding away while a ChatGPT sidebar had access to the browser.

One AI would suggest a solution. The other would explain why the first AI was wrong. Then the first AI would try again.

I basically allowed two AIs to argue with each other until they solved an integration problem I had been fighting.

That may not be the development process they teach in college, but the integration started working, and that is what mattered.

Somewhere, an actual web developer probably just felt a disturbance in the Force.

Carol has been bitching at me.

Carol has been bitching at me about how much time I have spent on this website.

In her defense, she is not wrong.

That does not mean eight peaceful hours sitting at a desk with a cup of coffee.

It means checking products, fixing filters, testing the cart, changing categories, working on accounts, arguing with integrations, writing content, creating images, checking mobile layouts and trying to understand why something that worked yesterday has suddenly decided it no longer believes in itself.

This is happening while I am still operating the actual business.

Apparently sleep is important.

Carol keeps bringing that up.

I am starting to suspect she may have a point.

Then I have to build the whole thing again

Once this website is functioning the way I want, I have to build another version for Frontiersmen Defense Systems.

That version will serve dealers, law-enforcement agencies, institutional buyers and government partners. It will need its own products, account structure, procurement tools, quote requests, agency information, dealer pricing and restricted access.

Basically, I need to take everything I have learned from this build and create another website with an entirely different audience and workflow.

That is going to be fun.

See, Carol? Web development can be fun.

Carol does not appear convinced.

Meet FRANK.

This is also a good time to introduce FRANK.

FRANK is the AI system I eventually want to build for Frontiersmen.

The long-term goal is to create an in-house AI that can help customers navigate products, understand services, find information and get pointed in the right direction without pretending it replaces actual human service.

FRANK will not replace me. Hopefully, he will prevent me from answering the same basic question 47 times while I am trying to finish something else.

The problem is that I need an in-home server powerful enough to host him.

Servers cost money.

This is apparently how normal businesses operate now.

You sell guns and night vision to fund the artificial intelligence living in your house.

Nothing weird about that.

Why put this much work into it?

Because I believe Frontiersmen is heading somewhere much larger.

The retail operation is growing. Night vision is becoming a major part of what we do. Frontiersmen Defense Systems is moving further into dealer sales, agency procurement and government-contracting work.

The systems supporting those businesses need to exist before everything gets busy.

I would rather struggle with the infrastructure now than wait until customers, orders, dealers and agencies are demanding something the system cannot handle.

If we do this correctly, customers should not need to think about most of it. They should be able to find a product, understand where it is coming from, place an order and know what happens next.

That is the goal.

Give us some time.

The website is not finished.

There are going to be bugs, missing products, incomplete sections and things that need adjustment after people begin using them.

I am sure someone will find a button I forgot to connect within six minutes of the site going live.

That is part of building something this large.

Give us some time, send useful feedback and understand that I am actively working on it.

This is the fifth design, the largest build we have attempted and the first version I genuinely believe gives us the foundation we need for what is coming.

I hope people enjoy the new site.

More importantly, I hope you enjoy what we are going to build on top of it.

Questions for everyone

Real feedback is considerably more useful than me guessing what 60,000 distributor SKUs people may or may not want.