Client reporting · Agencies and marketing leaders
First-Touch vs. Last-Touch Attribution for Client Reports
First-touch attribution gives the earliest eligible recorded touch all the credit for a conversion. Last-touch attribution gives that credit to the latest eligible recorded touch before the conversion. Compare them on the same website, dates and payment scope to understand which sources appear at the beginning and end of recorded buying paths.
What Changes Between First-Touch and Last-Touch Attribution?
The allocation changes. The purchase does not. First-touch highlights the start of the eligible recorded path; last-touch highlights its end. Neither rule gives a complete explanation of why the customer bought, and neither can assign credit to an interaction that was never recorded or linked.
| Rule | Question It Helps You Ask | Blind Spot |
|---|---|---|
| First-Touch | Where did this recorded path begin? | Later visits and closing interactions receive no credit. |
| Last-Touch | Which eligible recorded visit came before purchase? | Earlier discovery and consideration receive no credit. |
Use the wider marketing attribution models guide when you need to compare these rules with Linear. This article focuses on a narrower agency task: showing two views of the same revenue, then agreeing which one leads the recurring client report.
How Does One Sale Look Under Both Rules?
Suppose a customer arrives through your newsletter, returns through organic search, then comes back through Google Ads and makes a $120 purchase. Assume all three touches are eligible, recorded and linked to that sale. First-touch credits the newsletter with $120. Last-touch credits Google Ads with $120. Organic search receives zero under both single-touch rules.
Zero credit does not mean the middle visit was useless. It means these particular rules do not allocate any value to it. If you need to inspect its role in the sequence, read the recorded path. Do not turn a zero-credit row into a claim that the channel has no business value.
The explorer deliberately keeps one sale constant. Try the one-touch path: both rules give the same source the full amount. Try no eligible touches: the amount stays unassigned instead of being attached to a convenient source. A manual allocation spreadsheet can illustrate the same principle, provided you keep the record set and definitions consistent.
What Must Stay Fixed Before Comparing Client Reports?
Fix the reporting scope before changing the model. Otherwise, a different date range, website or payment definition can look like a model effect. Write down the baseline so another person can reproduce the comparison next month.
- Choose the client website and an agreed reporting period. Keep the same filters during both views.
- Confirm the currency and revenue definition. A collected payment, a refund-adjusted amount and a booked deal are different things.
- Confirm the outcome being credited. A lead-created report is not automatically comparable with a purchase-revenue report.
- Check material missing or unmatched records before making a channel claim. Note collection changes that affected the period.
HubSpot’s attribution definitions separate contact creation, deal creation and revenue reporting. That is a useful reminder when a client sends a CRM export: match the outcome before debating which model is right. A model label alone does not make two tools comparable.
Use the reporting date guide if you need an agreed period or comparison in Attriboo. A date comparison examines different periods; the model comparison in this workflow examines different credit rules over the same selected records.
How Do You Compare the Same Period in Attriboo?
Open Dashboard, find Where Is My Money Coming From? and select View all. Choose Sources. Keep the website, dates and report filters fixed, then select First-Touch. Record the source-level Net revenue and Order credit you want to discuss.
Where to find itDashboard → Where Is My Money Coming From? → View all → Sources → First-Touch
The overview includes the website, date range, model and source table they affect. The selected model controls are enlarged below; these are two views of the same report scope.
Select the overview to open the full screenshot.
Enlarged detail of the outlined area
- Read the discovery-side view
First-Touch allocates to the earliest eligible recorded touch. That may be later than the customer’s first encounter with your business.
Select Last-Touch without changing the other controls. Compare the same source names, not just the first row: sorting can move a source when its credit changes. Look for sources that gain closing-side credit and those whose discovery-side credit becomes less visible.
Where to find itDashboard → Where Is My Money Coming From? → View all → Sources → Last-Touch
The overview includes the website, date range, model and source table they affect. The selected model controls are enlarged below; these are two views of the same report scope.
