Skip to main content
RealHumanTests

AI and conversational IVR

IVR testing with real human callers

Real people call the AI IVR or conversational IVR your customers already reach, and use it end to end the way they would. They report where it routed them, what it misheard, and whether they would have hung up before reaching the right place.

Where testers reach it

  • Phone line

What you get

  • Every transcript or recording
  • Scores on the 8 criterion rubric
  • A one-page verdict
  • No fix to sell, ever

Most IVR testing checks that the call flow is wired correctly: press 2, reach billing, hear the right prompt. That kind of testing is necessary, and it is well served by automated tools. What it does not tell you is how a caller experiences the menu when they do not know which option they need.

Conversational IVR and AI IVR systems replace the keypad with open questions such as "Tell me why you are calling." That is friendlier when it works and more confusing when it does not. A caller who describes the problem in their own words can easily land in the wrong queue, get asked to repeat themselves, or loop back to the start.

Our IVR testing services use human callers with real reasons for calling. They describe problems the way customers do, change their minds, press zero, and say "operator" in several ways. Each call is scored on the fixed rubric and ends with a plain note on where the caller ended up and whether it was the right place.

We test the IVR you already have. We do not design, build or tune call flows.

What the panel pushes on

Where testers push hardest

The panel spends most of its time where this kind of AI tends to fail.

  • Routing by intent

    Testers describe their reason for calling in their own words and record which queue or answer they reached, and whether it was the right one.

  • Speech recognition

    Account numbers, dates of birth, names and addresses spoken naturally, with accents and at different speeds.

  • Menu depth and loops

    Testers count the steps to reach the right place and flag every loop back to the main menu.

  • Escape routes

    Pressing zero, saying "agent" or staying silent. Testers check whether a caller who needs a person can reach one.

  • Transfers and context

    When the call reaches a queue or a person, testers note whether they had to repeat everything they already said.

Sample scenarios

What a tester might try

Real scenarios, played by real people in character.

  • Confused first-timer

    Has a problem that fits two menu options and describes it in a long sentence instead of a keyword.

  • Asks for a human

    Presses zero repeatedly, then says "operator", then "talk to someone" to find which escape route works.

  • Accent or non-native speaker

    Reads out an account number with a regional accent and corrects one digit halfway through.

  • Angry customer

    Calls about a service outage, says so in the first sentence and gets impatient at each extra question.

  • Edge case

    Calls about an account that is not in their name, on behalf of an elderly relative.

You tell us your own concerns and we build the panel around them. These are starting points, not a fixed script.

The rubric

Scored on the same eight criteria as every test

Highlighted rows are where this kind of AI most often loses points. Every session is scored 1 to 5 on all eight, with a written reason.

The eight criterion RealHumanTests rubric, each scored 1 to 5 with a reason
#CriterionThe question the tester answersScore 1 to 5
01Understood meFocusDid it understand what I actually asked, including when I phrased it badly?1 to 5, with a reason
02Got it rightWas every fact, price and policy it gave me correct, with nothing made up?1 to 5, with a reason
03Got it doneDid I leave with my problem solved or my task completed?1 to 5, with a reason
04EffortFocusHow many turns, repeats and rephrasings did it take?1 to 5, with a reason
05HandoffFocusWhen I needed a person, could I reach one without a fight?1 to 5, with a reason
06ToneDid it stay patient and respectful when I was angry, confused or slow?1 to 5, with a reason
07Stayed in boundsDid it avoid promises, discounts or advice it had no authority to give?1 to 5, with a reason
08Would I stayFocusWould I have hung up, given up, or gone to a competitor?1 to 5, with a reason

Why people, not only software

What automated checks miss here

Automated IVR testing is excellent at checking that every path connects and every prompt plays. It follows the menu the way it was designed. Human callers do not, and the gap between the designed path and the path a real caller takes is where most IVR frustration lives.

Containment rates and average handle time can improve while callers are quietly giving up. A human tester tells you the exact prompt where they would have hung up.

What you receive

Transcripts, scores and a plain verdict

This is the report format, shown blank, with no invented scores.

The one-page verdict

Sample format, no real scores
System
Your support chatbot, website widget
Panel
12 sessions, 8 personas
Window
About ten days
Rubric score layout, blank in this sample
CriterionMean of 5
Understood meblank in this sample
Got it rightblank in this sample
Got it doneblank in this sample
Effortblank in this sample
Handoffblank in this sample
Toneblank in this sample
Stayed in boundsblank in this sample
Would I stayblank in this sample

Verdict, one of three

  • Keep it
  • Fix these settings
  • Reconsider it
  • Every transcript or recording

    Each session in full, with the tester's notes pinned to the exact moment something went right or wrong.

  • A score per criterion, per session

    Eight criteria scored 1 to 5 by the person who lived the conversation, each with a written reason.

  • What failed, named plainly

    When the verdict is fix these settings, it lists the behaviors testers saw fail, so whoever maintains your AI knows where to look.

  • No fix to sell

    We diagnose only. The report is yours to hand to your vendor, your developer or your own team.

Questions

Common questions about this test

Quick answers before you request a test.

Do you replace automated IVR testing tools?

No. Automated tools check that call flows are wired correctly at scale. We add the human side: whether a real caller understood the prompts, reached the right place, and would have stayed on the line.

Can you test a traditional keypad IVR as well as a conversational one?

Yes. Any IVR a customer calls can be tested. The rubric is the same, and testers note where touch tone and speech options behave differently.

How many calls does an IVR test include?

The Audit starts from $349 for a 12-session panel. Larger IVRs with many branches usually suit a Custom Panel, which we quote after a short scoping form.

See your AI the way your customers do

Tell us what your AI does and what worries you. We build a panel of real people around it and hand you every transcript, a score per criterion and a plain verdict. We never sell the fix.