Why PhoneTap Exists

In 2020, I started running my HVAC company remotely.

In 2022, we switched to ServiceTitan.

Like many contractors, I quickly learned something:

ServiceTitan is powerful. But if you do not fully understand it, it will not help you the way you expect.

There are many fields, settings, and reports working behind the scenes. If the data going in is wrong, the reports coming out will also be wrong.

When you manage a company from a distance, you must trust your numbers. I could not rely on walking into the office to see what was happening. I had to rely on data.

So I made it my job to understand:

• Every important field
• Every dashboard
• Every call metric
• Every report

Before I could make good decisions, I had to make sure the data was correct.

Bad Data Is Worse Than No Data

Many owners say they want to make data-driven decisions.

That is good. But only if the data is accurate.

Bad data is worse than guessing.

Why?

Because bad data gives you confidence in the wrong direction.

This problem is well-documented in operations research. A Harvard Business Review analysis estimated that bad data costs U.S. businesses $3 trillion per year — primarily by leading managers to make confident, well-reasoned decisions based on inputs that don't reflect reality. In a call center context, a manager who sees a 78% booking rate and acts on it has no way to know whether that rate is real or an artifact of misclassification.

You think you are making smart decisions. But you are not.

At first, I checked the basics:

• Payments were marked complete
• Jobs were closed correctly
• Recalls were tied to jobs
• Commissions made sense

Then I looked at call center metrics.

That is where the problem showed up.

The Call Numbers Were Not Right

I started reviewing:

• Total calls
• Booked calls
• Unbooked calls
• Booking percentage
• Abandoned calls
• Call reasons

I listened to real call recordings. So did my co-founders.

What we heard did not match the numbers.

Different CSRs had very different results. But when we listened to the calls, the classifications often did not make sense.

Even with call playback, we could only review one call at a time.

So we built a simple tool that:

• Pulled the recording
• Transcribed the call
• Turned it into readable text

Now we could read calls quickly.

Together, we reviewed thousands of transcripts.

For each call, we wrote down:

• What it should have been classified as
• What it was classified as
• If it was unbooked, why

Across multiple companies handling thousands of calls, we found accuracy was only about 50 to 60 percent.

To put that in dollar terms: if a well-run residential HVAC company takes 400 calls a month and the true booking opportunity rate is 65%, roughly 260 of those calls represent real revenue. If 40–50% are misclassified, up to 130 missed opportunities per month are hidden from reporting. At an average new-customer value of $1,500–$2,000, that misclassification could be masking $150,000–$260,000 in monthly lost revenue opportunity — without anyone realizing it.

That means almost half the calls were labeled wrong.

Some companies even had managers who listened and reclassified calls. Those were still inconsistent.

The issue was not effort. It was nuance. Calls are complex. People get tired. Standards drift over time.

Call classification is a perfect job for AI.

Why AI Makes a Difference

For the last few years, my co-founders and I have spent several hours a day working with AI and refining call analysis.

The goal was simple:

• Was this call a real opportunity?
• Was it booked?
• If not, why?

We built PhoneTap to answer those questions clearly.

PhoneTap:

• Analyzes every call
• Determines what really happened
• Identifies if it was a true booking opportunity
• Explains why it was not booked
• Syncs corrected data back into ServiceTitan

When your call data is accurate, your reports become useful.

AI-based call analysis typically achieves 85–95% classification accuracy on structured call categories — compared to the 50–60% human accuracy we observed. McKinsey's 2023 State of AI report found that AI adoption in customer operations reduced error rates by 30–50%, with the largest gains coming from classification tasks where human consistency degrades over time and shift. Call classification is exactly that kind of task.

You can coach better. You can train better. You can manage better.

Understanding Lost Revenue

We knew an unbooked call is not just a number.

It is potential revenue.

But how much?

Some new customers spend $100. Some spend $1,000. Some spend $20,000.

The residential HVAC market that Grand View Research valued at $117 billion in 2023 is built almost entirely on repeat relationships — maintenance visits, repairs, and eventually replacement. A new customer who stays with a contractor for 10 years typically generates $8,000–$15,000 in total lifetime revenue. An unbooked call doesn't just cost the first appointment. It costs the entire relationship.

And many keep spending over time.

So we built a Lost Revenue Dashboard.

PhoneTap looks at your real ServiceTitan history to calculate:

• What a new customer is worth in 30 days
• In 60 days
• In 6 months
• In 1 year

We remove extreme outliers so the numbers stay realistic.

Now, when you see an unbooked call, you also see an estimated dollar amount tied to it.

It is no longer just a missed call. It is a clear financial impact.

Tracking Recovered Revenue

We also track what happens next.

If a call was unbooked, did the customer later:

• Call back?
• Respond to an outbound call?
• Book through text?
• Submit a web form?

PhoneTap checks for that.

You can see:

• Revenue that was at risk
• Revenue that was recovered
• The real reasons customers did not book

This gives you the full picture.

Outbound recovery matters more than most contractors realize. Bain & Company research found that increasing customer retention by just 5% increases profits by 25–95%. In a call center context, every unbooked caller who is successfully recovered — by callback, outbound text, or web form — represents a prevented churn event. Knowing which calls were recovered, and by which channel, is what turns a cost center into a measurable revenue operation.

The Real Why

PhoneTap exists because contractors deserve accurate numbers.

You should not have to guess.

You should not have to rely on flawed reports.

You should not make decisions based on bad classifications.

When your data is accurate, everything changes:

• Better coaching
• Better hiring decisions
• Better marketing tracking
• Better forecasting

Accuracy comes first.

Then insight.

Then action.

The payoff from accurate data is measurable. Harvard Business Review's analysis of data quality in business operations found that employees in organizations with high data quality spend 50% less time on non-productive work — resolving data discrepancies, second-guessing reports, and manually verifying numbers. In a call center, that time goes back into coaching, training, and booking calls.

That is why PhoneTap exists.