Skip to main content

How do I match tracked phone calls to ServiceTitan jobs?

ServiceTitan Published August 23, 2026
Short Answer

Match on normalized phone number first, then narrow with a time window around the call, then confirm against the customer or location record. Expect a meaningful share to fail: people call from a number that is not on file, offices call back from a main line, and households share numbers. Report matched and unmatched separately and never spread the unmatched portion across sources to make the totals look complete.

The matching keys, in order of reliability

  • Normalized calling number to customer or location phone. Strip formatting and country prefixes on both sides before comparing. Most failed matches are formatting, not reality.
  • Call time to booking creation time. A booking created within a short window after a call is very likely the same event. Widen the window and precision collapses.
  • Agent identity. If the CSR who took the call is the user who created the booking, confidence rises sharply.
  • Service address confirmation. Useful as a tiebreaker when one phone number maps to several customers.

Why matches legitimately fail

A homeowner calls from a cell phone while the account carries a landline. A commercial account calls from a main switchboard shared by six properties. A spouse calls from their own number. A customer calls, hangs up, and the office calls back on an untracked line, so the booking is created with no inbound call attached at all.

None of these are integration bugs. They are how phones work. A setup that claims to match everything is either guessing or overwriting evidence.

Confidence tiers beat a yes-or-no match

Rather than a binary matched flag, assign tiers: exact number plus tight time window, number match with a looser window, probabilistic match on address or agent, and unmatched. Report on the top tier by default and let anyone who wants to inspect the weaker tiers do so.

This structure also lets you improve the pipeline honestly. When the number-normalization rule improves, you can show exactly how many records moved from a lower tier to a higher one, rather than announcing that attribution got better. Making that match rate visible is a basic requirement of call intelligence that anyone should ask about before trusting a source report.

What the join unlocks

Once calls and jobs are linked, the questions change. You can measure booking rate by advertising source rather than in aggregate. You can see which sources produce calls that book but produce small jobs. You can find the calls that were bookable and did not book, and route them to a follow-up list the same day.

That last one is usually the fastest return in the entire project, because it converts a reporting exercise into recovered work. It is also a good illustration of what marketing intelligence means in practice: not a nicer chart, but a list somebody acts on before the customer calls a competitor.

A note on recordings

Call recording and its use are governed by rules that vary by state, and requirements can differ depending on where the caller is. Treat consent language, retention and access controls as decisions to confirm with your own counsel rather than settings to copy from another company.

Topics: call tracking · matching · attribution · phone data

Have a version of this question about your own business?

The useful answer usually depends on which systems you run and how they're connected. That's a conversation, not a blog post.

Related Answers

People who read this also asked

Browse the Answer Hub →

AI is easy to access. Making it useful is hard.

Bluefrog makes AI useful by integrating it with the way your business actually works — your software, your calls, your customers, your marketing and your revenue.

Technology development since 1997 · AI integration platforms since 2001