Most care plans promise “uptime monitoring”, and most deliver exactly that: a ping every minute that confirms the home page returns 200. It is honest and it is nearly useless, because the things that actually go wrong on a client site — the contact form that stopped sending after a plugin update, the checkout button that a theme change disconnected — all return 200. The client finds out from a customer, and then you find out from the client. Here is the version of monitoring where you find out first.
Whatever a client sells, their site has at least one flow that makes them money and one that makes them reachable. Monitor those two and you have covered the calls you actually get:
A PylonMon customer journey is a real browser that walks those steps on a schedule, from outside the client’s hosting, and pages you with the failing step and a screenshot when one of them stops working.
Add a monitor of type Customer journey, give it the client’s URL, and write the steps in the words on the page:
goto /contact fill Name = Probe Monitor fill Email = probe@youragency.example fill Message = [monitor] automated check, please ignore click Send expect Thanks
Yes, this sends a real email through the client’s form every run. That is the point: the test is whether the form still delivers. Make it cheap to live with: the subject or first line carries [monitor] so a mailbox rule files it, the period is fifteen or thirty minutes rather than five, and the from-address is one you control so a broken autoresponder lands on you, not the client. If the form uses a CAPTCHA that blocks automation, stop one step earlier — expect Send proves the form rendered with its button — and monitor delivery through the mailbox instead.
goto /product/monitoring-probe/ click Add to cart goto /checkout/ expect Billing details fill First name = Probe fill Email address = probe@youragency.example expect Place order
This reaches the payment step and stops. It proves the cart works, the checkout renders, and the payment section loaded its gateway — which is where most WooCommerce breakage shows up — without placing an order. The probe product is a hidden $0 item the client will never sell. If the client wants the order completed, a 100%-off coupon and click Place order / expect Order received finish it, and the orders arrive tagged with the probe’s email for cleanup. Shopify stores get the same treatment; the Shopify walkthrough covers the three safe levels of mocking a purchase there.
Everything sits in one workspace. Three habits keep it tidy at forty sites:
DOWN client-a.example · contact form
step 5 (click "Send") failed: expected text "Thanks" missing after 15s
re-checked once · screenshot attached · incident opened · on call: MariaMaria opens the screenshot before opening the site: the form is there, the button is there, and under it is a red line from the SMTP plugin about an expired token. Ten minutes later the token is rotated, the next run passes, the incident auto-resolves, and the client’s status page shows a twelve-minute blip with a one-line note. The client’s Monday email is you telling them, not them telling you.
A journey is a line on the care plan: “we test your contact form and checkout every fifteen minutes and fix it before you notice”. That is a different sentence from “uptime monitoring” and clients understand it immediately. On PylonMon the cost side is flat: Team is $30 a month for 250 monitors and 20 journeys with on-call rotations included, and when the twenty run out, another five journeys are $10 a month — prepaid, never a surprise. Two clients’ care-plan add-ons cover the whole account.
Want us to write the first journey with you? setup@pylonmon.com, no charge. The step language: the docs.