Call Dispositions Explained: How to Tag Every Call Without Relying on Staff to Remember

A call disposition is a short label recording how a call ended: booked, quoted, follow-up needed, information only, spam or wrong number. Dispositions are what turn a call log into a report, and the reliable way to get one on every call is auto-disposition, where AI reads the transcript and applies the label, because manual tagging is skipped whenever staff are busy.

This guide covers what dispositions are for, a starter list you can adopt today, why manual tagging breaks down, how automatic tagging works, and what to do with the labels once you have them.

Why dispositions matter more than call counts

A call tracking system without dispositions can tell you that Google Ads produced 90 calls last month. It cannot tell you how many of those were customers, how many were sales reps, and how many were people who dialed the wrong business.

With dispositions, the same 90 calls become something like this: so many booked, so many quoted and waiting, so many needing a callback, and the rest noise. Now you can calculate booked rate, cost per booked call and how much money is sitting in unanswered follow-ups.

Almost every metric worth reporting depends on this label. The ten we recommend in our guide to call tracking KPIs nearly all start from a disposition, and the conversion definitions in our guide to phone call conversion tracking rely on knowing which calls were qualified and which were booked.

A starter list of disposition codes

Most businesses need fewer codes than they think. This set of six is the one CallFlux assigns automatically, and it covers the decisions that matter.

Disposition	Meaning	What happens next
Booked	The caller committed: appointment, order or dispatched job	Count as a conversion; match to revenue
Quoted	A price was given; no commitment yet	Follow up within a set time
Follow-Up Needed	The caller needs a callback, more information or a decision from someone else	Assign an owner and a deadline
Info Only	A genuine caller with a question that is not a sale: hours, directions, an existing job	No sales action; exclude from lead counts
Spam	Robocalls, solicitations and vendors	Exclude from every metric; consider blocking
Wrong Number	The caller wanted someone else	Exclude; watch for patterns pointing to a listing error

When to add a code

Add a disposition only if you will act differently because of it. Good candidates:

Outside service area. If these are frequent, your ad targeting needs tightening.
Existing customer. Separates service calls from new demand.
Not a service we offer. Shows where your ads or listings are attracting the wrong searches.
Price objection. Useful if you want to study lost quotes.

Poor candidates are codes that describe the caller's mood, duplicate another code, or exist because one manager asked for them once. Every additional option makes each choice slower and less consistent. Keep the list under about ten.

Rules for a list people can use

Make codes mutually exclusive. If a call could reasonably be tagged two ways, staff will split between them and neither count will be right. Write one sentence for each code saying what qualifies.
Order by outcome, not alphabet. Revenue outcomes first, noise last.
Define booked precisely. Is a tentative appointment booked? Is a customer who says "I will call back to confirm" booked? Decide once.
Separate the outcome from the reason. The disposition is what happened. Why it happened belongs in a note or a second field.

Why manual tagging fails

Asking the person who took the call to tag it sounds reasonable. In practice it breaks for predictable reasons.

The phone rings again. The moment a call ends is the moment the next one starts. The tag is postponed and then forgotten.
Calls answered away from the desk. A technician or owner answering on a mobile phone has no screen in front of them.
Default-option bias. When a tag is required to close the screen, people click the first option to make it go away.
Inconsistent judgment. One person tags every price inquiry as quoted; another tags the same call as info only.
Missed calls have no handler. Nobody tags a call nobody answered, yet those are often the most important to follow up.

The outcome is a dataset that is perhaps two-thirds complete and unevenly accurate, with no way of knowing which third is missing. Reports built on it look authoritative and are not.

How auto-disposition works

Automatic disposition replaces the click with a reading of what was actually said.

The call is recorded and transcribed. Speech is converted to text, with each speaker separated.
An AI model reads the transcript. It looks at the substance of the conversation, not just keywords: whether a time was agreed, whether a price was given, whether the caller was selling something.
A disposition is assigned. The call is labeled with the best-fitting outcome from your list.
Downstream actions fire. A booked call can be sent to an ad platform as a conversion; a follow-up needed call can create a task or a notification.

Because the label comes from the transcript, every call gets one, whether it was answered at a desk, on a mobile phone or at two in the morning. The same model can also produce a short written summary of the call, which is covered in our guide to AI call summaries, and a lead score, described in our article on AI lead scoring for phone calls.

