A DDoS attack on a website shows as traffic far above normal from unusual sources, one page or login hit far harder than the rest, errors at checkout and alerts from the host. The business loses sales, leads and staff time until the site is back, and recovery outlasts the flood.
Last reviewed 2026-11-16 by Manish Sharma.
A DDoS attack is a flood of traffic meant to make a service unavailable, and the term explained with its main types has its own page. The signs are hard to read: the UK National Cyber Security Centre (NCSC) notes in its Denial of Service (DoS) guidance, last reviewed 25 March 2024, that traffic from many sources is harder to tell apart from real visitors. For the wider job of keeping public sites online when floods arrive, start with the use case page.
This guide covers the signs, how to read them and what the business loses. It says nothing about how attackers run an attack.
Signs of a DDoS Attack on a Website: Five to Check
Five signs point to an attack, and Signisys’s working view is that no single sign settles it. The sourced parts below name their source, and we read every page cited on 16 November 2026.
| Sign | What it looks like | What else could cause it | Who notices first |
|---|---|---|---|
| Unusual sources | Traffic from regions or networks the site does not serve | A campaign or a news mention | Analytics owner |
| One page hit hard | Login, search or checkout carries the load | A popular offer | On-call engineer |
| Errors at key steps | Timeouts and error pages at checkout and sign-in | A release gone wrong | Customers and support desk |
| Host or cloud alerts | A provider alert about an attack | A scheduled test | Cloud owner |
| Busy servers or a busy link | Servers load while bandwidth looks normal, or the reverse | A failing server | Infrastructure team |

Traffic From Unusual Sources or Regions
A rise from regions or networks the site does not serve is the first thing many teams spot. If a firm sells only in India and the GCC, a surge from elsewhere deserves a second look. You need no address list or method to see it: the analytics view by country is enough.
One Page or Login Hit Far Harder Than the Rest
AWS describes attacks that flood an application with queries that are valid for the target, such as web request floods, in its page How AWS Shield and Shield Advanced work, which shows no date. Because the requests look valid, the load lands on one page. Login, search or checkout carries it while other pages stay quiet.
Errors at Checkout, Login and Search
Timeouts and error pages appear where sales and sign-ins happen. Customers and the support desk notice first, sometimes before any dashboard turns red.
Alerts From the Host or Cloud Provider
Azure DDoS Protection can send alerts at the start and stop of an attack, and the Microsoft Learn overview of Azure’s DDoS service, dated 8 July 2025, says it gives reports during and after one. AWS says it provides Shield Standard automatically to AWS users, while Shield Advanced is a subscription for higher levels of protection. The AWS page lists Shield Advanced detection logic for application layer threats. Ask your host which alerts you receive.
A campaign, a release or a failing server can produce the same graph, so read the signs together.
Busy Servers on a Quiet Network, and the Reverse
Application floods load servers while bandwidth looks normal. Network floods fill the link while servers stay lighter. NCSC’s testing and monitoring guidance, last reviewed 25 March 2024, says to ensure sight of the network, compute and storage aspects of your service, so watch all three.
DDoS Attack or Busy Day: Reading the Website Signs
A DDoS attack and a good day for sales can look alike on a graph, so compare the shape.
A campaign spike follows a send or an ad, while an attack lands on few pages and little of it buys.
A campaign spike follows a send or an ad, comes from regions you serve, spreads across pages and converts into orders. A slow server follows a release or a change and shows no rise in traffic. An attack, however, does neither: the traffic is unusual, it lands on few pages, and little of it buys.
Teams also ask how to check a DDoS attack on a website from outside the network. Load the key pages from more than one location. Then ask upstream providers for alerts, which NCSC says they may offer. Failure from every location points to more than a local fault.
Load the home page, login and checkout from two or three places and note the time and error text.
What Happens During a DDoS Attack on a Website
In the first hour, each group sees something different, because each watches a different part of the site. Customers see pages that hang or fail. The support desk sees calls and tickets rise. The on-call engineer sees servers or links under load. The host or cloud provider sees traffic they may flag. Leadership sees sales stop.
Sales and leads stop where pages fail. Support calls rise as customers ask what is wrong. Staff drop planned work to deal with the problem, so the next release waits.
Key Takeaway The signs show first at checkout, in the support queue and on the host’s alerts.
The escalation path to your provider and the runbook for the day are a separate topic.
The Business Impact of a DDoS Attack on a Website
Four effects follow, and each one can be sized before an attack happens.
Lost Sales and Leads
Orders and forms fail while the site is down. NCSC names loss of sales among the effects of an unavailable site. No figure is given here, because the number depends on what your site earns per hour.
Support Load and Staff Time
Calls and tickets climb, and NCSC also lists employees who are unable to work. Engineers leave planned work to restore service, and that work slips.
Missed Service Levels and Partner Trust
Portals and APIs that partners depend on stop answering, and agreed service levels are missed. Customers who try and fail are a risk to repeat business, a loss to size with your own numbers.
Recovery Time After the Site Is Back
NCSC says an attack consumes time and money to analyse, defend and recover from. Expect review meetings, rule changes and reports after the site returns.

Sizing the Impact of a DDoS Attack for Your Website
NCSC’s first practice is to understand your service. To begin with, list every public site, portal and API.
For each, write down what it earns or serves per hour in your own numbers, who owns it and who notices a failure first. Signisys’s working view is to size the impact per site before choosing anything else, as a brochure site and a payment portal need very different responses. No figure here applies to your business, so fill the list with your own.
Key Takeaway Size the impact per site, in your own numbers, before you choose anything else.
Where to Read Next on a DDoS Attack on a Website
For advice on planning before, during and after an attack, see the plan for before, during and after an attack. The rest, however, sits in separate guides.
- Steps to detect, absorb and recover: see ordered steps to keep a site online during a flood.
- A scored readiness check: see eleven scored checks for a public site.
- The network layer and firewall comparison: see how network filtering and a web firewall differ.
Signisys provides these services and partners with AWS and Microsoft: see our network and zero trust security solution and security engineering and support services. A working session lists your public sites, their owners, who notices first and what each earns or serves, with no promised result and no price. No charge, with a fixed scope.
DDoS Attack on a Website: FAQ
How can you check a website from outside during a suspected attack?
Does a DDoS attack on a website touch customer data?
Who should be told first when the signs appear?
Does a slow website mean a DDoS attack?
Security Insights, Delivered.
Practical, vendor-neutral analysis of threats, tools, and architecture decisions.
