Shopify
Stripe
Slack
Notion
Google Workspace
Confluence
How Do You Analyze a Year of Zendesk Tickets with AI? (2026)
A Macha Study reads up to 20,000 Zendesk tickets in one run and returns a row per ticket with the columns you define, such as sentiment and contact reason. Set a 12-month date range, test the columns on a small sample, then run the full year.
Key takeaways
- Macha's Studies feature analyzes up to 20,000 Zendesk tickets in a single run, returning one row per ticket with columns like sentiment, contact reason and resolution status.
- Before the full run, a Study tests the columns and guidance on a sample of 10 to 50 tickets, and the sample is charged only for its size.
- Study columns can be boolean, single select, multi select, number, short text or long text, and 4 to 6 focused columns work better than 15 fuzzy ones.
- Every Study run is a frozen snapshot you can download as CSV or push into a Knowledge Source, so a quarterly re-run shows whether fixes moved the numbers.
- A Study run charges only for tickets processed successfully, while errored and cancelled records cost nothing, and Macha's single plan starts at $299/mo for 750 tickets.
To analyze a year of Zendesk tickets with AI, run a Macha Study: set the date range to the last 12 months, define 4 to 6 columns such as sentiment and contact_reason, test on a sample of 10 to 50 tickets, then run it on up to 20,000 tickets. You get back one structured row per ticket (sentiment, the real reason they contacted you, whether they were resolved) that you can sort, count and export as CSV. This is the analysis playbook for Studies. (If you want to turn tickets into agent knowledge instead, see past tickets as a knowledge base.)
| Setting | What the Studies docs say |
|---|---|
| Source | Zendesk Tickets (more sources listed as coming) |
| Records per run | Up to 20,000 |
| Column types | Boolean, single select, multi select, number, short text, long text |
| Test run | A sample of 10 to 50 records before the full run |
| Output | A frozen result table, CSV download, or push to a Knowledge Source |
| Charge | Only for records processed successfully; errored and cancelled records cost nothing |
What can a year of tickets tell you?
Run AI across the whole set and you can answer questions that are normally guesswork:
- Why do people actually contact us? The true contact-reason breakdown, which is almost always different from your existing tags, because tags are applied inconsistently (or not at all).
- What's making customers unhappy? Sentiment plus the specific friction driving it.
- What's trending? Spikes in a topic after a release, a pricing change, or an outage.
- Where are the revenue signals? Refund requests, cancellation risk and upsell openings, hidden across thousands of tickets.
- What's missing from our docs? The questions your help center couldn't answer, which is your content roadmap.
How do you set up a ticket analysis Study?
The wizard is the same one you'd use for a knowledge-base Study. The difference is what you extract.
1. Source. Start a New Study, choose Zendesk Tickets, and pick the Zendesk connector to read from.
2. Scope the records. Set a date range. For a real analysis, widen it to 6 to 12 months (a single run is capped at 20,000 tickets, so a busier desk may need a narrower window or two runs). Add a search filter only if you want to focus (for example, one product line); for a broad analysis, leave it open. Then pick what the AI reads per ticket. The Full comment thread gives the richest signal for sentiment and reasons.
3. Extract: define your analysis columns. This is where an analysis Study differs from a knowledge-base one. Instead of question/answer, you define insight columns:
| Column | Type | What it surfaces |
|---|---|---|
sentiment | Single select | positive / neutral / negative |
contact_reason | Single select | the real reason they wrote in |
frustration_driver | Short text | what specifically annoyed them |
was_resolved | Boolean | did the ticket actually get solved |
refund_requested | Boolean | revenue signal |
doc_gap | Boolean | could this have been self-served? |
Add a line of guidance (for example, "Judge sentiment from the customer's messages, not the agent's. Pick the single best contact reason.") and pick a model. Classification like this rarely needs your most capable model.
4. Test, then run. Run a test on a sample of 10 to 50 tickets first to make sure the AI judges sentiment and reasons the way you would, and tweak the guidance if not. The test sample is charged too, but only for its size. Then run the full study. It processes in the background with live progress, and you're charged only for tickets it processes successfully.
What do you get back, and what do you do with it?
The result is a frozen table: one row per ticket, your columns filled. The run keeps the configuration it started with, even if you edit the Study later. From there:
- Spot the patterns. Sort and count: which contact reason dominates? What share is negative sentiment, and on which topics?
- Quantify the gut feelings. "We get a lot of billing complaints" becomes a count of negative-sentiment billing tickets and the reason behind most of them.
- Build your roadmap. The
doc_gapandfrustration_drivercolumns tell product and content teams what to fix and write. - Feed it forward. Download the CSV for a spreadsheet or BI tool, push the results into a Knowledge Source so agents benefit, or re-run quarterly to track whether your fixes moved the numbers.
Because each run is a snapshot, you can compare this quarter's analysis to last quarter's and see the trend.
Which ticket analyses are worth running first?
- Contact-reason audit: just
contact_reason+was_resolved, across the year. The single most clarifying report most teams never run. - Release post-mortem: scope to the two weeks after a launch; extract
topic+sentimentto see what broke. - Churn signals:
cancellation_risk(boolean) +reason, to find at-risk accounts hiding in support. - Self-service opportunity:
doc_gap+topic, to rank what to add to your help center.
What does a Study cost, and what are the best practices?
- Plan: Macha sells one plan with every feature on every tier, priced by ticket volume from $299/mo for 750 tickets. Studies also run on the free trial.
- Charge: a run is charged per record it processes successfully; errored and cancelled records cost nothing.
- Test before the full run: get the columns and guidance judging the way you would on a sample.
- Match the model to the job: analysis extraction rarely needs a premium one, and a faster model gets you through 20,000 tickets sooner.
- Keep columns focused: 4 to 6 well-chosen columns beat 15 fuzzy ones.
- Re-run on a cadence so you can watch trends, not just snapshots.
Frequently asked questions
Can AI really analyze my support tickets accurately? Yes, for well-defined extractions like sentiment, category, and yes/no flags, especially when you test the columns on a sample first and refine the guidance.
How many tickets can one analysis cover? Up to 20,000 per run; widen the date range to cover your year, or split a busier year into two runs.
What's the difference between this and the knowledge-base use case? Same feature, different columns. Knowledge base extracts question/answer; analysis extracts insights like sentiment and contact reason.
How much does it cost? A run is charged only for the records it processes successfully, and the test sample is charged only for its size. Studies is part of Macha's single plan, which starts at $299/mo for 750 tickets, and it runs on the free trial too.
Can I export or reuse the results? Yes. The run is a stored, frozen snapshot you can review, download as CSV, or push into a Knowledge Source for your agents.
Which sources can a Study read? Zendesk Tickets today. The docs list more sources as coming.
Why run this analysis at all?
You don't have to guess what your customers struggle with, because the answer is already in your tickets. Studies reads up to 20,000 of them in one run and hands you a structured table of sentiment, reasons, and signals. Scope the year, define a handful of insight columns, test, and run, and a pile of tickets becomes a report your product, content, and support teams can act on.
Find out what's really in your tickets: start with $50 of free usage, no credit card required, and talk to the Macha team about your first analysis Study. Start free.
