Phone intent scoring ranks contacts by the likelihood that they will answer a cold call. It measures reachability, not buying intent. Use it to set call order, pair it with an account signal and a relevance check, and test it against your own call log before trusting a vendor's connect-rate claims.
Phone Intent Scoring: The Short Answer
- The phrase is vendor vocabulary for answer likelihood. TitanX describes Phone Intent™ as a score of each prospect's likelihood of answering a cold call before a rep dials. The other pages that rank use "intent score" for something else: readiness to buy.
- Two questions, two scores. A buying-intent score asks whether an account is in a buying window. A phone intent score asks whether a person will pick up. A contact who answers is not a buyer, and an account in a buying window may have no reachable contact.
- The public pages do not show how the score is built or tested. They give tiers and connect rates and do not give the inputs, the cut points, or a test against a random call order.
- Use three questions in order. Is the account worth working now, is this person relevant, and will this number reach someone? The Three-Question Call Stack keeps the third answer from replacing the first two.
- Test the score with a call log. Compare connect rates by tier with intervals, then compare the share of connects that reach a relevant person. The two can differ.
What the Pages That Rank for Phone Intent Scoring Say
The results use "intent scoring" for two different things. One set of pages scores buying readiness from behaviour. A TitanX set scores whether a contact will answer.
| Source | What "intent score" means | Numbers and claims it gives | What it leaves open |
|---|---|---|---|
| TitanX, The Role of Intent in Modern Sales Strategies (April 15, 2026) | Ranks contacts by their probability of picking up; the engine uses patterns such as call history, engagement activity and broader buyer signals | A "2 percent versus 20 percent connect rate" contrast with no source or base | Which inputs, how they are weighted, and how the score was tested |
| TitanX, The World's First Precision Dialer is Here | Phone Intent™ scores each prospect's likelihood of answering a cold call before a rep dials | About 20% of any market will answer a cold call, across more than 100 million phone numbers; 3.5 times the meeting rate for single-line dialing in a study of more than 3 million calls across 42 companies; an early customer list in which about 98 of 500 contacts scored as likely to answer | The 3.5 times figure compares dialing modes, not scored against unscored order; the customer example has no control list |
| TitanX, TitanX vs Apollo | Scoring is derived from a dataset built across billions of dials, "not firmographic signals"; contacts fall into P1, P2/P3 and Bad Data tiers | A 3 to 7% industry baseline connect rate and 20 to 30% for TitanX customers; cites Cognism's 2026 analysis of over 200,000 calls for a 2.3% cold-call success rate | The TitanX customer rate has no stated sample; the other TitanX pages describe the inputs differently |
| 6sense, Magic Numbers (April 12, 2023) | Intent Score is a 1 to 100 measure of an account's readiness to buy; Contact Engagement Score is a 1 to 100 score of contacts' engagement with owned content | Score bands map to buying stages: Target below 20, Awareness 20 to 49, Consideration 50 to 69, Decision 70 to 84, Purchase 85 and above | Nothing on whether a person will answer a call |
| Saber, Intent Score (January 18, 2026) | A number that quantifies a prospect or account's likelihood to purchase from behavioural, firmographic and engagement signals | Performance claims attributed to Gartner and a 20 to 30 percent conversion gain, without links | Phone reachability; the source for the figures |
| Factors.ai, Intent Scoring via Website Visitor Identification | Numeric values on prospects or accounts from behavioural signals such as website visits and pricing page views; compares predictive and rule-based models | A discussion of black-box risk in predictive scoring | Phone reachability; outcome figures |
Three points follow. Among these pages, only TitanX uses the phrase for a call-answer score. The TitanX pages describe the inputs differently. None of the six pages shows a result for a scored call order against a random one, which is the comparison a buyer needs.
Three Scores, Three Questions
A single "intent" label hides three separate judgments. Each score answers one question and has its own source and shelf life.
| Question | Score type | Example from the pages | Unit | What it cannot tell you |
|---|---|---|---|---|
| Is this account worth working now? | Buying intent | 6sense Intent Score (1 to 100, mapped to buying stages) | Account | Whether any person will pick up |
| Is this person relevant to the purchase? | Contact relevance and engagement | 6sense Contact Engagement Score (engagement with owned content) | Person | Whether the account is in a window |
| Will this number reach someone? | Answer likelihood | TitanX Phone Intent™ tiers | Phone number or person | Whether the person is a buyer |
The Three-Question Call Stack
The Three-Question Call Stack is a rule for the order of use. Ask the account question first, the person question second and the number question third. The answer score only reorders contacts who have already passed the first two questions. A reachability score applied first fills the call list with people who answer and do not buy.
Percepture's guide to what B2B intent data is defines the first layer, and the how intent data is collected and scored guide covers how account scores are built. For the difference between observed events and vendor-reported behaviour, see buying signals vs intent data, and for why a contact record and an intent signal prove different things see intent data vs contact data.
