GUIDE · INCIDENT RESPONSE

DDoS attack: preparation and response for businesses with online services

By the end of this guide you will know what to prepare before a DDoS attack, how to tell an attack from a legitimate surge in customers, and which steps to take during the outage. You will also see how to communicate, which mistakes to avoid, and how to verify that your defences can actually be activated.

Updated october 2026
7 min read
Altovar Technical Editorial Team

1. What a DDoS attack is and why small businesses are targeted too

A DDoS attack, short for distributed denial of service, sends so much traffic to a website or online service that it becomes slow or unreachable. The traffic comes from many different sources, which makes it hard to block with a simple rule. You do not need to be an important target: attackers often choose a business that is not prepared.

The damage is mainly commercial: orders not completed, customers unable to log in, lost bookings. Sometimes the attack is designed to distract staff while other access attempts are made. The heaviest cost is often confusion: those responding do not know whether the problem is technical or commercial. The problem needs to be addressed before it starts, because during an attack decisions are made under pressure and with incomplete information.

2. Preparation: what must be ready beforehand

Defence against a DDoS is built upstream, with your connectivity provider, with the provider hosting your services and, if used, with a dedicated mitigation service. These solutions absorb or filter abnormal traffic before it reaches your infrastructure. Ask your provider which protection is included and what limits are stated in writing. Also ask whether protection switches on automatically or requires a manual request, because the timing differs greatly.

You also need a clear view of your critical services: sales site, customer portal, bookings, email. Not all of them need the same level of defence, and if the system is saturated it helps to know what to keep running first. The playbook must list these services, with the owner of each and the supplier's contact. Also prepare a message for the status page, ready to adapt in a few minutes.

  • List of online services with priorities and owners
  • Emergency contact at the connectivity and hosting provider
  • Anti-DDoS protection included or purchased, with documented limits
  • Holding message or status page ready to publish
  • Procedure for distinguishing a legitimate surge from an attack
  • Alternative channels for customers if the website does not respond

3. The first steps, in order

When services slow down, the first task is to establish whether the problem is an attack, a fault or a surge in legitimate requests. The playbook should offer a quick verification route that does not require deciding everything from scratch. Record each step with time and observations, so you can compare trends over time. The route is there to avoid wasting time in discussion, but it does not replace the judgement of those who know the systems.

Activating mitigation or changing configuration should be decided by a designated manager, not by whoever notices first. The supplier must know in advance whom to call and what information to provide. A confused request slows down the response, whereas precise details, such as time, service and symptom, allow immediate action. It helps to keep these details on a short form, ready to send.

  1. Check whether the problem really affects the external service, including checking the internal connection.
  2. Compare incoming traffic with normal patterns and note the time the anomaly started.
  3. Notify the technical manager and the connectivity provider, with the data collected.
  4. Activate the anti-DDoS protection provided under the contract, following the written procedure.
  5. Keep critical services running and temporarily switch off non-essential ones, if the provider recommends it.
  6. Inform management and the communications lead, with a short, verified message.
  7. Record every configuration change made, so it can be reversed once the attack is over.

4. Telling an attack from a surge in customers

Not every rise in traffic is an attack. A promotion, an article going viral or an event can generate a genuine surge, which needs to be handled differently. Confusing the two leads to opposite errors: blocking real customers, or leaving the door open to attackers. For this reason the playbook should require comparing origin, type of requests and timing before deciding.

Typical signs of an attack are repetitive requests to the same pages, unusual origins, traffic that does not follow normal usage patterns, and no increase in sales or enquiries. No single sign is enough to reach a conclusion. The decision to block portions of traffic has a cost: it can also exclude legitimate customers. It should be taken by the technical manager, with management informed, and the reasoning recorded.

5. Mistakes that prolong the outage

The first mistake is trying to respond to hostile traffic manually, blocking addresses one by one. The source of the traffic changes faster than a team can keep up. Even a hurried response can make things worse, because every change must then be reverted. The second is changing the infrastructure in a rush without recording the changes, so that eventually nobody knows what was altered.

The third mistake is failing to communicate. Customers and partners who receive no news tend to look for it elsewhere, and improvised messages can create confusion. A short, verified and updated message is more useful than a prolonged silence, and it must stay consistent with what is published on the website and on social channels. Also avoid dealing with anyone who demands money to stop the attack without a legal and technical assessment.

6. How to get started

This week, run a desk simulation of a slowdown of your most important service. Ask the manager who will make the call, which provider will respond and what protection can be activated in practice. Note every point where someone replies 'we will look at it later': these are the gaps to close. A short simulation is enough to bring out the biggest gaps. Repeat the test after each correction, to verify that the gap is genuinely closed.

Then ask your connectivity and hosting providers in writing which anti-DDoS protections are included, what limits are stated and how they are activated. If the answers are unclear, the first job is to obtain them, before you need them for real. Keep the answers in the playbook, dated, so they remain available even when people change roles. Also check that the supplier's contact is not your only available channel.

Frequently Asked Questions

Is an antivirus or a firewall enough against a DDoS?

Usually not. A firewall can filter some requests, but it cannot absorb a volume of traffic that saturates the connection before it reaches your infrastructure. This is why protection must be planned upstream, with your connectivity provider and with services dedicated to mitigation. The firewall remains useful for other purposes, but it should not be considered the only defence.

Should I pay someone who demands money to stop the attack?

The payment demand is often part of the attack. Paying does not guarantee that the attack will end, and it may signal to others that you are willing to pay. The decision should be assessed with your legal adviser and your provider, and the competent authorities should be involved if the situation requires it. The playbook should say who makes this assessment, so the choice is not left to the moment.

How do I know whether traffic is an attack or a surge in customers?

Compare origin, type of requests, pages affected and timing with normal patterns. A genuine surge in customers usually produces requests to sales pages and results in orders or enquiries. An attack tends to concentrate on a few paths, come from unusual origins and produce no purchases. No single signal is decisive: the technical manager must assess them together.

Should I inform customers during an attack?

A short, verified message is generally useful, because it prevents customers from searching for information in a chaotic way. There is no need to describe the attack technically, nor to promise restoration times you cannot guarantee. The communications lead must approve the text and update it when the information changes.

Does a DDoS attack bring notification obligations?

Not automatically. It can become relevant if the attack causes a significant incident for an entity falling under the NIS2 Directive, on the security of network and information systems, transposed into Italian law by D.Lgs. 138/2024, or if it involves personal data. Under NIS2, the first report must be made within 24 hours. Check with your legal adviser whether your case falls within either scope.

About this content

General informational guide. Regulations and technical requirements change over time: always check the official sources and assess your specific case with a qualified technical or legal advisor.

GET IN TOUCH

Talk to our team.

We'll show how Altovar integrates into your infrastructure. No commitment, no sales pitch. Just a technical conversation.

BOOK A DEMO →