The short version
- We score products on six published criteria, using rules anyone can apply and check.
- The biggest criterion, true cost with AI on, is arithmetic on the vendor's own list prices. Nobody assigns it by hand.
- Every price is stored once, with its source and the date we read it, and every page in every language is built from that record.
- We have not run hands-on tests yet. Scores rate list prices and public documentation, and no page on this site says or implies otherwise.
- The order of our lists is an editorial decision. Scores are shown next to every product so you can compare them yourself.
- There are no affiliate links, paid placements or sponsored rankings here.
What we do, and what we do not do yet
Most "best help desk" lists are written from vendor marketing pages and other lists, then presented as testing. We think that is the central problem with software reviews, so we start by being exact about our own work.
What we do today: desk research. We read vendor pricing pages, billing documentation, legal terms and security pages; published case studies from named customers; and independent reporting. We record what each source says, link to it, turn the list prices into a bill at three team sizes, and score products on what a buyer can verify before signing.
What we do not do yet: hands-on testing. We have not set these products up side by side, so we publish no measurements of our own, such as setup time or AI answer quality, and award no points for them. Words like "tested" and "hands-on" do not appear in our reviews for that reason. When that changes, this page and the changelog will say so, and the results will be published in full.
Where our facts come from
- The vendor's own pricing pages, documentation and legal terms, for that vendor's prices, billing definitions, channels, languages, hosting regions and certifications.
- Case studies that name the customer, for results achieved in production, read with the knowledge that vendors choose which customers to publish.
- Independent reporting and data, such as established press, regulatory filings and analyst research. A study paid for by the vendor is labeled as such.
If we cannot find a source for something, the page says "not documented" rather than guessing. That label describes the public record on the date we checked; it does not mean the feature does not exist.
One product database, so pages cannot contradict each other
Review sites often quote one price in a ranking and a different one in a comparison, because each article was written separately. Here, every price lives in a single product record with its source and the date it was read. Reviews, pricing pages, comparisons, the calculator and all four language editions are generated from that record. Our writers cannot type a price into a sentence; they can only point to the record. When a vendor changes a price, we change it once and every page updates.
Prices are shown in the currency the vendor publishes. Comparisons and the calculator use each vendor's US dollar list price, because every vendor we cover publishes one. Where a vendor also publishes its own euro or sterling price list, pricing pages in those markets show it alongside. We never convert a price ourselves, because a number produced with an exchange rate is not a price anyone can buy at.
How the true-cost figure is calculated
A help desk's advertised seat price is only part of the bill. Most vendors now sell AI in two further pieces: an agent-assist add-on charged per agent, and a customer-facing AI agent charged by usage, such as per resolution or per session. Our true-cost calculator adds the three together: seats on a given plan tier, the agent-assist add-on for every agent, and a number of AI resolutions a month.
Every product is costed at the same three team sizes: 3 agents on the entry tier, and 10 and 50 agents on the mid tier, with about 250 tickets per agent a month and one ticket in five resolved by AI. The assumptions made for each product, such as which plan we treat as its mid tier, are listed under the calculator. Tiers are not equal across vendors: one vendor's mid tier includes features another sells only at the top. The figure tells you what a plan costs, not that two plans do the same job.
How we score
Rubric v1.0 has six criteria. Each product gets a level from 1 to 5 on each criterion. The cost level is computed from the 10-agent scenario; the other five are assigned using the written rules below. The level is multiplied by the criterion's weight and the results are added up to give a score out of 10. Because the rules are public, you can check any score yourself and tell us if we got one wrong.
