One small change can break your tracking.
When someone makes a change live, you know within minutes. When your tracking breaks, you get an alert.
Free while we are still testing it (in beta). Sign in with Google.
Bring your own AI. Plug in the AI helper you use, and it can find and fix problems for you. See how
Google Tag Manager
Google Analytics
Slack
Your AI (MCP)
Tracking breaks without a sound.
Someone on your team makes a small change. Your site looks fine. But your sales and leads stop showing up in your reports.
You find out late
Nothing warns you. You find out weeks later, when a report looks wrong or a client asks why.
You can't see what changed
It is hard to tell what is different from one version to the next. So you guess.
Fixing it is slow
You dig through old versions to find the bad one. Each hour you wait, you lose data you can't get back.
Now you hear about it fast, not weeks later.
TagStrike tells you what changed, in plain words. It checks that your tracking still works. And if a change goes wrong, you can go back to a version that worked in a few clicks.
How it works
Here is what happens from the day you sign up.
A few words first
- Tag
- A small bit of code on your site. It sends info, like a sale or a form fill, to a tool like Google Analytics.
- Trigger
- The rule for when a tag runs. For example: when someone sends the contact form.
- Container
- The box in Google Tag Manager that holds all the tags for one site.
- Variable
- A saved value a tag can use, like a form's name.
- Event
- Something a visitor does, like sending a form.
- Publish
- To make your changes live on the site. Each publish makes a new version.
Step 1: Sign in and pick your containers
Sign in with Google. At first, we only ask for your name, email and photo.
Next, you let us look at Google Tag Manager. We can read, not change. Then you pick the containers to watch and group them by client.
Step 2: You see every publish within minutes
Every 5 minutes, we look at the live version of each container. So when someone publishes, we see it fast.
Is a container quiet for a week? Then we check it every 15 minutes. We save each new version from the day you connect.
Step 3: You get a Slack message in plain words
When someone publishes, you get one short message in Slack. It lists the top 5 changes. You see which tags, triggers and variables were added, changed or removed.
Another publish in the same hour goes in the same thread, as a short note.
One link opens the full list. It shows each field that changed, side by side.
Step 4: You get warned about risky changes
Each new version gets 10 safety checks. In the app, we call them lint rules. They catch mistakes like these:
- A sales or lead tag was removed or paused.
- The trigger on a sales or lead tag changed.
- An event name changed only in its capital letters, like Lead to lead.
- An email or phone number is typed right into a tag.
- Google Analytics is set up twice.
- Your container has a cookie banner tool. But an ad tag has no consent settings, so it may run before a visitor says yes.
Step 5: You know your tags still fire
A tag fires when it runs and sends its data. To test this, a real web browser opens your pages, just like a visitor.
When you connect, give us your site's address. We then set up a few checks for you. You can add more pages any time.
Each check runs every 6 hours. It also runs a few minutes after each publish, as soon as we see it. You can pick every 1, 6 or 24 hours instead.
If a check fails once, we mark it in the app. If it fails twice in a row, you get an alert.
Step 6: You hear when your real numbers drop
Some problems only show up with real visitors. So we can watch two kinds of numbers for you.
- Your key events in Google Analytics, like sales or leads. Each hour, we compare them to the same hour and day over the last 4 weeks. You hear from us if they stay far below normal for 2 hours, or drop to zero.
- The monitor tag, if you add it. It counts which tags and events fire for real visitors. By default, it counts 1 page view in 10. If a count stays below half of normal, you get an alert. For a tag or event, it must stay low for 2 hours in a row.
When a number drops, the alert names the last publish before the drop. One click shows you what changed.
Step 7: Go back to a version that worked
Found a bad change? Pick an older version that worked. A page asks if you are sure. You click to confirm, and that version goes live again.
The first time, you give us one more permission. We log each time someone goes back, with who did it and when.
Step 8: Or ask your AI helper to do the digging
Plug in the AI helper you already use. Then ask it things in plain words. Why did leads drop? What changed last week?
It looks at your versions, changes and alerts, and gives you the answer. It can even draft a fix. You say yes before anything goes live.
Built for agencies with lots of clients
You look after many client sites. Now you can see them all in one place.
