Customer Service
Help Desk Software: Features, Types and When AI Pays Off
Help desk software brings service requests into one place. Features, types compared, the threshold where AI starts to pay, and the metrics that prove it.
Help desk software brings service requests from email, phone, chat and contact forms into one place, manages them as tickets and makes their handling traceable. It answers the question of who is working on what and what is still open. What it does not answer by itself is how many of those requests needed to exist at all. That is where AI comes in, and that is where it either pays off or does not.
Key takeaways
- Definition: Help desk software collects service requests across channels, manages them as tickets and documents how they are handled.
- Distinction: The service desk is the broader, ITIL-defined term. In customer service, the help desk is usually what is meant.
- Six core features: shared inbox, ticket management, assignment, SLA monitoring, knowledge base, reporting.
- AI pays off when volume, repetition and documented answers come together. Without documentation, no model helps.
- OMQ: The Ticket System Assistant works inside your existing help desk, while the Intelligent Help Page, Smart Contact Form and AI Chatbot resolve enquiries before a ticket is created.
- What is help desk software?
- Help desk or service desk?
- Which features does help desk software have?
- Which types of help desk software exist?
- When does AI pay off in a help desk?
- Where AI reaches its limits
- Which metrics show the effect?
- What should decide the selection?
- What do GDPR and the EU AI Act require?
- Extending your help desk with OMQ
- Key terms at a glance
- Conclusion
- Frequently asked questions
What is help desk software?
Help desk software is a system that brings service requests from different channels into one place, manages them as tickets and makes their handling traceable. Every request becomes a case with a status, an owner and a history.
The value lies less in storing than in ordering. Without such a system, requests end up in personal inboxes, callback notes and chat threads. Nobody reliably knows what is open, how long someone has been waiting, or whether two people are working the same case.
Typical deployments are external customer service, internal IT support, and shared service centres in areas such as HR or facilities.
Help desk or service desk?
The two terms are often used interchangeably but describe different scopes.
| Help desk | Service desk | |
|---|---|---|
| Origin | Older term from IT support | Function defined in ITIL |
| Purpose | Handle incidents and requests | Single point of contact for all services |
| Scope | Focused on incident handling | Also service requests, problem and change management |
| Mindset | Reactive, case-by-case | Process-oriented, service-wide |
| Typical use | Customer service, basic IT support | IT service management under ITIL |
ITIL defines the service desk as the single point of contact between the service provider and its users. The help desk is the older, narrower term and was originally aimed at fixing faults.
For customer service, the distinction is mostly academic. A team answering external customer enquiries does not need an ITIL framework, it needs a system that sorts requests reliably. The difference starts to matter when tender documents call for ITIL alignment.
Which features does help desk software have?
Six features make up the core. Everything else is equipment.
| Feature | What it does |
|---|---|
| Shared inbox | Requests from email, forms, chat and phone arrive in one interface |
| Ticket management | Each request gets a status, a priority and a traceable history |
| Assignment and rules | Tickets are routed automatically by topic, language or workload |
| SLA monitoring | Response and resolution targets are tracked, breaches flagged |
| Knowledge base | Stored answers are available while working a ticket and feed self-service channels |
| Reporting | Volume, handling times, topics and satisfaction become measurable |
The fifth row is the one most often underestimated. Many teams treat the knowledge base as an add-on, but it is what keeps answers consistent and it is the precondition for automating anything later.
Beyond the core, most systems offer macros and canned responses, automation rules, a customer portal, post-resolution satisfaction surveys, and integrations with your shop, CRM, ERP or telephony.
Which types of help desk software exist?
The differences lie in the operating model and the feature set rather than in the basic idea.
| Type | Characteristic | Fits when |
|---|---|---|
| Cloud solution | Hosted by the vendor, billed per seat | You need to be live quickly without tying up IT |
| On-premise | Installed on your own infrastructure | Data must not leave the building |
| Open source | Source-available, self-hosted or managed | Adaptability matters and development capacity exists |
| Shared-inbox tool | Shared mailbox without full ticketing | Volume is low and the team is small |
| ITSM suite | Full ITIL-aligned service management | Internal IT services with problem and change management |
| Sector-specific | Tailored to one industry | Specialist processes must be reflected, e.g. public sector |
Widely used systems include Zendesk, Freshdesk, Salesforce Service Cloud and Jira Service Management, alongside the open-source systems Zammad and OTRS.