What it does well and where to check it

Clear outcomes are the easiest to classify. A robocall, a caller who asked for a different business, and a caller who agreed to Tuesday at ten are all distinct in a transcript.

The harder cases are the same ones people disagree on: a caller who got a price and said they would talk to their spouse, or a long conversation about an existing job that ended with a question about new work. Audio quality matters too; a caller on a bad connection produces a weaker transcript.

The practical safeguard is review, not blind trust. Listen to a sample of calls every week across the categories, and where the label and your judgment disagree, sharpen the definition of that code. Staff should be able to correct a label when it is wrong, and corrections are worth reviewing for patterns.

One compliance note: auto-disposition depends on recording and transcribing calls. Recording laws vary by state, with some requiring all parties to consent, and compliance is the responsibility of the business doing the recording. Our overview of call recording consent laws lists the questions to settle first.

What to do with dispositions once you have them

Report on outcomes

Replace total calls with booked calls and cost per booked call by channel. A channel that looks cheap on cost per call can turn out to be the most expensive once only booked calls are counted.

Feed the ad platforms

Send booked calls back to Google and Meta as conversions so bidding learns from outcomes, not rings. The mechanics are in our guide to importing call revenue into Google Ads.

Automate the follow-up

A disposition is a trigger. Quoted with no booking after a day can prompt a callback. Follow-up needed can notify an owner immediately. Spam can add the number to a block list, as described in our spam call filtering guide. Rules like these are what CallFlux's automation engine is for; see the automation page for how rules are built.

Coach the team

Compare outcomes by the person who answered. If one team member's calls end as quoted far more often than booked, there is a conversation to have, and call recordings to have it with. Our guide to AI call scoring and QA scorecards goes further on structured review.

Clean up your marketing

A cluster of wrong number or not a service we offer calls from one keyword or one listing is a targeting problem you can fix the same day.

A rollout plan

Write the list. Six to ten codes, one sentence each.
Agree on booked. Get the people who answer the phone to sign off on the definition.
Turn on recording, transcription and automatic classification, with the consent notice your state requires.
Run for two weeks without changing anything else. Review twenty calls a week against the labels.
Adjust definitions where labels and judgment disagree.
Switch reports to booked-based metrics, and tell stakeholders in advance that lead counts will fall because noise is being removed.
Add automation rules for quoted and follow-up needed.

Frequently Asked Questions

What is a call disposition?

A call disposition is a short label that records the outcome of a phone call, such as booked, quoted, follow-up needed, information only, spam or wrong number. It is set at the end of the call, either by the person who handled it or automatically by software, and it turns a list of calls into data that can be counted, filtered and reported.

What call disposition codes should a small business use?

Start with six: booked, quoted, follow-up needed, information only, spam and wrong number. That set separates revenue outcomes from noise and is short enough to apply consistently. Add a code only when you will make a different decision because of it, such as outside service area or existing customer, and keep the full list under about ten.

What is auto-disposition?

Auto-disposition is the automatic assignment of a disposition to each call by software, without anyone clicking a button. The system transcribes the call, an AI model reads the transcript, and the call is labeled with the outcome that best fits what was said. It produces a disposition on every call, including the ones staff are too busy to tag.

How accurate is AI call disposition?

Accuracy depends on audio quality, how clearly your categories are defined and how closely they match what is actually said on calls. Clear outcomes such as spam, wrong number and a confirmed booking are generally easier to classify than ambiguous ones such as a caller who says they will think about it. The practical approach is to review a sample of calls each week and correct the definitions where the labels disagree with your judgment.

Why do manual call dispositions fail?

Because tagging is extra work at the moment staff are busiest. After a call ends the phone often rings again, so the tag is skipped, guessed later or set to whatever option is at the top of the list. The result is partial data that looks complete, which is worse than none because reports built on it are trusted.

What can you do with call dispositions once you have them?

Dispositions let you report booked rate and cost per booked call by channel, send booked calls to ad platforms as conversions, trigger follow-up automatically when a call is tagged as quoted or follow-up needed, exclude spam from lead counts, and spot coaching needs when one person's calls are quoted far more often than they are booked.

Get an outcome on every call

CallFlux transcribes calls and uses AI to summarize, score and classify them, so outcomes do not depend on someone remembering to click. See AI call insights and automation, or check the pricing page for which plan includes the auto-disposition engine and custom dispositions.