New stores: 14 days of Protector freeFirst 25 stores.Uptime, daily scans, Slack and Telegram alerts. No card.See what’s included

Store Monitoring

Checking the store twice a day means finding out up to twelve hours late.

Almost every store owner monitors their store already — by opening it. It is a reasonable habit and it catches obvious problems. What it cannot do is bound how late you find out, and lateness is where the cost sits. This is the arithmetic of manual checking versus automated alerting, and the specific failures each one is good and bad at.

Scan your store free See pricing

Free plan, no credit cardRead-only connectionAlerts to email, Telegram, Slack or Discord

The arithmetic of a manual check

A manual check has an average detection delay of half the gap between checks. Check twice a day and the average is six hours, with a worst case of twelve — and the worst case lands overnight, which is also when you are least likely to be checking and, for a store selling internationally, not necessarily when you are quietest.

The gap is not the only problem. Manual checks are correlated with your schedule, not with risk: they cluster around working hours, they stop on weekends and holidays, and they thin out exactly when trading peaks. Black Friday is the day you are least able to check every hour and the day an hour costs the most.

How you check	Average delay	Worst case	Runs at 3am?
Twice a day	6 hours	12 hours	No
Every few hours during work	1–2 hours	16 hours overnight	No
Customer tells you	Unbounded	Unbounded	Only if they email
Automated check every 5 minutes	2.5 minutes	5 minutes	Yes
Order-flow analysis every 1–3 hours	30–90 minutes	1–3 hours	Yes

What manual checking genuinely catches

This is not an argument that looking at your own store is worthless. A person notices things no check is configured for: a layout that broke on mobile, an image that is the wrong product, copy that reads badly, a promotion that reads as confusing rather than compelling. Judgement is the part automation does not have.

Visual and layout problems that render fine but look wrong.
Copy, tone and merchandising mistakes.
Anything where the question is “is this good?” rather than “is this working?”

What manual checking reliably misses

The failures that hurt most share one property: they are invisible from the homepage. You cannot see them by looking, which is why the habit feels reassuring right up until the month-end report.

Checkout failing while every other page loads normally — a 200 OK all the way to zero orders.
A price error on one variant out of eight hundred, or a bulk edit that hit the wrong collection.
A certificate expiring at 4am on a Sunday.
A gradual decline in orders — down 20% for three days is invisible day to day and obvious only in hindsight.
A sold-out bestseller you are still paying to advertise.
Broken images or links on deep pages nobody browses to from the homepage.

The useful division of labour

Automation is for the questions with a definite answer — is it up, are orders arriving, did a price change, is the certificate still valid, did a link break. Those questions have thresholds, they can run every few minutes, and they do not get bored on a Sunday.

Keep the manual pass for judgement, and let it be occasional rather than anxious. The point of monitoring is not that you stop looking at your store. It is that you stop needing to look in order to know.

What Cassian™ actually does

Specific checks, with the plan they run on.

Cassian™ analyses order flow against a weighted historical baseline every 1 to 6 hours depending on plan, and raises a critical alert on a period that should have orders but has none.

Every plan. 3-hourly on the free Inspector plan, hourly on Guardian and above.

Significant-drop detection compares the current period against expected volume, and gradual-decline detection catches the multi-hour slide a daily glance cannot see.

Significant drop on Analyst and above. Gradual decline on Protector and above.

Cassian™ compares every product variant against yesterday’s snapshot, so a single mispriced variant in a catalog of thousands is found without anyone opening a product page.

Every plan with a connected store.

Alerts route to email on every plan, and to Telegram, Slack or Discord on Protector and above, each with its own minimum severity — so the overnight channel only carries what is worth waking up for.

Email on every plan. Telegram, Slack and Discord on Protector and above.

FAQ

Common questions

Is automated monitoring better than checking my store myself?

They catch different things. A person is better at judgement — layout, copy, merchandising, anything where the question is whether something is good. Automation is better at questions with a definite answer, and better at asking them at 3am. The measurable difference is detection delay: checking twice a day averages six hours late, an automated check every five minutes averages two and a half minutes. Most owners keep both, and stop checking anxiously once something else is watching.

Won’t automated alerts just become noise I ignore?

They will if they are configured badly. Two things prevent it: confirming a failure before alerting, so a single dropped request never pages you, and per-channel severity floors, so critical incidents reach your phone while informational findings collect quietly in email. Cassian™ also groups related changes — twenty price changes from one bulk edit is one alert, not twenty.

How long does it take to set up?

Connecting a store is a few minutes and read-only. Zero-order alerting works from the first baseline. The one thing worth doing deliberately is choosing which severity goes to which channel, which takes about as long as reading this page.

Keep reading

Store Monitoring Real-Time Sales Alerts Downtime Cost Calculator Signs You’re Losing Sales Order Pulse — order-flow monitoring Alert channels and connectors

Imagine if your store could tell you when something's wrong.

Free scan, results in minutes. Zero-order alerts on every plan, including the free one.

Scan your store free See pricing