Day Seven: The Night It All Connected

Day seven. The day everything I had been building separately finally clicked into one connected system.

Let me take you deeper into this one, because today was as much about the engineering underneath Pawtrips as it was about what you can see on the surface. If you only want the story, stay with me. If you want to understand how the thing actually works, even better.

The bug that had been lying to me

For days, some of my changes had not been saving.

I would make an edit. The system would confirm success. I would move on. And quietly, invisibly, the change had not actually stuck to the file.

I caught it when I went to add affiliate product recommendations across the whole site. I had been adding them to dozens of guides over a couple of sessions. When I ran a count to check my progress, it came back saying only seven guides had them.

Seven. After all that work.

Finding the cause instead of thrashing at it

The old me would have panicked and assumed everything was broken.

Seven days of building has taught me to do something different. Stop. Say it out loud: this is fixable. Then go find the actual cause rather than swinging at the symptoms.

The cause was almost elegant once I saw it. The way I had been making large changes was by pasting big blocks of code into the server's command line. When those blocks got large enough, the terminal was quietly corrupting them as they came in. Dropping the occasional character. Breaking the script's structure. So the script would run against a broken version of itself, fail to do what it was meant to, and report success anyway.

It had been confidently wrong for days.

The fix, and why it matters technically

Pawtrips content lives in code. Every guide is an entry in a large structured file, with its title, its sections, its safety warnings, and now its product recommendations. That file is the single source of truth. When I want to change the site, I edit that file, commit the change to version control, and push it. A deployment pipeline picks up the change and rebuilds the live site automatically. No manual uploads, no downtime.

That pipeline is the part I moved onto properly this week, and it is the reason the bug was so sneaky. The pushing and deploying worked perfectly. It was the local editing step, the pasting, that was broken. The healthy half of the system was faithfully publishing the broken output of the unhealthy half.

The fix was to change how I make big edits. Instead of pasting a large script into the terminal, I now write the script to its own file, transfer that file across cleanly, and run it directly. No pasting means no corruption. Then, critically, I verify the result with a quick count before I push anything live. Trust nothing until the numbers confirm it.

That last habit, verify before you trust, is the single most valuable thing I learned this week. The bug taught it to me the hard way.

Seven became fifty-two

With the reliable method in place, I ran the real job.

Seven guides with product recommendations became fifty-two in one clean pass.

But I want to be clear about what that actually involved, because it was not just dumping links everywhere. I mapped specific, relevant products to each guide based on what someone reading that exact guide genuinely needs. The first aid guides got proper first aid kits and wound care. The senior dog guide got the orthopaedic bed and the car ramp that protect ageing joints. The rescue dog guide got the GPS tracker that gives anxious new owners real peace of mind. The water safety guide got the life jacket with a rescue handle.

The system I built for this is reusable. Each product is a small structured object with a title, a category badge, who it suits, an honest description, the price, and the link. They are designed to slot in after specific sections of a guide so the recommendation appears exactly where it is relevant, not bolted onto the end. I wrote the logic so it even checks how many sections each guide has and adjusts placement automatically, so a recommendation never points at a section that does not exist.

I checked the result three times before letting it go live. Fifty-two earning pages where this morning there were seven.

The deals engine

The other big build today was making the commercial side of the site feel alive.

The first version of the deals section on the homepage was too quiet. It used the same colours and styling as everything around it, so it blended in and got lost. People scrolled straight past it.

So I rebuilt it as something deliberately different. A dark, self-contained section that breaks the visual rhythm of the page. Bold red deal badges. A row of products that moves on its own, slowly scrolling so it feels active without being annoying, and pausing when you hover so you can actually read it. The whole thing is built to make you stop scrolling for a second, which is all a good deal section needs to do.

I paired it with a dedicated bulk buys page, because I know from how people search that "bulk buy" and "value pack" are real, high-intent terms. Someone searching those words is ready to spend. Meeting them with a page built exactly for that intent is just good sense.

And I added an Our Picks section to the navigation, front and centre, so the products I genuinely stand behind are never more than one tap away from anywhere on the site.

How this actually scales

Here is where my head is at on the bigger picture, because today made the path clearer than it has ever been.

Right now Pawtrips is a content platform with a commercial layer. Guides bring people in through search. Product recommendations and a growing email list turn that attention into something durable. That is the foundation, and it is working.

The next technical step is moving the content out of code and into a proper content management system. At the moment, adding or editing a guide means touching the codebase. That is fine for me, but it does not scale and it slows down the one thing the site needs most, which is more content faster. A visual editing system means I can write, edit, and add images to guides without writing code, and it opens the door to eventually having others contribute too.

After that comes the feature I am most excited about. A trip planner that asks you a few simple questions, your dog's breed, your state, how many days you have, what you love doing, and then builds you a genuine personalised itinerary drawn from the real guides on the site, with the relevant gear recommendations layered in. The reason this is powerful is that it is only as good as the content behind it, and Pawtrips already has the deepest, most specific Australian pet travel content anywhere. The planner turns that depth into something no competitor can easily copy. It also collects an email and gives me real data on what people actually want, which tells me exactly what to build next.

Beyond that sit the things that make it a business rather than a blog. Verified property listings. A trust badge program for genuinely pet-friendly stays. A community where owners answer each other's questions, which compounds into search traffic on its own. Each layer funds and feeds the next.

On the work itself

I am obsessed with this. I will say it plainly because this diary is built on honesty.

I work at Bunnings. I do not have a computer science degree. I learned every piece of this by sitting down and refusing to get up until I understood it. A bug that quietly broke my work for days did not make me want to quit. It made me want to understand it so completely that it can never happen again. That instinct is the whole reason I think this might work.

The research backs the direction. The feedback from people I have shown it to is the kind you cannot fake. The data says build it exactly the way I am building it. What is left is simply the work. Keep showing up. Keep building with care. Trust that effort this consistent, on something this needed, eventually gets noticed.

Seven days in, Pawtrips is a fast, properly architected platform with over a hundred guides, fifty-two of them now earning, a growing list, a lead magnet, and a commercial engine that finally feels alive.

Tomorrow

A search bar so people can find the right guide in seconds. Template fixes so every guide feels as polished as the best ones. Then the content management system that lets all of this scale.

But that is tomorrow.

Today, the disconnected pieces became one connected system. And it is live.

Day seven. The night it all connected.