What an Answer-Likelihood Score Can and Cannot Be
TitanX publishes the shape of its score: tiers P1 for high propensity to answer, P2 and P3 for lower propensity, and Bad Data for wrong or inactive numbers. It says the scoring is built on a dataset across billions of dials. It does not publish the inputs, the weights, the tier cut points, how often a score is refreshed or the match rate to a given list.
Those omissions matter because the TitanX pages describe the inputs differently. The April 2026 post names call history, engagement activity and broader buyer signals, while the Apollo comparison says the scoring is built on a dataset of dials and not on firmographic signals. The statements may describe different parts of one system. Without the input list a buyer cannot tell which parts are first-party call outcomes and which are modelled.
The "about 20% will answer" figure is reported across more than 100 million phone numbers in one post and described through billions of dials in another. The units differ, numbers in one case and dials in the other, so the two statements do not combine into one base. The Cognism 2.3% figure that TitanX cites is a cold-call success rate and not a connect rate, and the two measures are not interchangeable.
Source, Timestamp and Confidence for Each Input
This table covers the inputs a call-order decision draws on. Confidence is this guide's judgment and not a measured accuracy rate.
| Input | Type | Source | Timestamp to keep | Confidence in this guide's view |
|---|---|---|---|---|
| Public buying event (hire, funding, posted role) | Observed | Public record | Event date and link | Strong that it happened; weak on what it means for a purchase |
| Account buying-intent score | Reported or inferred | Intent platform | Score date and stage | Weak to partial; treat as a reason to look |
| Contact relevance (title, role in the purchase) | Reported | Contact record | Date last verified | Partial |
| Number validity | Reported | Data vendor or carrier lookup | Verification date | Partial |
| Past call outcomes for this number | Observed (first party) | Your dialer or CRM log | Call dates | Strong that the calls happened |
| Vendor answer-likelihood tier | Inferred | Dialer vendor | Score date | Weak until tested on your list |
| Seller's call outcome and disposition | Observed (first party) | Your CRM | Call date | Strong |
Past outcomes on your own numbers are the only input in the table that is both first party and about phone behaviour. A vendor tier is an inference until it is tested against that log.
The Call Order Card
The Call Order Card is one row per contact that records the three answers and the next action. Each row names the signal, the source and the date, so a reader can check them.
| Account signal | Person relevant | Answer likelihood | Next action |
|---|---|---|---|
| Dated event or intent stage | Yes | High | Call in the next block; open the record to the signal first |
| Dated event or intent stage | Yes | Low | Use another channel for the first touch, or find a second relevant contact; verify the number |
| Dated event or intent stage | No | High | Do not call on reachability alone; research the right contact at the account |
| None | Yes | High | Put in a nurture or routine queue; an answered call does not create a buying reason |
| None | Yes | Low | Suppress from the call list |
| Any | Any | Bad data tier | Verify or replace the number before dialing |
The card has a seller step and a return step. A seller accepts or returns a flagged account with a reason, and a call that connects is logged with a disposition. The signal-to-touch latency guide covers how fast a flagged account should reach a seller. Percepture's guide on using intent data to identify sales qualified leads covers the step from signal to qualified lead, and its B2B intent data services page describes the service side. For ranking vendors that supply the account layer, see the best intent data providers comparison.
Testing a Phone Intent Score on Your Own List
A vendor connect rate is measured on the vendor's customers. Your list, your numbers and your reps differ, so the score needs a test.
- Score the list and freeze the tier. Record each contact's tier before any call is made.
- Randomise call order within each tier. If reps call the top tier first, time of day and week confound the result.
- Define a connect. A connect is a live conversation with a person. Count voicemails and gatekeeper-only pickups separately.
- Count by tier. Record dials, distinct contacts reached and connects.
- Add an interval. Report the 95 percent Wilson interval beside each connect rate.
- Check what the connects were. Count connects that reached the relevant person for your offer. This is the figure the score does not claim to predict.
An Invented Worked Example
Halden Route Software is an invented company, and every number below is invented for this guide. It scored 600 contacts and called each up to three times over two weeks in random order within each tier.
| Tier | Contacts | Connects | Connect rate | 95% Wilson interval |
|---|---|---|---|---|
| P1 | 210 | 63 | 30.0% | 24.2% to 36.5% |
| P2 and P3 | 240 | 33 | 13.8% | 10.0% to 18.7% |
| Bad Data | 150 | 6 | 4.0% | 1.8% to 8.5% |
| All contacts | 600 | 102 | 17.0% | 14.2% to 20.2% |
The intervals do not overlap, so in this example the tiers separate on connect rate. The team then checks who answered.
| Tier | Connects | Reached the relevant person | Share | 95% Wilson interval |
|---|---|---|---|---|
| P1 | 63 | 21 | 33.3% | 22.9% to 45.6% |
| P2 and P3 | 33 | 9 | 27.3% | 15.1% to 44.2% |
The intervals overlap. The score ranks who answers, and the data do not show that the P1 answers are more relevant. A team that read the first table as proof of better leads would be reading a difference the second table does not support. The decision for Halden is to use the tier for call order, to keep the relevance and account checks on the Call Order Card, and to pull more records into the second table before concluding anything.