One practical point for the selection: the type of system is hard to change later, because tickets, history, roles and automations all have to be migrated. The AI layer on top can be swapped comparatively easily. It is worth treating those as two separate decisions.
When does AI pay off in a help desk?
AI in a help desk pays off when three conditions hold at the same time: meaningful contact volume, a high share of recurring standard enquiries, and documented answers to them.
| Condition | Question to ask | If not met |
|---|---|---|
| Volume | Is there enough traffic for the setup effort to amortise? | Sort out processes and templates first |
| Repetition | Can enquiries be reduced to a manageable set of topics? | Automation reaches too few cases |
| Documentation | Do the answers exist in writing, or only in people’s heads? | Build the knowledge base first |
The third row is the decisive one, and the one most often skipped. AI can only reproduce what is documented. Without that foundation, the main thing a rollout produces is disappointment.
The size of that gap is visible in a Gartner survey of 5,728 customers: 73 per cent use self-service at some point in their service journey, yet only 14 per cent fully resolve their issue there. Among the reasons given, 43 per cent could not find relevant content and 45 per cent said the company did not understand their issue. Both causes sit in the knowledge base, not in the language model.
Where AI should act first
A common misconception is that AI belongs inside the ticketing system. It works there, but only once the ticket exists. Considerably more volume can be intercepted earlier.
- Before the ticket: help page, chat and contact form answer the enquiry before a case is created.
- On arrival: incoming emails are analysed and known questions answered automatically.
- Inside the ticket: agents receive reply suggestions from the knowledge base and accept them with one click.
All three stages should answer from the same source. Maintain content per channel and it will go stale in at least one of them.
Where AI reaches its limits
The limits below apply regardless of vendor. They follow from the architecture rather than from any product’s quality.
- AI reproduces what is documented. Gaps in the knowledge base are not flagged, they are paraphrased around.
- Inside the ticketing system it acts after the fact. The case already exists by then.
- Individual cases stay with people. Goodwill, complaints and negotiation cannot be answered from a database.
- A poor first impression is costly. In a Gartner survey of 3,566 customers, only 27 per cent said they would try a chatbot again after a bad experience.
- The route to a human must stay visible. 87 per cent of customers consider access to a person essential once generative AI is used in service.
- Maintenance stays manual. Prices, deadlines and processes change, and nobody outside your team notices.
Which metrics show the effect?
| Metric | What it tells you |
|---|---|
| First contact resolution | Share of enquiries settled at the first contact |
| Average handling time | How much team time a case consumes |
| First response time | How long customers wait for the first reply |
| SLA attainment | Whether committed targets are being met |
| CSAT | How satisfied customers are with the resolution |
| Ticket deflection rate | How many enquiries are resolved before a ticket is created |
| Zero-result searches | Which content is missing from the knowledge base |
On automated channels, deflection and CSAT belong in the same report. Falling ticket counts alongside falling satisfaction is not a saving, it is displacement into channels nobody measures.
The last row is also the most actionable: every search term returning nothing is a missing entry.
What should decide the selection?
Eight questions worth settling before the decision:
- Which channels come together? Email, phone, chat, forms and messengers should land in one interface.
- How is it billed? Per seat, by volume, or outcome-based per resolved enquiry. The models behave very differently as you grow.
- Is there an integrated knowledge base? And does it also feed your self-service channels?
- Which integrations exist? Shop, CRM, ERP and telephony.
- How much administration does it need? Who creates rules, and does that require IT?
- Can data be exported? Tickets, history and knowledge base in an open format.
- What compliance evidence exists? Processing agreement, processing location, subprocessors, retention periods.
- What are the exit terms? Notice periods and treatment of historical data.
Points six and eight are the ones that get skimmed, and they determine how expensive a future switch becomes.
What do GDPR and the EU AI Act require?
Help desk software processes personal data, often at scale. UK GDPR and EU GDPR therefore require a processing agreement under Article 28, named subprocessors, transparency about where data is processed, and defined retention periods.
Once AI is involved, Article 50 of the EU AI Act has applied since 2 August 2026: people must be able to tell that they are interacting with an AI system, unless that is already obvious from the circumstances. For chatbots the duty applies with no transition period.
