Why I Built Hoppla

There is no shortage of monitoring tools. Datadog, New Relic, Uptime Robot, Oh Dear, Pingdom, Better Uptime, Cronitor. The list keeps going. If all you want is to know whether a website is up, you have dozens of options.

So why build another one?

Most monitoring tools do too much

Strange thing to say as someone who just shipped a monitoring product, but here it is: most monitoring tools are overbuilt.

Not bad. Overbuilt. They show you everything, and most of it you'll never look at. You log into Datadog and there are 47 dashboards, 200 custom metrics, APM traces, log aggregation, infra maps, and a bill that makes your CFO twitch. You wanted to know if the site is up. Now you're debugging a query to find out why your custom dashboard stopped rendering.

Even the focused tools (the ones that only do website monitoring) pile on features that add weight without touching the actual problem. So you spend your evening configuring the tool instead of, you know, monitoring your sites.

This kept happening to me. Sign up for a service, lose an hour to setup, then realise I was using maybe 20% of it. The rest was noise. Features built for someone else's job, cluttering my dashboard, firing alerts I didn't care about.

I wanted something opinionated

What I needed was small. Watch my sites, tell me when something breaks. Not "here are 15 graphs of your response-time distribution over the last 72 hours." Just: is the site up, is the cert about to expire, did the DNS records change, is the cron job still running?

I wanted a tool that decided things for me. Check from more than one place and only wake me when the site is actually down, not because one route flapped. Funnel every alert into one channel instead of making me write routing rules. Show me red and green, not a wall of time-series.

Opinionated software is underrated. The tools I reach for every day (Laravel, Tailwind, Linear) all have a strong view on how things should work. They make the call so I don't have to. I wanted that for monitoring.

So I started building

I've been a developer a long time. Started in PHP, moved into the Laravel world, ended up leading a small team. At some point curiosity won and I started prototyping.

It wasn't a straight line. I tried a few architectures, experimented with different ways to run the checks, and threw away more code than I kept. The queue-driven setup Hoppla runs on today, a scheduler dispatching checks to Redis Streams workers, was the third attempt. The first two worked. They just didn't scale the way I wanted.

I tested on my own sites first, then a handful of client sites. The check types grew one at a time: uptime, then SSL, then DNS, then cron, then broken links, then Lighthouse. Each one because I actually needed it that week, not because a roadmap told me to.

The GEO audit, the AI-readiness scoring, came out of plain curiosity. AI search was clearly growing and I wondered how well my clients' sites were set up for it. Nobody was measuring it, so I built a check that does: 31 separate signals across structured data, AI crawlability, content structure and technical metadata. Turns out nobody else in monitoring offers it yet.

At some point I had a tool that did exactly what I wanted and nothing else. Clean dashboard. Alerts that meant something. Two-region confirmation so I stopped getting woken for nothing. It just worked.

The reseller problem that became the business model

Here's where the story turns in a direction I hadn't planned.

I was running websites for clients and naturally wanted to offer monitoring alongside. The snag: every tool I looked at was either too expensive to resell or didn't do white-labeling at all.

Walk through it as an agency. A tool charges €49/month for 100 sites, you manage 30 client sites, you pay €49/month. To make any margin you'd have to charge your client more than they'd pay going direct. That isn't a business model. Your client Googles the tool, sees the price, and wonders why you're adding a markup.

The enterprise tools are worse. Datadog and New Relic bill per host, per metric, per whatever, and by the time you're monitoring enough to matter you're staring at hundreds or thousands a month. Nobody resells that to a small business on a WordPress site.

So there was a gap: a monitoring tool priced for reselling. Cheap enough that an agency or hosting company can fold it into a standard package, charge a fair price, and still keep a margin. With white-labeling that genuinely works, not a $500/month enterprise add-on but a core feature.

That's what Hoppla turned into. The tool I built for myself, priced and white-labeled so other people can build a service on top of it. The model is wholesale per monitor, tiered by volume. You set what your clients pay.

What "white-labeling that works" actually means

Let me be specific, because most tools that say "white-label" mean "you can swap the logo in one spot."

In Hoppla, white-labeling means:

• Your brand name, logo and colours everywhere: the dashboard, alert emails, status pages
• Alerts sent from your own email domain ([email protected]), not from ours
• Status pages on your client's custom domain (status.clientname.com), not a hoppla.dev subdomain
• A reseller API so you can provision tenants and manage monitors programmatically
• Portal domains, so you can host the whole monitoring dashboard on your own domain
• Your clients never see the word "Hoppla" anywhere

This isn't an enterprise-tier line item. It's in because it's the entire point. If you can't resell it under your own brand at a price that makes sense, white-labeling is just a checkbox in a feature grid.

The opinionated choices I made

A few decisions that shape how Hoppla works, and why.

Dual-check verification by default. When a check fails at one location, it's re-run from a second one before anything alerts. On by default, not a toggle you have to find. I was done with 3am false alarms.

One alert pipeline. Every check type (uptime, SSL, DNS, cron, all of it) runs through the same alert system. One set of rules, one set of channels. I'm not building a tool that makes you configure routing per check type.

Status pages included. Not a separate product, not an add-on. Every plan gets status pages; putting them on your own domain starts at Starter. If you're monitoring something, you probably want a way to tell people how it's doing.

No per-seat pricing. I can't stand per-seat pricing. Your whole team should be able to open the tool without anyone doing math on whether the next seat is worth it. You pay for how many monitors you run, not how many people look at the dashboard.

Tiers change size, not shape. I'd love to tell you every feature sits on every plan. It isn't quite true: the heavy checks (Lighthouse, GEO audits, crawls) start at Starter, and the small tiers carry four notification channels instead of all seven. What a tier mostly changes is how much you can watch — monitors, status pages, how long the history sticks around. There's no "contact sales" tier, and nothing important hides behind "Business."

Where things are now

Hoppla is live and watching real sites. I use it every day. The core is solid: uptime, SSL, DNS, cron, broken links, Lighthouse, GEO audits, server agents, security headers, status pages, on-call schedules and escalation policies.

It's still early, and I'll say the honest part out loud: this is monitoring, not an APM. No request traces, no flame graphs. If you need to profile a slow database query, Hoppla isn't your tool. It tells you what broke and where, then gets out of the way. I ship something most weeks: more probe locations, tighter integrations, whatever I trip over while using it myself. The short feedback loop is the best part of building for yourself.

If any of this lands, if monitoring tools have felt like either too much or too expensive to resell, try Hoppla. It's the tool I wanted when I started looking.

And if it doesn't fit, I'd honestly like to hear why. Still an engineer at heart, and blunt feedback is the most useful thing you can send me.