Skip to content
Tutorials Center
My learning

Topics

Lesson 29 of 37 ยท Reporting and Dashboards

Custom Report Builder and Marketing Attribution

Learn HubSpot's custom report builder, cross-object joins, and revenue attribution models: first touch, last touch, linear, U-shaped and time decay, with runnable Python calculations.

Single-object reports answer simple questions. Real questions cross objects: "Which campaigns produced the deals that closed?" "How long do tickets take for our biggest customers?" For these, HubSpot offers the custom report builder, which can join several objects, and attribution reports, which credit revenue to marketing touches.

HubSpot cannot run here. This lesson explains the feature and uses Python stand-ins to compute attribution so you understand the models. The custom report builder is a Professional-and-above feature, and multi-touch revenue attribution is limited to higher plans; check yours.

Custom report builder

Choose Reporting > Reports > Create custom report > Custom report builder. The steps are:

  1. Select data sources: one or more objects (contacts, companies, deals, tickets, activities, ...). HubSpot joins them using associations, so the relationships must be recorded correctly.
  2. Add fields (properties) from each object; filters apply per object.
  3. Choose a visualization: table, bar, line, pivot, and so on.
  4. Save and add it to a dashboard.

The join logic is important. If a deal has two associated companies, a report that lists deals by company can count the deal twice. Learn to ask: "what does one row represent?" and check totals against a simple report.

Python Join deals to companies and watch for double counting

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

Attribution

Attribution reporting assigns credit for a contact, deal or revenue to the marketing touchpoints along the way (a blog visit, an ad click, an email). HubSpot reports include models such as:

  • First touch: all credit to the first interaction.
  • Last touch: all credit to the final interaction before conversion.
  • Linear: equal credit to every touch.
  • U-shaped: 40% first, 40% the touch that created the contact, 20% shared among the rest (HubSpot's definition for contact creation; check the current documentation for exact percentages).
  • W-shaped, Full path and Time decay variants give extra weight to key moments or recent touches.

Attribution models

First touch Credit the discovery
Last touch Credit the closer
Linear Split evenly
Time decay Recent touches count more

Models tell different stories

1 / 2

Computing the models

The stand-in below distributes a $10,000 deal across one customer's journey under several models. The U-shaped weights here are the simple textbook 40/40/20 version used for teaching.

Python Revenue attribution under four models

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

Rolling it up per channel

With many deals you add the credits by channel to see which channels produce revenue.

Python Channel revenue by model

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

Notice that organic search looks great under first touch, while email looks great under last touch. Neither is "right"; use several models and talk about patterns, not single numbers.

Data quality for attribution

  • Tracking code must be on the site so page views map to contacts.
  • UTM parameters need consistent naming (utm_source=newsletter, not Newsletter and news).
  • Contacts created by import have no touchpoints; they will not appear in channel reports.
  • Offline channels (events, phone) need to be logged as activities or properties.
  • Privacy consent affects tracking: users who decline cookies may be invisible to the report.
Python Normalize UTM sources before reporting

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

Check your understanding

Quick check ยท 3 questions

of 3 answered You scored / 3 ยท perfect!

  1. 1. Why can a cross-object report double count revenue?

    Joins create one row per association, so check what a row represents.
  2. 2. Which model gives all credit to the first interaction?

    First touch credits the earliest known interaction.
  3. 3. What is a good habit when comparing channels?

    Different models answer different questions; consistent patterns are more trustworthy.

Summary

  • The custom report builder joins objects through associations; always sanity-check totals.
  • Attribution credits marketing touches using first, last, linear, U-shaped or time-decay models.
  • Clean UTMs, tracking and consent settings produce reliable channel reports.

Next: Funnels, Conversion Rates and Sales Forecasting