Audit Your Call Tracking Integrity Before You Judge a Lead Source

Before you cut a vendor, prove the phone system told you the truth

Every store has that meeting. Somebody pulls up the lead source report, points at a vendor, and says "we got 41 calls from them and sold two." Then a decision gets made on that number.

The problem is that the number is the output of about six systems stapled together: a tracking number pool, a routing tree, a BDC phone system, a CRM integration, a human typing a source into a deal screen, and a DMS match on the sold side. Any one of those can be quietly broken for months and the report still prints.

I've seen a store swap a tracking number on a third-party listing during a website redesign and lose attribution on every call from that vendor for seven weeks. The report said the vendor collapsed. The vendor was fine. The number on the listing pointed to the main line.

So before you judge a lead source, audit the plumbing. Here's how to do it in an afternoon.

Step 1: Call every tracking number yourself

Not a spot check. All of them. Pull the full list of tracking numbers from your call tracking provider — every pool, every vendor, every campaign, every Google Business Profile, every window sticker and buyback mailer still floating around.

Then sit down with a cell phone and a spreadsheet. For each number, log:

• Does it ring? Dead numbers happen more than you'd think, especially on retired campaigns that someone forgot to unpublish.
• Where does it land? Sales, service, parts, the receptionist, or a voicemail box nobody checks.
• How many rings before pickup or rollover?
• Does the call appear in the tracking dashboard within five minutes, tagged to the right source?

That last one is the whole point. You're not testing whether the phone works. You're testing whether dealership call tracking is actually recording what you think it's recording.

Do the same thing in reverse. Pull up every place your number is published — your website header, your VDP pages, your third-party listings, your Google profile, your Facebook page, your service reminder texts, your plate frames — and check which number is displayed. One mismatch between "number we're tracking" and "number the customer sees" kills an entire lead source in your reporting.

Common finding: the website header shows the tracking number, but the "Contact Us" page and the footer show the main line. Mobile shoppers tap the footer.

Step 2: Map the routing tree out loud

Dealership phone routing is almost always older than anyone currently working at the store. Nobody has drawn it since the last phone vendor change.

Draw it now. On paper is fine. For each entry point, write out what happens:

1. Caller dials the sales tracking number at 11 a.m. Tuesday.
2. Rings the sales hunt group for 4 rings.
3. Rolls to the BDC queue for 20 seconds.
4. Rolls to a voicemail box owned by... who, exactly?

Then run the same trace for: Saturday at 4 p.m., Monday at 7:45 a.m. before the store opens, and 8 p.m. after close. Those three windows are where routing falls apart, because after-hours schedules were set by whoever installed the system and never revisited when hours changed.

Specific things worth checking:

• Holiday schedules. Still set for last year's calendar?
• The orphan voicemail box. Every store has one. Find whose extension it belongs to and whether anyone has the password.
• Simultaneous vs. sequential ring. If you're ringing six desks at once and five reps have their phones face-down, the call rolls before anyone notices.
• Mobile forwarding. If a rep's calls forward to a cell, does the tracking system still capture the recording and duration?

Step 3: Test missed call recovery end to end

A missed call is not a lost lead unless you let it be. But most stores have never tested whether their missed call recovery actually produces a callback.

Run the test. Have someone the store doesn't recognize call in from an unfamiliar number during a known busy window, let it ring to voicemail or abandon, and then do nothing. Just watch.

Measure three things:

• Did the missed call generate a CRM record? Not a voicemail notification — an actual lead or task with an owner.
• How long until someone called back? Timestamp it.
• What did the callback sound like? Pull the recording.

The callback script matters more than the speed. "Hi, this is Dana at the store, I saw a missed call from this number" is weak. Try:

"Hi, this is Dana from the dealership — looks like we missed your call about fifteen minutes ago and I didn't want to leave you hanging. Were you calling about a specific vehicle, or something on the service side?"

Run this test five times across different days and times. If two of the five never produce a callback, you have a 40% hole in your lead flow that no lead source report will ever show you, because an abandoned call that never got logged simply doesn't exist in your data.

Step 4: Follow a single call into the CRM

Now trace the handoff. Place a test call to a known tracking number, have a rep take it and work it like a real up, then go look at the CRM record.

Check, in order:

• Did a lead get created automatically, or did the rep have to create it?
• What source got stamped on it? The vendor, or something generic like "Phone Up" or "Walk-In"?
• Is the source field editable by the rep? If yes, how often does it get changed?
• Is the recording attached to the lead record, or does it live only in the tracking platform?

This is where most phone lead attribution dies. The call tracking system knows the source perfectly. The CRM gets a generic "Inbound Call" and the rep picks a source from a dropdown based on what they remember the customer saying. Two systems, two answers, and the report pulls from the weaker one.

Run a quick count: pull last month's phone leads from your call tracking platform and from your CRM. If tracking says 312 calls over 60 seconds and CRM says 180 phone leads, you have 132 calls that either never became leads or became leads with no source. Find which.

Good CRM source tracking means the source is written by the integration, locked from rep editing, and visible on the deal jacket all the way to delivery.

Step 5: Check the sold side

Attribution breaks twice — once when the lead comes in, once when the deal goes out.

Take last month's sold units. For each one, answer: does the deal have a lead source attached, and does that source match the original first-touch record?

You'll find three failure patterns:

• Blank source. Desk created the deal fresh instead of pulling the existing lead.
• Source overwrite. The customer came in on a third-party lead, then called the main line a week later, and the second touch overwrote the first.
• Duplicate records. Same customer, two leads, two sources, one sold — and the vendor that actually produced the customer isn't the one credited.

Set a rule and write it down: first touch owns the credit, and the deal must be merged into the original lead record before it can be marked sold. Make it a desk checklist item, not a suggestion.

What to do with what you find

Don't try to fix everything. Build a short list with a name and a date:

Then re-run the whole audit in 90 days. Phone systems drift. Websites get redesigned. Vendors change their listing numbers. An audit you ran once is a photograph, not a control.

Once the plumbing is clean, the lead source conversation gets a lot shorter. You're arguing about vendor performance instead of arguing about whether the data is real. And when you can tie recordings to the right source and the right outcome — which is where tools like MoreSignal earn their keep — the next vendor review takes twenty minutes instead of an hour of hedging.