RATING METHOD · RATED SEPTEMBER 2026
How we rate IT service desk software: seven criteria, published weights
Last reviewed 27 September 2026 · Service Desk Compare reviewers
What do we score, and how much does each criterion weigh?
Scroll for more
| Criterion | Weight | What it measures |
|---|---|---|
| ITIL practice coverage | 15% | Which ITIL 4 practices the product documents: incident, service request, problem, change enablement, release, service level, knowledge, service configuration and IT asset management. Certifications (PinkVERIFY, ITIL-certified practices) count when the vendor states them. |
| Automation and AI assistance | 18% | Documented AI and automation across the ticket lifecycle: intake, triage, routing, autonomous resolution, closure and knowledge upkeep. Informed by the Lifecycle Automation Census. |
| Time to value | 18% | How quickly a mid-size team can go live: self-serve trial or free tier, published plans, templates and defaults, and the configuration or partner effort vendors and reviewers describe. |
| Configuration management (CMDB) | 12% | CMDB and asset depth: discovery, relationship and service mapping, data health, and which plan includes it. |
| Usability for non-IT requesters | 14% | The experience of the employee raising the ticket: portal, catalog, chat and Teams or Slack intake, virtual agents, portal user limits. |
| Integrations | 10% | Documented connectors, API and MCP support, integration catalogue size where stated, and fit with common IT stacks. |
| Pricing transparency | 13% | Whether a buyer can model cost from public pages: published per-agent prices, minimums, add-on prices. |
The weights reflect a mid-size IT team (roughly 5 to 100 agents). Time to value and AI assistance carry the most weight (18% each) because they decide whether a team this size gets a working desk in weeks rather than quarters. ITIL coverage (15%) and CMDB (12%) matter less than they would for a large enterprise, but still decide the outcome for teams that run formal change and configuration processes. An enterprise buyer should read the per-criterion scores, not the total.
What does each point on the 0 to 10 scale mean?
| Score | Meaning |
|---|---|
| 9 to 10 | Documented in depth on public pages, with plan availability stated |
| 7 to 8.5 | Documented, with gaps or plan gates |
| 5 to 6.5 | Partly documented or dependent on add-ons |
| 3 to 4.5 | Thin documentation or major capability missing |
| 0 to 2.5 | Not documented or not published |
Scores move in half points.
How is the total calculated?
Total = sum of (criterion score × weight) ÷ 100. We calculate it in code from the scores and weights on this page, keep three decimals for sorting, and display one decimal. Tools that display the same total share a rank, shown as '2='.
| Criterion | Score × weight = product |
|---|---|
| ITIL practice coverage | 4 × 15 = 60 |
| Automation and AI assistance | 9 × 18 = 162 |
| Time to value | 8.5 × 18 = 153 |
| Configuration management (CMDB) | 2 × 12 = 24 |
| Usability for non-IT requesters | 8.5 × 14 = 119 |
| Integrations | 7.5 × 10 = 75 |
| Pricing transparency | 8 × 13 = 104 |
| Total | 697 ÷ 100 = 6.97, shown as 7.0 |
Where does the evidence come from?
- Vendor pricing pages, product pages and documentation. Each fact on the site links to the page it came from.
- Where a vendor does not publish a price we write 'Not published'. Where a capability is not on the pages we read we write 'Not documented on public pages reviewed'. Neither means the product lacks it.
- Vendor performance figures (for example SysAid's 'more than 60% of incidents' resolved before they become tickets) are quoted in a 'Vendor states' box and not verified.
- G2 category scores and review themes appear on some pages as context, in a 'What G2 shows' box with a link. They are not inputs to our scores, and we do not show G2 star averages or review counts, which changed between two fetches on the same day.
- Every comparative statement about ServiceNow links to a G2 page or a ServiceNow page.
How does the Lifecycle Automation Census work?
For each vendor we read the pricing page, the main product page and the AI or automation page, and recorded whether they describe automation at six stages: logged, categorised, routed, resolved, closed and reviewed. 'Documented' means the page describes the capability specifically. 'Partial' means it is described in general terms or only for another use case. 'Not documented' means we did not find it. The census informs the AI criterion but is not a score by itself.
