Map Pack Call And Interaction Attribution

Local conversions mostly happen off the website: a tap on the call button, a direction request, a message from the profile. None of that reaches site analytics, so a business winning locally reports flat traffic and the work looks ineffective. The obvious remedy, putting a call tracking number on the profile, creates a NAP conflict when done carelessly. Most of this is instrumentation set up before anyone asks the reporting question.

The skill file

---
name: map-pack-call-attribution
description: Use when local visibility has improved but nobody can say whether it produced calls, direction requests or walk-ins, and the reporting stops at map pack positions.
---

# Map Pack Call And Interaction Attribution

Local conversions mostly happen off the website: a tap on the call button, a direction request, a message from the profile. None of that reaches site analytics, so a business winning locally reports flat traffic and the work looks ineffective. The obvious remedy, putting a call tracking number on the profile, creates a NAP conflict when done carelessly. Most of this is instrumentation set up before anyone asks the reporting question.

## What you need first

- Profile performance data exported monthly: calls, direction requests, website clicks, messages, and the search type split
- Website analytics with UTM parameters already on the profile website link
- A call tracking arrangement that keeps the verified main number in place, if calls need attributing
- Grid rank history covering the same months, so visibility and outcome can be compared

## Method

1. Tag the profile website link with UTM parameters before anything else, because untagged profile traffic lands in organic or direct and cannot be separated retrospectively.
2. Use dynamic number insertion on the website and keep the verified number as the primary on the profile, adding any tracking number as an additional number. Replacing the primary number splits the identity that every citation depends on.
3. Export profile interaction data every month rather than reading the in-product chart, since the retention window is short and the history simply disappears.
4. Separate discovery searches from direct searches in that data. Growth in direct searches is brand demand created somewhere else, while growth in discovery searches is what local work is meant to produce.
5. Line the interaction counts up against the grid history week by week. Position gains with no interaction change usually mean the gains landed in low-demand parts of the area.
6. Attach a value to each interaction type using the business own close rates for calls and for visits, and label it as a modelled figure rather than measured revenue.
7. Report calls, direction requests and website sessions as three separate numbers. Summing them into a single local actions total hides which one moved and is the first figure an executive stops believing.

## What this produces

A monthly local performance record with tagged website sessions, interaction counts split by search type, and a stated model for turning interactions into value.

## Where this goes wrong

- Swapping the verified phone number for a tracking number on the profile and on citations at different times, which breaks NAP consistency for months
- Relying on the in-product charts, whose short retention makes any year-on-year comparison impossible later
- Counting direction requests as new customers, when a large share are repeat customers and staff
- Attributing all call growth to local work in a period when paid or offline advertising also ran

---

From the QuQi skill library - https://www.quqi.io/skills/map-pack-call-attribution

What you need first

• Profile performance data exported monthly: calls, direction requests, website clicks, messages, and the search type split
• Website analytics with UTM parameters already on the profile website link
• A call tracking arrangement that keeps the verified main number in place, if calls need attributing
• Grid rank history covering the same months, so visibility and outcome can be compared

Method

1. 01 Tag the profile website link with UTM parameters before anything else, because untagged profile traffic lands in organic or direct and cannot be separated retrospectively.
2. 02 Use dynamic number insertion on the website and keep the verified number as the primary on the profile, adding any tracking number as an additional number. Replacing the primary number splits the identity that every citation depends on.
3. 03 Export profile interaction data every month rather than reading the in-product chart, since the retention window is short and the history simply disappears.
4. 04 Separate discovery searches from direct searches in that data. Growth in direct searches is brand demand created somewhere else, while growth in discovery searches is what local work is meant to produce.
5. 05 Line the interaction counts up against the grid history week by week. Position gains with no interaction change usually mean the gains landed in low-demand parts of the area.
6. 06 Attach a value to each interaction type using the business own close rates for calls and for visits, and label it as a modelled figure rather than measured revenue.
7. 07 Report calls, direction requests and website sessions as three separate numbers. Summing them into a single local actions total hides which one moved and is the first figure an executive stops believing.

What this produces

A monthly local performance record with tagged website sessions, interaction counts split by search type, and a stated model for turning interactions into value.

Where this goes wrong

• Swapping the verified phone number for a tracking number on the profile and on citations at different times, which breaks NAP consistency for months
• Relying on the in-product charts, whose short retention makes any year-on-year comparison impossible later
• Counting direction requests as new customers, when a large share are repeat customers and staff
• Attributing all call growth to local work in a period when paid or offline advertising also ran

Use this skill in your own AI

The download is a plain markdown file with the name and trigger in its frontmatter. Where an assistant supports skills it can load itself, that frontmatter is what it reads to decide this one applies.

Questions about this skill

More in Local SEO