CallRail Implementation Strategy

Overview

CallRail is a call tracking platform that connects inbound phone calls to their marketing source — paid ads, organic search, direct traffic, etc. When implementing CallRail, there is a fundamental architectural decision to make: dynamic number insertion (DNI) versus replacing the main business number with a CallRail number. Each approach has meaningful tradeoffs for attribution completeness and operational simplicity.

This decision came up during Asymmetric's engagement with ^[1], where call attribution was identified as a gap in measuring Google Ads conversion quality.

The Core Decision

Option 1: Dynamic Number Insertion (DNI)

DNI uses JavaScript to swap the phone number displayed on a website based on how the visitor arrived. A visitor from a Google Ad sees one number; a visitor from organic search sees another; a direct visitor sees yet another.

Pros:
- Preserves the client's primary business number for all non-tracked contexts (business cards, directories, signage)
- Granular source-level attribution per session
- No disruption to existing NAP (Name, Address, Phone) consistency for local SEO

Cons:
- Requires JavaScript to load correctly — calls made before the script fires may be missed
- More complex setup; requires careful QA across devices and page types
- Does not capture calls from offline sources or channels where the number appears statically

Option 2: Replace Main Number with a CallRail Number

The business's primary phone number is replaced entirely with a CallRail tracking number, which forwards to the real line. All calls — regardless of source — flow through CallRail.

Pros:
- Comprehensive call capture; no calls fall outside tracking
- Simpler implementation — no JavaScript dependency
- Easier to report total call volume alongside source attribution

Cons:
- The CallRail number becomes the canonical business number, creating dependency on the vendor
- NAP consistency risk: if the number appears in Google Business Profile, directories, or citations, those must all be updated
- If CallRail service lapses, calls may be disrupted

Recommended Decision Framework

For most established businesses with existing citation profiles, DNI is the safer default. For newer businesses or those where comprehensive attribution outweighs citation risk, main number replacement offers simpler, more complete tracking.

Implementation Checklist (Either Approach)

1. Create CallRail account and configure a pool of tracking numbers appropriate to traffic volume
2. Define conversion events — align with what Google Ads counts as a conversion (calls of minimum duration, form fills, etc.)
3. Connect CallRail to Google Ads via the Google Ads integration to import call conversions
4. Connect to Google Analytics (GA4) to see call data alongside session data
5. Set minimum call duration threshold (e.g., 60 seconds) to filter out wrong numbers and short non-intent calls
6. QA across devices — confirm numbers swap correctly on mobile, tablet, and desktop
7. Update Google Business Profile if the tracking number becomes the primary number
8. Notify client of any number changes that affect their printed or offline materials

Open Questions Before Deployment

Before finalizing the approach, confirm the following with the client:

• Is the current phone number published in directories, Google Business Profile, or offline materials?
• What is the approximate split between online-sourced calls and offline/referral calls?
• Is there an existing Google Ads conversion action for calls, or is this net new?
• What call duration should count as a meaningful conversion?

Client Context

During the November 2025 BluePoint ATM review, Asymmetric had already created a CallRail account and was evaluating implementation approach. The team was weighing DNI against main number replacement to ensure comprehensive tracking of leads generated by the Google Ads campaigns (Cashless, Traditional ATM, Reverse ATM, PMAX).

Next step at time of meeting: Asymmetric to send BluePoint a structured list of questions to resolve the implementation decision.

Related: ^[1] · ^[2]