What the browser is really saying
You open your website and see a warning such as “Your connection is not private” or “Warning: Potential Security Risk Ahead”. The browser could not validate the secure connection to that address. Save the message and error code; do not bypass it to enter passwords or customer data.
It helps to tell this apart from other failures. A 500 error means the server did not know how to answer; a DNS failure means the browser never reached the server; a site down shows a hosting message or a timeout. This warning is different: the browser reached the server, started negotiating the encrypted channel and found something that does not add up, so it stops before painting the page.
That «something that does not add up» almost always has to do with the certificate your site presents, or with the chain of authorities that backs it. It is not automatic proof that you have been hacked: most of the time it is a certificate management issue, not an intrusion. The browser prefers to stop before sending visitor data through a channel it does not trust, and that caution gives you room to investigate calmly.
Common causes behind the warning
Before asking for help, get a sense of what may be causing the warning. In a small-business site these causes keep appearing:
- Expired certificate. Every certificate has an end date. If it is not renewed in time, the browser blocks the page.
- Name that does not match. The certificate was issued for
mydomain.comand the visitor comes in throughwww.mydomain.com, a new subdomain or an old domain after a change. - Missing intermediate chain. A certificate relies on an authority which relies on another. If the server does not deliver that chain, the browser cannot reach a trusted root.
- Self-signed certificate. The server presents one that does not come from a recognised authority. Common in test environments that, by accident, end up serving the public.
- HTTP resources inside an HTTPS page. They may be blocked or trigger mixed-content warnings. This is a related issue worth checking separately, but it is not the same as a certificate error that stops the initial connection.
- Incorrect date or time on the visitor’s device. The browser may treat a certificate as expired or not yet valid. This can explain why one device shows a warning and another does not.
- Corporate antivirus or proxy. Some antivirus tools and corporate networks inspect HTTPS with their own certificate. If it is not properly installed, the browser does not recognise it.
The first three usually mean the server must be looked at. The last two mean the problem may sit on the visitor's device. That distinction shapes who to call.
Before you touch anything: check and document
The exact message and error code help support distinguish a site problem from a device or network issue. You can gather this information without changing server configuration:
- Take screenshots of the warning where it appears, with the full URL and the exact text.
- Expand the detail and note the error code: Chrome shows a
NET::ERR_..., Firefox aSEC_ERROR_...orSSL_ERROR_.... That code is the most useful clue for whoever diagnoses it. - Try another browser on the same device. If Chrome flags it and Firefox does not, the problem may sit in the certificate chain or the local configuration.
- Try another device and another network: mobile on cellular data, a home connection against the office one. If it only appears on the company network, look at the proxy section.
- Open the site in a private window and record whether the warning changes. This uses a separate session, but does not rule out every extension or a problem with the website.
- If it happens only on a company network, ask its IT team whether HTTPS inspection is involved. Do not disable antivirus or network protections as a test.
- Write down how long it has been happening, which pages show it and whether any recent change happened (domain renewal, hosting change, new installation).
This preservation gives you real data to share with whoever steps in and prevents rushed decisions, like forcing a renewal when the issue was the device clock, or disabling HTTPS to «see if that works».
How to read the error detail
The code the browser shows is not decorative. Each one points to a different cause, and though names change across browsers, the operational meaning is similar. The ones that appear most often:
| Code | Operational meaning | Where the cause usually sits |
|---|---|---|
NET::ERR_CERT_DATE_INVALID | The certificate date is not valid: it has expired or is not yet effective. | Certificate outside its validity period, or an incorrect date and time on the visitor’s device. |
NET::ERR_CERT_COMMON_NAME_INVALID | The domain in the URL does not match the names covered by the certificate. | Certificate issued for another domain, missing www or a subdomain, recent name change. |
NET::ERR_CERT_AUTHORITY_INVALID | The browser does not recognise the authority that signs the certificate. | Missing intermediate chain, self-signed certificate, or antivirus inspecting HTTPS without proper configuration. |
SEC_ERROR_EXPIRED_CERTIFICATE | Firefox equivalent to an expired certificate. | Pending renewal or incorrect date and time on the visitor’s device. |
SSL_ERROR_BAD_CERT_DOMAIN | Firefox detects that the certificate is not valid for that domain. | Same scenario as COMMON_NAME_INVALID: certificate issued for another name. |
The table orients the investigation but does not close the case. Two sites with the same code can have different causes: one because the server does not deliver the complete chain, another because the visitor's antivirus injects its own certificate. For the technical base behind these codes, the Mozilla reference on TLS and the official Chrome help are useful reads.
