SaaS monitoring
Upcheck watches your SaaS from where your customers are: an HTTPS request to every hostname every 5 minutes, the certificate verified twice a day, DNS and the domain registration checked on their own clocks — and one alert, in the channel your team reads, when any of them needs a human.
$10/month base · $1 per hostname · four checks on every one
What is SaaS monitoring?
SaaS monitoring is the continuous, automated checking of a software-as-a-service product from outside its own infrastructure, to confirm that customers can reach it and that it is being served correctly. A monitor requests each hostname on a schedule and records whether it answered, how fast, and with what status. It verifies the certificate the server presents, reads the DNS records that route customers to it, and watches the domain registration underneath all of it.
The distinction that matters is outside the stack, not inside it. Your APM agent, your Kubernetes probes, and your cloud provider's health checks all live inside the product. When the certificate expires, DNS stops resolving, or the load balancer serves a 502, every one of them is still reporting a healthy process. The customer, meanwhile, is looking at a browser error. External monitoring is the only vantage point that agrees with the customer.
The phrase is also used the other way round — monitoring the third-party SaaS applications your company depends on. The same external checks apply, and that use is covered below. This page is mostly about the SaaS you build.
How Upcheck monitors a SaaS application
- 01
Add every hostname your product answers on
The app, the API, the marketing site, the docs, the login host — each one is a website in Upcheck. No agent, no SDK, no code in your deploy. Upcheck reaches each hostname over the public internet the way a customer's browser does.
- 02
Four checks run on their own clocks
An HTTPS request every 5 minutes records the status and response time. A TLS handshake every 12 hours verifies the certificate. DNS is resolved every 6 hours and compared against the last answer. The domain registration is read daily over RDAP.
- 03
Alerts go where the on-call already is
An outage is confirmed by a second failed probe before anyone is paged, then delivered to email, Slack, Discord, Microsoft Teams, or your own webhook. Certificate, DNS, and domain warnings fire at the thresholds you set, and a recovery alert closes each outage.
The four ways a SaaS goes dark without a server failing
Uptime is the fastest failure and the only one most SaaS monitoring tools watch. The other three happen on slower clocks, outside your deploy pipeline, and each one ends with a customer who cannot get in.
- Every 5 minutes
The app stops answering
A 502 from the load balancer, a deploy that never came back, a database that filled its disk. Probed every 5 minutes, confirmed on the second failure, alerted with what the site answered and for how long.
- Every 12 hours
A certificate expires on the API host
The web app renews fine; api.yourproduct.com does not, and every customer integration fails its TLS handshake at once. Validity, chain, and hostname coverage verified every 12 hours, with expiry warnings weeks ahead.
- Every 6 hours
A DNS record quietly changes
A migration leaves an A record pointing at the old provider, or a nameserver change drops a record nobody documented. Resolution is checked every 6 hours and A, AAAA, NS, and MX changes are reported against the previous answer.
- Daily
The domain itself lapses
The registrar invoice went to a founder who left. Registration expiry and registry holds are read daily from the registry itself, not from a calendar reminder somebody set three years ago.
A SaaS is a set of hostnames, not one URL
The app, the API, and the marketing site usually sit behind different providers, carry different certificates, and fail independently. A monitor pointed at one URL to represent the product will report the marketing site healthy while the API has been refusing every integration for an hour.
Upcheck monitors and prices per hostname, so the right move is to add every one a customer or an integration can reach. Alert rules cover the whole workspace by default or scope down to specific websites — so the API pages the on-call phone while the docs site posts to a channel.
Monitoring the SaaS you depend on
Your product cannot take a payment if the payment provider is down, cannot log anyone in if the identity provider is down, and cannot send a receipt if the email API is down. Those outages arrive as your support tickets, not theirs — and the vendor's status page is updated by the vendor, usually after you already know.
Add the vendor hostnames your product calls as websites in Upcheck. Each one is probed every 5 minutes with the same confirmation, its certificate is verified, and its DNS is watched, exactly like your own. The alert names the hostname, so an outage at the auth provider reads as theirs rather than yours in the channel.
Be clear about what this is: a reachability and correctness check from outside, not an authenticated API call. It will tell you the vendor's API host stopped answering or is presenting a bad certificate. It will not tell you that a specific endpoint is returning errors for your account. For that you still want the vendor's status page and your own error tracking.
What to look for in SaaS monitoring software
The category runs from a checkbox on an observability platform to a single-purpose uptime pinger. Whichever direction you come from, these are the capabilities that decide whether it works at 3am.
- 01
Checks from outside your stack
Your APM agent runs inside the app. When DNS, the certificate, or the load balancer fails, the agent is either reporting healthy or not reporting at all. Only an external request sees what the customer sees.
- 02
Confirmation before paging
A monitor that alerts on every blip is muted inside a month, and a muted monitor is worse than none. Failure should be confirmed by a second probe before a human hears about it.
- 03
All four ways a SaaS goes dark
Downtime is the fastest failure and the only one most tools watch. The certificate, the DNS records, and the domain registration take a product down on their own schedules, and each needs its own check.
- 04
Per-hostname, not per-product
A SaaS is a set of hostnames with different providers, certificates, and blast radii. The tool should monitor and price each one, not ask you to pick a single URL to represent the product.
- 05
A webhook for your incident tooling
Slack is where people read alerts; the webhook is where your incident process starts. A structured event per outage lets you open an incident, update a status page, or page through your own rotation.