True cost with AI on — 30% (computed from list prices)
What does a 10-agent team pay per agent per month on the vendor's mid-tier plan, with the agent-assist add-on switched on for everyone and 500 conversations a month resolved by the vendor's AI agent? This level is computed from list prices by the same code as the calculator; nobody assigns it by hand.
| Level | What it means |
|---|---|
| 5/5 | Under $60 per agent per month. |
| 4/5 | From $60 to under $100 per agent per month. |
| 3/5 | From $100 to under $150 per agent per month. |
| 2/5 | From $150 to under $200 per agent per month. |
| 1/5 | $200 or more per agent per month, or the vendor's list prices do not allow the sum. |
Pricing transparency — 20%
Can a buyer work out the full bill before talking to sales?
| Level | What it means |
|---|---|
| 5/5 | Every self-serve plan and AI add-on is priced in public, the billable unit is defined in writing, and minimums, mandatory fees and overage rules are stated. |
| 4/5 | Plan prices and AI prices are public, but the billable unit, the minimums or the overage rules are not fully defined in writing. |
| 3/5 | Plan prices are public, but the AI or usage prices, or a material fee, are not. |
| 2/5 | An entry price is public, but most plans or all AI features are quote-only. |
| 1/5 | No public prices. |
AI, as documented — 15%
What AI does the vendor document, and can you tell how it is sold and what happens to your data?
| Level | What it means |
|---|---|
| 5/5 | Everything in level 4, plus a published statement on whether customer data trains AI models and which AI sub-processors are used. |
| 4/5 | Agent-assist AI and a customer-facing AI agent, both with public prices and a defined usage unit. |
| 3/5 | A customer-facing AI agent is documented, but its price or its billing unit is not public. |
| 2/5 | Writing help or summaries for agents only; no customer-facing AI agent. |
| 1/5 | No AI features documented. |
Access and switching — 10%
How easily can a team start, and how easily can it leave?
| Level | What it means |
|---|---|
| 5/5 | A free plan, monthly billing with no mandatory fees, and documented import tools. |
| 4/5 | Self-serve trial with no mandatory fees, plus either a free plan or a documented tool for importing from other help desks. |
| 3/5 | Self-serve trial with no mandatory fees; no free plan and no documented import tool. |
| 2/5 | Self-serve trial, but paid plans carry a mandatory onboarding fee or a seat minimum above three. |
| 1/5 | A sales call is required to start. |
Channels and languages — 10%
Does it cover the channels your customers use, in the languages your agents work in?
| Level | What it means |
|---|---|
| 5/5 | Email, live chat, a help center, native phone and native messaging, with the agent interface documented in both German and Spanish. |
| 4/5 | Email, live chat and a help center, plus native phone or native messaging (WhatsApp, SMS or social). |
| 3/5 | Email, live chat and a help center; phone and messaging channels only through third-party apps. |
| 2/5 | Email plus live chat or a help center. |
| 1/5 | Email only. |
Data protection and compliance — 15%
Can you keep customer data where you need it, and prove the vendor's controls?
| Level | What it means |
|---|---|
| 5/5 | Everything in level 4, plus a documented HIPAA business associate agreement and a published list of sub-processors. |
| 4/5 | A data processing agreement, SOC 2 or ISO 27001, and EU data hosting available on self-serve plans. |
| 3/5 | A data processing agreement, SOC 2 or ISO 27001, and EU data hosting that is limited to higher plans, a paid add-on or a sales request. |
| 2/5 | A data processing agreement and SOC 2 or ISO 27001, but no documented EU data hosting. |
| 1/5 | No data processing agreement or security certification documented. |
The ranking
| Product | True cost with AI on | Pricing transparency | AI, as documented | Access and switching | Channels and languages | Data protection and compliance | Score |
|---|---|---|---|---|---|---|---|
| Jira Service Management | 5/5 | 4/5 | 5/5 | 4/5 | 3/5 | 4/5 | 8.7 |
| Help Scout | 4/5 | 5/5 | 5/5 | 5/5 | 4/5 | 2/5 | 8.3 |
| Zoho Desk | 5/5 | 3/5 | 3/5 | 5/5 | 4/5 | 4/5 | 8.1 |
| Freshdesk | 3/5 | 4/5 | 5/5 | 3/5 | 5/5 | 5/5 | 8.0 |
| HubSpot Service Hub | 3/5 | 4/5 | 5/5 | 2/5 | 5/5 | 4/5 | 7.5 |
