PostHog has an alert system that almost nobody uses.

Which is wild, because your conversion rate can tank for three days before anyone notices. Leads just… stop. Traffic looks fine. Everything seems normal until someone finally checks the numbers and goes “wait, what happened last Tuesday?”

Alerts fix that. We’re going to walk through the two types that actually matter and one you should skip entirely.

How to Setup Alerts in PostHog

1. Create a trend in Product Analytics
2. Add the metrics you want to track
3. Save the insight
4. Click the three dots and select Create Alert
5. Set your upper and lower bounds
6. Connect it to Slack, email, or a webhook

That’s it. Once it’s saved, PostHog checks your metric on whatever schedule you set (daily, hourly, etc.) and pings you if it goes outside your bounds.

Now let’s get into the two types worth setting up.

Alert Type 1: Conversion Rate Monitoring

This one saves you when something breaks.

If your conversion rate drops below a certain percentage or spikes way above normal, something changed.

Form broke. Traffic source shifted. A dev pushed code that killed your funnel.

Whatever it is, you want to know now, not in a week.

Setting It Up

Go into Product Analytics and create a new trend.

Add two metrics:

• Metric A: Total page views (or unique users if you’re using the identify call)
• Metric B: Your lead generation event

Enable formulas and create: B / A * 100

That gives you conversion rate. Set the Y-axis to percentage, save it, hit the three dots, create alert. Bada bing bada boom.

Thresholds

We use 1% as the lower bound and 5% as the upper.

Below 1%? Something’s probably broken.

Above 5%? Also weird. Could be bots. Could be a tracking issue. Either way, worth checking.

Run it daily, connect it to Slack or email, assign someone specific if you want accountability. That’s it.

Alert Type 2: Volume Monitoring

Different thing entirely. Not about percentages. Just raw numbers.

You should be getting X leads per day. If you’re suddenly getting zero, that’s a problem.

Setting It Up

Create a trend with your lead event, save it, create an alert.

If you normally get 5-10 leads per day, set your lower threshold at 1. Simple.

Some of our clients run this hourly instead of daily. If you’re getting hundreds of leads, you don’t want to wait 24 hours to find out something broke at 9am. Run it every hour, average the last day, catch it fast.

Where Do These Live?

This trips people up all the time.

You set up like 15 alerts over three months. Where’d they go?

Data Pipelines → Destinations

They’re all there. History, when they fired, everything. That’s your central hub.

Skip This One: Error Tracking Alerts

PostHog has error tracking alerts. Don’t turn them on.

I know it sounds useful. It’s not.

What actually happens: you enable it, and suddenly someone’s getting 500 alerts because of some edge case JavaScript error that doesn’t even matter. Then people start ignoring alerts entirely. Or someone gets blamed for “not fixing the errors” when half of them aren’t real issues.

Just… don’t. Trust me.

Bonus: Manual Slack Notifications

You can use Data Pipelines to fire Slack messages on specific events. Not monitoring a metric. Just “when this exact thing happens, ping me.”

Works great for low volume, high intent stuff:

• Enterprise form submissions
• Demo requests
• High-value signups

If you get 2-3 enterprise inquiries per week, you probably want to know the second it comes in. Not in some daily digest.

Set it up under Data Pipelines → Create Destination → Slack. Filter for your specific event, configure the message, done.

You can throttle it too. Run once per 8 hours, cap at X events. Keeps you from getting spammed.

TL;DR for PostHog Alerts

Two alerts worth using:

• Conversion rate for catching broken funnels
• Volume for catching lead droughts

One to skip:

• Error tracking (noise)

One bonus:

• Data pipeline destinations for instant pings on specific events

Set them up once. Let them run. Get back to building.