Operations
Synthetic Monitoring: can customers still sign up and log in?
Synthetic Monitoring is a robot customer that runs your sign-up and login from start to finish every 2 hours, and emails you the step that failed. From £39 a month, with nothing to pay up front. Running in 3 working days.
Why is checking that the homepage loads not enough?
A basic uptime checker only asks whether the homepage loads. The homepage can load while nobody can sign up, the welcome email never arrives or the login button does nothing - and those are the failures that cost money.
Those are the failures that cost money, and they are invisible from the front page. You hear about them days later, if the customer bothers to tell you.
The fix is a test customer that does what a real one does, from the first form to the last click, on a timer. That is what runs on GigXchange every 2 hours.
How is synthetic monitoring different from an uptime checker?
An uptime checker tells you the homepage loads, so it misses a broken login. Synthetic Monitoring tests the whole journey step by step every 2 hours, runs uptime checks too, and shows visitors a branded fallback page if the site dies.
| Waiting for complaints | A basic uptime checker | Synthetic Monitoring | |
|---|---|---|---|
| What is checked | Nothing | Does the homepage load | The whole journey, step by step |
| How often | Never | Every few minutes | Every 2 hours, plus uptime checks |
| Catches a broken login | Days later, from a customer | No - the homepage still loads | Yes, and which step |
| If the site dies | An error page | An error page | A branded fallback page |
| Customers see | Nothing until you post | Usually nothing | An honest status page |
What does Synthetic Monitoring check, and what do you get?
Synthetic Monitoring gives you a test customer running your journey 12 times a day, separate uptime checks on site, app and database, alerts naming the failed step, an honest status page, a fallback page and test accounts cleaned up.
- 01
The journey, tested
A robot customer signs up, confirms the email, logs in, resets the password and cleans up after itself.
- 02
Every 2 hours
12 full runs a day, so a broken step is caught within hours, not days.
- 03
Uptime checks
Your site, app and database checked separately, so you know which part failed.
- 04
Alerts that say where
An email naming the exact step that failed, not just "site down".
- 05
An honest status page
A public page showing what is working, measured from real checks - never a hand-typed "all good".
- 06
A fallback page
If the site dies, visitors see a branded "back soon" page that refreshes itself, not a raw error.
- 07
Clean test data
Every test account is deleted after its run and kept out of your sign-up figures.
What does one test run look like?
One Synthetic Monitoring run goes from the sign-up form to the welcome email, the email link, login, password reset and onboarding, then deletes the test account. Every step passes or fails, and a fail emails you.
Run 14:00 - all passed
- Sign-up form: ok
- Welcome email arrived: ok
- Email link verified: ok
- Account created: ok
- Log in: ok
- Password reset: ok
- Onboarding finished: ok
- Test account deleted: ok
- Next run: 16:00
- 12full sign-up and login runs a day
- 8steps checked on every run
- 5sign-up paths rotated in turn
- 3separate uptime checks: site, app, database
Source GigXchange's own figures, as of 25 October 2026
How quickly can Synthetic Monitoring be running?
Synthetic Monitoring is running in 3 working days. Day 1 maps your sign-up, login and any checkout. Day 2 builds the test customer, uptime checks and pages. Day 3 breaks a step on purpose to prove the alert reaches you.
- 01
Day 1
Map
We walk your sign-up, login and, if you have one, checkout, and agree what counts as a pass.
- 02
Day 2
Build
The test customer, the uptime checks, the fallback page and the status page set up.
- 03
Day 3
Prove
We break a step on purpose to show the alert fires, then switch it on for real.
Do you need synthetic monitoring on your site?
Synthetic Monitoring suits a site where customers sign up, log in or pay, and a broken step costs money the same day. It is the wrong buy for a site with no forms or logins, or a large shop needing checks every minute.
A good fit if
- Customers sign up, log in or pay on your site
- A broken step costs you money the same day
- You have no one watching the site at weekends
Not for you if
- Your site has no forms or logins
- You already run journey tests in-house
- You need checks every minute for a large shop
Questions
07Asked and answered
Synthetic monitoring is a robot pretending to be a customer on a timer. Instead of waiting for real people to hit a problem, a script fills in your sign-up form, waits for the email, clicks the link, logs in and checks each step worked. On GigXchange it runs 12 times a day and checks 8 steps, from the first form to deleting the test account.
From £39 a month, with nothing to pay up front for a standard sign-up and login journey. That covers the test runs, the uptime checks, the alerts, the status page and the fallback page. A longer journey, such as a full checkout with a test payment, can raise the monthly price, and we quote it before you start. There is no contract length, and the price does not rise with your traffic or the number of people who use your site.
An uptime checker only asks whether a page loads; this checks whether the journey works. Your homepage can load while the login is broken or the welcome email has stopped sending, and an uptime checker will say all is fine. We run both: Cloudflare health checks for site, app and database, plus the full journey every 2 hours. Each catches failures the other misses.
No, every test account is marked, deleted after its run and filtered out of your numbers. Each one uses a special email pattern, so your welcome emails, newsletters and sign-up counts skip it. If a run crashes halfway, a daily sweep clears up any account left behind. If you also take the Operations Command Centre, its sign-up figures exclude them automatically.
Yes, but it is written for your site during setup, not copied over. On GigXchange the robot tests sign-up and login, so checkout is the part built fresh each time. With Stripe we use its test mode, so no real card is charged. We agree what a pass looks like on day 1 and prove it fails properly on day 3.
Visitors see a branded fallback page instead of a raw error, and you get an email. The fallback page refreshes itself every 30 seconds, so people drop back into the live site the moment it recovers. The status page shows what is affected. What you do next belongs in a written plan, which is what the Incident & Continuity Plan gives you.
3 working days. Day 1 we map the journey and agree what counts as a pass. Day 2 we build the test customer, the uptime checks, the status page and the fallback page. Day 3 we break a step on purpose to prove the alert reaches you, then switch it on. You need to give us one test login and access to the inbox the test emails go to.
Yes, any month, with no notice period. When you cancel, the tests and alerts stop, the test email address is retired and any leftover test accounts are deleted. The status page and fallback page can be handed over to you to keep, or switched off. Nothing in your live site depends on the monitor staying on, so switching it off cannot break anything for your customers.