Freshservice
01
Logged
Partial02
Categorised
Not documented03
Routed
Documented04
Resolved
Documented05
Closed
Not documented06
Reviewed
DocumentedHalo
01
Logged
Documented02
Categorised
Documented03
Routed
Documented04
Resolved
Documented05
Closed
Documented06
Reviewed
DocumentedJira Service Management
01
Logged
Documented02
Categorised
Partial03
Routed
Partial04
Resolved
Documented05
Closed
Not documented06
Reviewed
DocumentedManageEngine ServiceDesk Plus
01
Logged
Partial02
Categorised
Documented03
Routed
Documented04
Resolved
Partial05
Closed
Not documented06
Reviewed
DocumentedServiceNow ITSM
01
Logged
Documented02
Categorised
Partial03
Routed
Partial04
Resolved
Documented05
Closed
Not documented06
Reviewed
DocumentedSysAid
01
Logged
Partial02
Categorised
Not documented03
Routed
Not documented04
Resolved
Documented05
Closed
Not documented06
Reviewed
DocumentedZendesk for IT (Employee Service)
01
Logged
Documented02
Categorised
Documented03
Routed
Documented04
Resolved
Documented05
Closed
Not documented06
Reviewed
Documentedmonday service
01
Logged
Documented02
Categorised
Documented03
Routed
Documented04
Resolved
Documented05
Closed
Not documented06
Reviewed
DocumentedIvanti Neurons for ITSM
01
Logged
Documented02
Categorised
Documented03
Routed
Documented04
Resolved
Documented05
Closed
Not documented06
Reviewed
DocumentedShow notes
| Vendor | 01 Logged | 02 Categorised | 03 Routed | 04 Resolved | 05 Closed | 06 Reviewed |
|---|---|---|---|---|---|---|
| Freshservice | PartialFreddy AI collects and verifies requester details (general service page) | Not documented | DocumentedFreddy AI routes the issue to the right agent | Documented'Domain-specific AI resolves issues instantly' (vendor wording) | Not documented | DocumentedProblem management on Pro; Freddy AI Insights surfaces emerging issues |
| Halo | DocumentedEnd-user virtual agent triages and logs requests on chat, Teams, Slack and portal | DocumentedAI suggestions for categorisation and prioritisation | DocumentedAI suggestions for team routing and SLA-breach warnings | DocumentedProcess agents auto-resolve known issues | DocumentedA requester's thank-you reply is read as resolution | DocumentedTicket records turned into standardised knowledge |
| Jira Service Management | DocumentedPortal, email, chat and forms on every plan | PartialRovo Agents for AI-powered employee support; no auto-categorisation claim | PartialQueues and workflows; no AI routing claim | DocumentedVirtual service agent (Premium) | Not documented | DocumentedGroup incidents to problems for root cause analysis |
| ManageEngine ServiceDesk Plus | PartialZia virtual support agent; intake channels not detailed on the page read | DocumentedPredictive intelligence for ticket triage | DocumentedPredictive routing and assignment | PartialZia virtual agent and generated replies; no autonomous resolution claim | Not documented | DocumentedProblem management (Enterprise or add-on) |
| ServiceNow ITSM | DocumentedOmnichannel self-service across portal, chat, voice, email, mobile, Teams, Slack | PartialNow Assist AI-guided recommendations | PartialAI agents streamline incident management and request fulfillment | DocumentedAutonomous AI specialists for password resets, access and VPN issues | Not documented | DocumentedProblem management on the same platform and CMDB |
| SysAid | PartialAI aims to resolve incidents before they become tickets | Not documented | Not documented | DocumentedPre-built AI agents handle issues start to finish | Not documented | DocumentedProblem and knowledge management listed |
| Zendesk for IT (Employee Service) | DocumentedAI agents work on any channel | DocumentedDetects intent and tone | DocumentedOmnichannel routing assigns tickets to qualified agents | DocumentedAI agents resolve multi-step workflows | Not documented | DocumentedAI monitoring across 100% of interactions |
| monday service | DocumentedIntake through email, WorkForms and the portal | DocumentedAutomatic categorization and summarization of cross-channel requests | DocumentedSkill-based and load-balanced routing | DocumentedIntake & Triage Agent auto-resolves common requests | Not documented | DocumentedKnowledge Agent flags stale articles and learns from resolved tickets |
| Ivanti Neurons for ITSM | DocumentedVirtual agent, chatbots and self-service portals | DocumentedAutomatically classifies incoming tickets | DocumentedRoutes tickets to the right team | DocumentedAgentic AI detects and remediates issues | Not documented | DocumentedKnowledge articles generated from resolved incidents |
What are the limits of this assessment?
- Public sources only. No hands-on testing, no demos, no trial accounts, no vendor interviews.
- Vendors document different things in different places. A capability that exists but is described only in a customer-only help centre will score lower here than it would after a trial.
- Prices change often and many are shown only for annual billing. Check the linked pricing page before you buy.
- The weights suit mid-size IT teams. Large enterprises with formal change advisory boards and complex CMDBs should weight ITIL and CMDB higher, which would move ServiceNow and Ivanti up.
- Scores reflect September 2026. We re-check every price and headline feature each quarter.
How do corrections work?
Vendors can send corrections with a link to a public source to editors@servicedeskcompare.com. We check and update within one working week and log the change on the page's 'Last reviewed' line.
What do readers ask about the method?
Did you test these service desks?
No. It is drawn from vendor pricing pages, documentation and G2 category pages, last reviewed in September 2026. We did not run trials, demos or vendor interviews.
Why does ServiceNow rank fifth on the full ITSM ranking if it has the most features?
ServiceNow scores 10 on ITIL coverage and CMDB, the highest in the set, but 4 on time to value and 2 on pricing transparency because it publishes no prices and has no self-serve trial. With the full ITSM weights, that places it fifth at 7.2.
Do G2 ratings affect the scores?
No. G2 figures are shown as context with a link to G2. Scores come from vendor documentation and our published rubric.
Why does monday service rank differently on the two rankings?
monday service ties for first of 9 on the business-teams ranking and ties for seventh of 9 on the full ITSM ranking. The same scores apply to every vendor: it scores 9 on AI assistance and 8.5 on time to value and requester experience, but 4 on ITIL coverage and 2 on CMDB, since its public pages do not document problem management, change management or a CMDB. Only the full ITSM ranking scores ITIL and CMDB.