One place for every container
See every client on one screen. A green, yellow or red dot shows which ones need you.
A Slack channel for each client
Send each client's alerts to its own channel. Your team sees what matters to them.
Less noise
Mute a type of alert for 24 hours. Follow-ups and "back to normal" notes stay in the same thread. Every Monday, you get a short recap of the week.
Works for one team, too
Just have one site? It works just as well for a team that runs its own tracking.
Use the AI helper you like. It does the digging. You say yes.
Finding out why tracking broke can take an hour. An AI helper can do that work for you.
You don't need a new AI tool. Plug the one you already use into TagStrike. It links up through MCP, a common way for AI tools to connect to apps. You make a key for it in the app.
Here are some things you can ask:
- "Why did leads drop on Acme Dental yesterday?"
- "What changed in the last version?"
- "Add a GA4 event when someone clicks Book a visit."
AI can make mistakes. So your AI helper gets a safety lock:
- It can read your tags, triggers, versions, alerts and check results.
- It can draft changes in a separate draft space, away from your live site.
- We run our safety checks on every draft and send you a note in Slack.
- Nothing goes live until a person on your team says yes in the app.
- It can never publish, delete a container or change who has access.
Works with the AI you use
These tools can plug in today with a key from the app. More tools soon.
- Claude Code
- Claude Desktop
- Cursor
- VS Code with GitHub Copilot
- Windsurf
- Gemini CLI
- OpenAI Codex CLI
Your clients' setups stay safe
You are giving us a look at your clients' tracking. Here is how we keep it safe.
We ask for a little at a time
Sign-in asks only for your name, email and photo. We ask for more only when you turn on a feature that needs it.
Your keys are locked up
The keys that let us reach your Google account are encrypted. That means they are scrambled so others can't read them.
Your data stays yours
Each team's data is walled off. No other team can see it. We never sell it or use it for ads.
You stay in charge
We never change your live site on our own. A person on your team always clicks to confirm.
Questions you may have
Does it change my container?
Not on its own. Most of the time, TagStrike only reads your containers. Two features can add things to a separate draft space in your container. They are the monitor tag, if you ask us to add it, and drafts from your AI helper. Nothing goes live until a person on your team says yes. That is also true when you go back to an old version.
What does it cost?
Nothing right now. It is free while we are in beta. There is no billing yet.
Do I need to add code to my site?
No. Most features need no new code. The monitor tag is the one extra, and you don't have to use it. It goes in Google Tag Manager, not in your site's code. We can put it in a draft for you. You choose when to publish it.
Who sees my data?
You, your team, anyone in the Slack channels you send alerts to, and any AI helper you connect. Our own small team can reach the system only to run it. We don't look at your data unless you ask us to. Or unless we must, to keep things safe or to follow the law. We never sell your data or use it for ads.
Can I use my own AI tool?
Yes. Make a key in the app and paste it into your AI tool. Today that works with Claude Code, Claude Desktop and Cursor. It also works with VS Code with GitHub Copilot, Windsurf, Gemini CLI and OpenAI Codex CLI. More tools soon. Your AI helper can look and draft. But a person on your team must say yes before any change goes live.
Does it work with server-side containers?
Not during the beta. For now it works with web containers. Those are the kind that run in your visitor's browser.
How fast will I hear about a change?
Most of the time, within about 5 minutes of a publish. If a container has had no change for a week, we check it every 15 minutes.
What do I need to start?
A Google account that can see your containers in Google Tag Manager. Slack is a big help, but you don't need it. To watch sales and leads, you also need access to Google Analytics. During the beta, you also need an invite code.
What if I remove your Google access?
Then we can't see your containers or your Google Analytics. So we stop looking for new publishes, and key event drop alerts stop too. The app tells you, and you can connect again at any time. The versions we already saved stay in the app. Scheduled browser checks and the monitor tag keep going until you turn them off.
Start watching your tracking today
Sign in with Google, let us read your containers, and pick the ones to watch. Add Slack, and you hear about the next publish within minutes.
Free while we are still testing it (in beta).
Sign in with Google
For now, we only ask for your name, email and photo. We ask to see your containers later, when you connect them.