Assess how much of your site is affected and the real risk
With the code written down and tests from different devices in hand, it is worth sizing the problem before asking for an intervention. The response depends on how widespread the warning is:
- Who sees the warning? If only you see it, it may be your device or network. If every visitor sees it, the server or certificate is almost certainly involved.
- Which part of the site stops? It is not the same to affect only a new landing page, a blog subdomain, the customer area or the home page. The more central, the more urgent the response.
- Since when? Ask the team if there was a hosting change, domain change or a renewal that did not finish cleanly.
- What impact can you observe? Record customer reports and interrupted functions. Analytics may never run if the browser blocks the page; a traffic drop does not count how many people saw the warning or prove its cause.
- Does it affect forms or payments? If the warning appears where the visitor was about to leave data or pay, commercial impact adds to the technical one.
Based on the answer, decide what to communicate while you work on it. If only you see it, no need to notify outside the team. If everyone sees it and the site is your main channel, a sober note through another channel (email, social, business profile) prevents anyone from thinking you have disappeared. Better to avoid long technical explanations before knowing the cause.
Who to call and in what order
Start with whoever manages the certificate or the affected service. Depending on the cause, they may need to coordinate with another provider. Knowing what each does avoids bouncing back and forth:
- Hosting provider. If the certificate is managed through your hosting panel, start with support to check renewal, installation and the certificate chain. They can also check how the domain is configured on the server.
- Domain registrar. Whoever sells you the domain does not have to manage the certificate, unless they also host the site. Calling them for this warning wastes time, unless the problem is DNS or the domain name.
- Website maintenance. Checks HTTP to HTTPS redirects and whether any page tries to load resources over HTTP. If the problem is with the server or certificate installation, they coordinate with whoever manages it.
- Internal team. If the server is in-house or there is a corporate proxy inspecting traffic, the internal team watches the server clock and the certificate policy.
As the owner, what you can do without touching the server matters: contact the right provider with the error code and screenshots, document what you tried and authorise the intervention once it is explained. If the fix is not clear, asking for a short explanation is entirely reasonable.
If nobody really owns the certificate renewal or the maintenance contract has been abandoned, you are facing a bigger problem than this warning and it is worth considering how to prepare a provider handover properly.
Mixed content and HSTS: two different things to check
Mixed content occurs when an HTTPS page requests images, scripts or other resources over HTTP. Blocking those resources may break parts of the page, so the team needs to review their addresses. This differs from an invalid certificate on the initial connection. MDN’s documentation explains the distinction.
HSTS tells the browser to use HTTPS for that domain. If the certificate fails, it prevents continuing over an invalid connection. This protects visitors: the remedy is to correct the certificate or its configuration, not ask people to bypass the warning. The HSTS reference explains how it works.
Depending on the cause, work may involve hosting, a service that delivers the website from other servers, or the page itself. Our guide to HTTPS, WAF and CDN helps distinguish their roles.
Preventing a repeat
Once it is resolved, agree who checks expiry and verifies renewal. The warning alone does not prove neglected maintenance, but it is a useful reason to review these responsibilities:
- Expiry calendar. Write down the end date of every certificate your site uses (main domain, subdomains, email, connected services) and who renews it. If no one owns it, that is already a problem to solve.
- Renewal alerts that do not drown. Automated emails from the certificate authority or hosting must reach an address someone actually reads.
- External checks. An alerting service can check the website’s response and certificate validity. Test where alerts go and who acts on them rather than assuming it detects every fault.
- Clear responsible party in the contract. The contract should state who looks after the certificate and what happens if it is not renewed. If it does not, asking for it to be added prevents future arguments.
- Minimal renewal protocol. A short document with steps and contacts stops renewal from depending on a single person's memory.
Fitting this inside a periodic technical review, instead of a loose task, is what turns it into a habit. The piece on website maintenance and technical review describes how to organise that review without making it a burden.
If you want us to walk you through it
If you need help understanding the warning, tell us the message, browser and error code. We can assess which review is needed and agree scope, price and timeframe before making changes.
For an explanation of what an encrypted connection protects and what it leaves outside its scope, see our guide to HTTPS, WAF and CDN or our security page.