- 06
Honest scope
SaaS monitoring software that promises traces, logs, real-user metrics, and uptime in one bill is usually an observability platform with an uptime checkbox. Know which product you are buying, and what it will cost when the data volume grows.
Comparing against a tool you already use? See the honest UptimeRobot alternative and Oh Dear alternative comparisons, including where each of them wins.
What this is not
Upcheck answers one question well: can customers reach the product, and is it served correctly. These are the adjacent questions it deliberately leaves to other tools.
One probe location
Each request comes from Cloudflare's edge, not a quorum across regions. Confirmation comes from two probes separated in time, not geography. Regional outage detection and per-continent latency need a multi-region tool.
No logins or multi-step journeys
Upcheck requests the root of each hostname and reads the answer. It does not sign in, fill a form, or call an authenticated API. Synthetic transaction monitoring is a different product.
No logs, traces, or real-user metrics
Upcheck answers whether the product is reachable and correctly served. Why a request was slow inside your code is an APM question, and you should keep an APM for it.
No public status page
Alerts go to your team, not your customers. If you publish a status page, the webhook can feed it, but Upcheck does not host one.
Priced per hostname, with everything included
$10 per month for the workspace, plus $1 per website per month. A SaaS with an app, an API, a marketing site, and a login host is $14 per month for all four checks on all four hostnames. The per-website charge is metered daily, so a hostname added mid-cycle costs only the days it was tracked. Every alert channel and the full check history are included at every size, with no seat fees. A workspace can track up to 500 websites.
Want to see what a customer sees right now? The free website down checker requests any hostname from outside your network, no signup.
Frequently asked questions
What is SaaS monitoring?
SaaS monitoring is the continuous, automated checking of a software-as-a-service product from outside its own infrastructure, to confirm that customers can actually reach it and that it is being served correctly. At minimum that means requesting each hostname on a schedule and recording whether it answered, how fast, and with what status. Done properly it also covers the things that take a SaaS offline without the servers ever failing: an expired TLS certificate, a DNS record that changed, or a domain registration that lapsed. The term is also used for monitoring third-party SaaS applications a company depends on, such as its payment or email provider, and the same external checks apply to those hostnames.
What does SaaS monitoring software do?
It runs the checks a customer would otherwise run for you by complaining: an HTTPS request to each hostname at a fixed interval, a TLS handshake that verifies the certificate the server presents, a DNS resolution compared against the last answer, and a read of the domain's registration status. When a check fails, or a threshold such as certificate expiry is crossed, it alerts a responsible team where they already work. Upcheck does all four checks on every hostname you add, and delivers alerts to email, Slack, Discord, Microsoft Teams, or a webhook.
What is the difference between SaaS monitoring, APM, and observability?
They answer different questions from different places. SaaS monitoring asks, from outside, whether the product is reachable and correctly served right now. APM asks, from inside the application, why a request was slow or threw an error, using traces and code-level instrumentation. Observability is the broader practice of collecting logs, metrics, and traces so you can ask questions you did not plan for. An APM agent cannot tell you that DNS stopped resolving or that the certificate expired, because in both cases the agent is still reporting a healthy process. Most SaaS teams need external monitoring and an APM, and should not expect one product to be both.
What should a SaaS company monitor?
Every hostname a customer or integration can reach: the app, the API, the marketing site, the login or SSO host, the docs, and any customer-facing hostnames on white-labelled domains. For each one, monitor uptime and response time, the TLS certificate, the DNS records, and the domain registration. Then monitor the hostnames of the third-party services your product cannot work without, so you learn about a payment or auth provider outage from your own alert rather than from support tickets.
How often should a SaaS application be checked?
Upcheck requests every hostname over HTTPS every 5 minutes and confirms an outage on a second failed probe, so the worst case from going down to a confirmed alert is about ten minutes. Certificates are re-verified every 12 hours, DNS every 6 hours, and the domain registration daily, which matches how quickly each of those can actually change. Any website can also be checked on demand from the dashboard.
Can Upcheck monitor third-party SaaS apps my team depends on?
Yes, for any hostname reachable over the public internet. Add the hostname of your payment processor's API, your authentication provider, or your transactional email service and Upcheck probes it every 5 minutes and checks its certificate and DNS like any other website. It is a reachability and correctness check from outside, not a login or an authenticated API call, and it does not replace the vendor's status page. It is independent of it, which is the point: vendor status pages are updated by the vendor, usually late.
Does Upcheck monitor customer custom domains on a white-labelled SaaS?
Yes. Each customer hostname is a website in Upcheck at $1 per month, with the same uptime, certificate, DNS, and domain checks as your own hostnames. This is the case where external monitoring earns the most: a customer's CNAME or certificate on a domain you do not control breaks without any change on your side, and the first you hear of it is usually a support ticket. A workspace can track up to 500 websites.
Does SaaS monitoring replace a status page?
No. A status page tells customers what you know; monitoring is how you come to know it. Upcheck sends alerts to your team and can post a structured webhook event per outage that your status page tooling can consume, but it does not publish a customer-facing page itself.
How much does Upcheck cost for a SaaS?
$10 per month for the workspace plus $1 per website per month, metered daily. A SaaS with an app, an API, a marketing site, and a login host is $14 per month for all four checks on all four hostnames, with every alert channel included. There is no per-seat fee and no tier that withholds Slack, webhooks, or history.
Hear it from your monitor, not your customers.
Add the hostnames your product answers on and Upcheck starts watching all four ways they can fail — from the outside, with the alerts routed wherever your team lives.