Where This Approach Falls Short
- The score ranks reachability and does not create demand. A person who answers may have no buying reason.
- Vendor figures are the vendors' own. The connect rates, the 20% answer share and the 3.5 times meeting rate are reported by TitanX, with the bases given above.
- A tier fitted to one list may not transfer. Verify on your own numbers before changing rep behaviour.
- Calling rules are separate from scoring. The FTC's guidance on the Telemarketing Sales Rule says most business-to-business calls are exempt but not calls that solicit individual employees to buy for their own use, and the FCC ruled in February 2024 that AI-generated voices fall under the TCPA's artificial or prerecorded voice rules. Whether a given call is allowed is a question for counsel and not for a score.
- The page set is thin. Only one vendor uses this phrase for a call-answer score, so the public evidence base is one vendor's own material.
Fact, Inference and Our View
| Statement | Type | Basis |
|---|---|---|
| TitanX describes Phone Intent™ as scoring each prospect's likelihood of answering a cold call, with P1, P2/P3 and Bad Data tiers | Fact (vendor page) | TitanX pages |
| 6sense's Intent Score is a 1 to 100 measure of an account's readiness to buy, mapped to five buying stages | Fact (vendor page) | 6sense blog, April 12, 2023 |
| Most business-to-business calls are exempt from the TSR, except calls that solicit individual employees for personal purchases | Fact (regulator guidance) | FTC |
| AI-generated voices fall within the TCPA's artificial or prerecorded voice restrictions | Fact (regulator ruling) | FCC 24-17, February 8, 2024 |
| TitanX's pages describe the score's inputs differently | Fact (about the pages) | TitanX pages |
| A reachability score applied before an account question fills a list with people who answer and do not buy | Inference | Follows from the score measuring answer likelihood |
| A vendor tier is an inference until tested on your own call log | Our view | This guide's framework |
| The Three-Question Call Stack and the Call Order Card make the score usable | Our view | This guide's framework |
Frequently Asked Questions
What is phone intent scoring?
Phone intent scoring ranks contacts by how likely they are to answer a cold call. TitanX uses the term for its Phone Intent™ tiers, which sort contacts into high propensity to answer, lower propensity and bad data. It measures reachability. It does not say that a contact wants to buy.
Is phone intent scoring the same as buy-intent data?
No. Buy-intent scores such as 6sense's Intent Score estimate whether an account is ready to buy, and phone intent scoring estimates whether a person will answer a call. A contact can score high on one and low on the other, so use the account question first and the answer question second.
How do you test whether a phone intent score works?
Freeze each contact's tier before calling, randomise call order within each tier, define a connect as a live conversation and compute the connect rate by tier with a confidence interval. Then count how many connects reached the person relevant to your offer, because answer likelihood and relevance are different measures.
Does an answered call mean the lead is qualified?
No. An answered call shows that a person picked up. Qualification needs a fit judgment and a buying reason, and a seller decides whether the contact is accepted or returned, with a reason recorded.
Sources
- TitanX, The Role of Intent in Modern Sales Strategies (April 15, 2026), The World's First Precision Dialer is Here and TitanX vs Apollo: Why Phone Intent Beats Database Size.
- 6sense, Magic Numbers: Breaking Down 6sense's Intent and Engagement Scores (April 12, 2023).
- Saber, Intent Score: Definition, Examples and Use Cases (January 18, 2026); Factors.ai, Intent Scoring via Website Visitor Identification.
- Federal Trade Commission, Complying with the Telemarketing Sales Rule; Federal Communications Commission, Declaratory Ruling FCC 24-17 (February 8, 2024).
- Wilson score interval, the standard binomial confidence interval used for the connect-rate ranges above (95 percent, z = 1.96).
- Percepture, B2B intent data services, What Is B2B Intent Data and How to Use Intent Data to Identify Sales Qualified Leads.
- Lead Seeker, best intent data providers, how intent data is collected and scored, buying signals vs intent data, intent data vs contact data, reducing time between buying signal and sales touchpoint, how Lead Seeker works and Lead Compass. Vendor pages are an October 2026 snapshot.
About the Author
Bob Generale is President of Percepture. He works across SEO, AI search, digital PR, sales intelligence and AI-powered revenue systems, with a focus on connecting visibility, buyer intent and sales action.
Next Steps
Add three columns to your call list: the account signal with its date, a relevance check and the answer-likelihood tier if you have one. Run the connect-rate test by tier before changing call order. Read how Lead Seeker works to see how a signal, a verified person and a source link sit on one record, and use Lead Compass to choose the accounts to test. Then start free with five verified contacts to compare a dated, source-linked record with a contact already on your call list.