For regulated sectors, public bodies and universities, these points are usually shortlisting criteria rather than contract details.
OMQ is a German company based in Berlin and provides a processing agreement under Article 28 GDPR including a named external data protection officer. Personal data is stripped out of free-text fields before the text is stored. Further details for tender responses are available through the contact form.
Extending your help desk with OMQ
OMQ does not replace a ticketing system. Its products act before it and inside it, and all of them draw on the same central knowledge base. An answer is maintained once and served identically everywhere.
| Product | Where it acts | Effect |
|---|---|---|
| Intelligent Help Page | Help page and FAQs | Resolves enquiries before a ticket exists |
| Smart Contact Form | Contact form | Suggests the answer before the message is sent |
| AI Chatbot | Website and messengers | Answers free-text enquiries in real time |
| AI Email Bot | Email inbox | Answers known questions automatically |
| AI Voice Bot | Phone | Handles recurring standard questions by voice |
| Ticket System Assistant | Inside the ticket | Suggests replies to agents, accepted with one click |
| AI Agent | Backend systems | Carries out tasks such as address changes |
The knowledge base is edited in a no-code editor, and existing website content can be imported into it in one step. Entries can be translated with a single click, with automatic translation available in over 30 languages, so the same answer serves an international audience without a separate content set per market.
The Ticket System Assistant plugs into existing help desks including Zendesk, Salesforce, Freshdesk, Zammad and OTRS. For Zendesk there is an app in the Zendesk Marketplace.
We use OMQ on our help and contact page and can thus answer most of our customer requests directly.Meike Schönwandt, Senior Manager Voice of Customer at Tchibo
Key terms at a glance
- Help desk software: A system that consolidates service requests across channels, manages them as tickets and documents their handling.
- Service desk: The ITIL-defined single point of contact between a service provider and its users, broader in scope than a help desk.
- ITSM: IT service management, the end-to-end governance of IT services along defined processes.
- Ticket: The case record under which a service request is tracked, with status, owner and history.
- SLA: An agreement on response and resolution targets, monitored by the system.
- First contact resolution: Share of enquiries settled conclusively at the first contact.
- Ticket deflection rate: Share of enquiries resolved before a ticket is created.
- Intent recognition: The method an AI uses to determine the purpose behind a free-text message rather than matching keywords.
Conclusion
Help desk software creates order: requests arrive in one place, get an owner and become traceable. The six core features are similar across systems, and the real differences sit in the operating model, the administration effort and the integrations.
AI pays off where volume, repetition and documentation come together. Where documentation is missing, the knowledge base is the first job. And it works hardest before the ticket rather than inside it, because an enquiry that never becomes a case also consumes no handling time.
For the assessment, two numbers side by side are enough: the ticket deflection rate and satisfaction. If one rises while the other falls, nothing was solved, it only moved.
Frequently asked questions
What is help desk software?
What is the difference between a help desk and a service desk?
Which types of help desk software exist?
When does AI pay off in a help desk?
Where does AI in a help desk reach its limits?
Which metrics show whether a help desk is working?
What does help desk software cost?
Is help desk software with AI GDPR compliant?
About OMQ
OMQ is AI software for automated customer service. It is operated by OMQ GmbH, a SaaS company founded in Berlin in 2010 by Matthias Meisdrock and Sven Engelmann. All products draw on the same central knowledge base: AI Chatbot, AI Email Bot, AI Voice Bot, AI Agent, Intelligent Help Page, Smart Contact Form and Ticket System Assistant. The software works in over 30 languages, and customers include Deutsche Bahn, Tchibo, Mister Spex and alltours.
About the author: Yasmin Altmann leads marketing at OMQ GmbH and writes in the OMQ lexicon about customer service automation, AI in service and self-service channels.
Current as of 6 October 2026. This article is reviewed at least quarterly.
Sources
- Gartner: Only 14% of Customer Service Issues Are Fully Resolved in Self-Service, survey of 5,728 customers
- Gartner: Only 27% of Customers Would Try a Chatbot Again After a Negative Experience, survey of 3,566 customers
- Gartner: 87% of Customers Say Companies Using GenAI for Customer Service Must Provide Access to a Human Agent
- Zendesk Marketplace: OMQ Assist for Zendesk Support