| Gorgias | 4/5 | 4/5 | 3/5 | 4/5 | 4/5 | 3/5 | 7.4 |
| Intercom | 2/5 | 5/5 | 4/5 | 3/5 | 5/5 | 3/5 | 6.9 |
| Front | 3/5 | 3/5 | 5/5 | 3/5 | 4/5 | 3/5 | 6.8 |
| Zendesk | 1/5 | 3/5 | 3/5 | 3/5 | 5/5 | 3/5 | 5.2 |
| Product | with AI on · per agent | True cost with AI on |
|---|---|---|
| Jira Service Management | $51 | 5/5 |
| Help Scout | $83 | 4/5 |
| Zoho Desk | $23 | 5/5 |
| Freshdesk | $109 | 3/5 |
| HubSpot Service Hub | $112 | 3/5 |
| Gorgias | $93 | 4/5 |
| Intercom | $164 | 2/5 |
| Front | $130 | 3/5 |
| Zendesk | $225 | 1/5 |
Scores and the order of our lists
The order in which products appear on our lists and in our ranking table is set by our editors. It usually follows the score, but not always: where our editors place a product higher or lower than its score would put it, the score shown next to it is still the one produced by the rules on this page, unchanged. If you want pure score order, the table of levels above lists every product by score.
Vendor claims versus named-customer figures
Vendors publish impressive numbers about time saved, tickets deflected and return on investment. Wherever we show one, we label who stands behind it:
- Vendor claim: a number from the vendor's marketing, usually an average or a best case.
- Study paid for by the vendor: research by an analyst firm that the vendor commissioned.
- Named customer, published by the vendor: a figure a named company reported in a case study the vendor chose to publish.
- Independent third party: press, filings or research not controlled by the vendor.
None of these figures were measured by us, and none of them earn points in the rubric. They are there so you can see the gap between the brochure and the case study.
What the score does not tell you
The score tells you how affordable, transparent, well documented, accessible, broad and compliant a product is, judged from public information. It does not tell you whether your agents will like working in it, how good its AI answers are on your content, or how responsive its support team is. Those depend on your own tickets and people, and the way to find out is a trial. Every product we list offers one.
The rubric gives cost the largest weight. That is deliberate, because it reflects the question this site exists to answer, but it means capable, expensive platforms score lower here than their feature lists alone would suggest. Our enterprise list reads the same data with that in mind.
Planned: a hands-on setup test
The next version of the rubric will add measured criteria. The protocol we intend to publish before running it:
- A fresh trial account of every product, set up by the same person following a written script: connect an email address, import the same 200 sample tickets, build the same five routing rules and two SLA policies, and publish the same ten help articles.
- Time to each milestone recorded, with dated screenshots.
- The same 50 customer questions put to each vendor's AI agent, scored for correct answers, wrong answers given with confidence, and appropriate handoff.
- All recordings, transcripts and scoring sheets published, so anyone can re-score them.
Keeping pages current
Every page shows three dates: when it was published, when it was last updated, and when its prices and figures were last checked against their sources. We re-verify the product database monthly and whenever a vendor announces a pricing change. Material changes are recorded in the changelog.
Independence and funding
If the publisher sells a product in this category, that product is labeled "Publisher's product" wherever it appears and is scored with the same rules as everything else. We do not use affiliate links, and links to vendors carry no tracking parameters.
AI assistance
Pages are researched and drafted with AI assistance, working from the sources listed at the bottom of each page. Every page is published by the help-desk.reviews team, which is responsible for its accuracy.
Corrections
If a price is stale, a fact is wrong or a score does not follow from the rules above, use the "Report an error" link on any page. We fix confirmed errors in the product database, which corrects every page at once, and we log material corrections in the changelog.
This page was researched and drafted with AI assistance from the sources listed on it. We have not run hands-on tests of these products. Method: How we review