What result should you attribute first?

Choose one result that your team can confirm. For an online shop, use a completed order. For a service business, you might start with a qualified enquiry, meaning a request that fits your service and buying requirements.

Keep enquiries and sales separate. HubSpot distinguishes lead attribution from reporting on purchases or revenue. A lead is a potential customer, rather than a completed sale. Its attribution guide explains the distinction.

Agree on a small shared vocabulary

• Conversion: the outcome you are studying, such as an order.
• Touchpoint: an interaction along the journey, such as an advertising click.
• Attribution model: the rule or calculation that distributes credit among those interactions.

Google calls an important recorded action a key event. Check the selected key event before interpreting an attribution report. These terms follow Google's explanation of attribution and models.

Hypothetical example: a small furniture retailer receives an enquiry after someone clicks a social advertisement, returns through search and follows an email link. Ask the sales team whether the enquiry became a confirmed order. Decide which outcome this review will cover.

Sketch that sequence before opening more reports. The customer journey mapping guide provides a useful next step for organising the exercise.

How do attribution model types distribute credit?

Choose a model comparison around your question. First touch recognises the opening interaction; last touch recognises the final interaction. Multi touch attribution spreads credit across multiple interactions. HubSpot describes these approaches.

First touch vs last touch attribution

The allocation rules in this table come from HubSpot's model descriptions. Treat them as general concepts, rather than a list of options available in every product.

In our hypothetical furniture journey, first touch credits social advertising and last touch credits email. Use that contrast to ask better questions about the journey. Avoid declaring a winning channel before examining the underlying enquiry and sale.

Where does data driven attribution fit?

Google's data driven attribution uses account data to allocate credit. It evaluates journeys with and without the selected outcome, and the model is specific to the advertiser and key event. Google explains how this calculation works.

My recommendation is to document the model beside every exported table. When comparing results, hold the outcome and date range steady. Ask whether a different allocation rule would change your proposed spending decision.

Where is the Google Analytics attribution report?

Open Advertising → Attribution, then choose Attribution models or Attribution paths. These are the report locations in Google's current instructions.

Check the available models

Google lists three: Data-driven attribution, Paid and organic last click, and Google paid channels last click. First click, linear, time decay and position-based models are no longer available in these reports. The official model list confirms this.

Paid and organic last click ignores direct visits and credits the last eligible channel. Google paid channels last click prioritises the last Google Ads click, falling back to paid and organic last click when none exists. An entirely direct journey is an exception to the exclusion of direct visits. See Google's rules and examples.

Review settings before changing them

1. Select the correct Analytics property, meaning the reporting space for your site or app.
2. Open Admin → Data display → Events → Attribution settings.
3. Review the reporting model and channels eligible for credit.
4. Write down the existing choices before making a change.

Google requires the Marketer role or higher to select attribution settings. The menu path and permission requirement are documented in its setup guidance.

For your first review, choose one key event and inspect its paths. Write down the steps you recognise and the ones you need someone to explain. Keep unfamiliar labels on a short investigation list.

What should you check before trusting the numbers?

Trace a recent order yourself. Ask someone to show the order record, the selected outcome and its appearance in reporting. Start with a case your team understands before discussing the whole account.

Build a simple reconciliation sheet

I recommend these columns: internal order reference, order status, recorded source, recorded outcome, missing information and person responsible for checking it. Keep customer contact details out of the shared review unless they are needed.

Hypothetical example: a home cleaning company records a website enquiry, then confirms the booking by telephone. During your review, you find that the report covers enquiries only. Label it accordingly and check confirmed bookings separately.

Use a consistent naming convention for campaigns. For practical link labelling, work through the UTM parameters guide. UTM parameters are campaign labels added to links; agree on names before distributing links to the team.

• Keep unknown sources marked as unknown.
• Check whether repeated records refer to the same order.
• Separate cancelled orders from confirmed business results.
• Record customer answers to “How did you hear about us?” separately.

Use those answers as investigation notes. Ask a colleague to explain any mismatch between the customer's account and the recorded path. Avoid silently replacing one with the other.

For service businesses, also review what happens after the enquiry. The sales follow-up guide can help you organise that next part of the work.

How should attribution influence ad spend allocation?

Use the report to propose a test. I would not move the entire budget because one channel receives the largest share of credit. Put spending, confirmed orders and the chosen allocation model on the same review page.

Write the decision before changing the budget

Use this template: “We will change this campaign's spending, watch this business outcome, and review the decision after a period that allows enquiries to become orders.” Set an acceptable spending limit and name the person who can stop the test.

Hypothetical example: an online stationery shop sees email receiving the final touch on several order paths. Before cutting social advertising, ask where those customers first entered the recorded journey. Then choose a limited spending test with a clear review date.

Keep the question narrow. Do you want more confirmed orders, a lower acquisition expense, or better enquiries? Choose one primary goal and record other changes, such as a new offer, alongside it.

When you are ready to formalise spending controls, use the ad budget management guide. Agree on the decision rule before automating any adjustment.

Keep marketing channel credit in perspective

Do not present attributed credit as a precise forecast of sales that would disappear if you stopped a channel. That is my recommended limit for using the report in a budget meeting. Ask what additional test would make the decision more convincing.

If people reach the shop but do not complete an order, investigate the buying experience as well. The conversion rate optimisation guide is a practical companion for that investigation.

What should your team do this week?

Finish the week with one clear decision and a short list of unresolved questions. You do not need to explain every customer journey in your first review. Use this working sequence.

1. Define success. Write the exact outcome being reviewed and ask sales to confirm the definition.
2. Inspect examples. Trace a few recent records and mark missing steps.
3. Open attribution reports. Keep the outcome and period consistent while reviewing models and paths.
4. Discuss one difference. Pick a channel whose role deserves investigation.
5. Assign the next action. Fix measurement, continue collecting evidence or run a limited spending test.

End the meeting with a useful note

Write what you observed, what you think it means and what you will test. Keep those three statements separate. Include the model, period, owner and next review date so the next meeting starts with the same question.

For a wider sequence of business experiments, continue with the small business growth marketing guide. Choose the next experiment from your unresolved questions.