Select the overview to open the full screenshot.
Enlarged detail of the outlined area
- Read the closing-side view
Keep dates fixed, select Last-Touch and compare source credit below. A changed allocation does not create another payment.
In these example views, Google Ads changes from $599 under First-Touch to $997 under Last-Touch. Perplexity changes from $499 to $0. The source allocations still sum to $2,840 in each view. That movement is the reporting question: where does credit move when the rule changes? It is not evidence of extra sales.
The attribution model reference explains the product’s allocation rules. Attriboo’s First-Touch and Last-Touch controls use eligible recorded touches, rather than a universal rule shared by every analytics platform. Check the exact report definition before copying a comparison into a client deck.
Read customer journeys when you need the underlying recorded sequence. Journeys uses its own visit and purchase-linking definitions, so its linked purchase value is not a substitute for the source revenue total under the chosen model. Use the path to explain a sequence, and use the model report to explain credit.
Why Is the First Recorded Touch Not Always the First Real Touch?
A report starts with the evidence available to it. A recommendation in a private message, an earlier untracked visit or activity on another device may not be part of the eligible path. First-touch means first within those records, not the first moment the person heard your company’s name.
Last-touch has the same limitation at the other end. If a late return visit is not recorded or cannot be linked, the latest eligible touch can be earlier than the actual final interaction. More confident wording does not repair missing history. Check the collection and matching problem before rewriting a source label.
For instance, a client might remember a customer first arriving from a partner, while your report begins with search. Record that as additional evidence with its source, rather than silently replacing the tracked path. A customer survey and a tracked referral answer related but different questions.
Google Analytics documents a specific last-click rule that generally excludes Direct unless the path is entirely Direct. It also notes that its first-click model is no longer available in current attribution reports. Do not assume an old GA tutorial or the words “last touch” describe the same handling as your other tools.
Which Model Should Lead Your Client Report?
Choose the lead view to match the reporting question and keep it consistent. If the review concerns the start of recorded buying paths, first-touch can lead. If it concerns the recorded visit before purchase, last-touch can lead. Show the alternative view when it materially changes the story.
My recommendation is to agree the convention before presenting a performance win. Do not switch the headline model each month to favor the channel your agency manages. State the model beside the figure, keep the other view available and explain substantial differences with the relevant recorded paths.
| Client Question | Useful Lead View | Follow-Up |
|---|---|---|
| Where do recorded buying paths start? | First-Touch | Inspect the earliest eligible visits and missing-history limits. |
| What source appears before purchase? | Last-Touch | Inspect the closing sequence and checkout handoff. |
| Should we increase this channel’s budget? | Neither rule is sufficient alone | Review costs, margins and suitable causal evidence. |
For the budget question, separate attribution from incrementality. Moving credit between sources does not measure how many extra purchases a campaign caused. Use a suitable experiment when the decision depends on that causal effect.
How Do You Explain the Difference Without Confusing a Client?
Start with the unchanged sale, then explain the changed rule. A useful example is: “This $120 purchase is credited to the newsletter under First-Touch and Google Ads under Last-Touch. The first view describes the recorded introduction; the second describes the eligible visit before purchase. These are two allocations of one sale.”
For a period-level report, name the website, dates, model and revenue definition before the source totals. Add one explanation of a meaningful change and one next action. If the total also changes, investigate report scope, matching, filters, currency and rounding before attributing that difference to the model.
The guide to showing SEO revenue to clients turns those definitions into a business discussion. Bring the model choice into that discussion early, especially when search appears at a different stage from the work your agency performed.
Common questions
Is first-touch or last-touch attribution more accurate?
Neither is universally more accurate. Each applies a different allocation rule to the available records. Choose the view that fits the reporting question, inspect missing history and use the alternative view to check how sensitive the story is to the rule.
Do first-touch and last-touch revenue totals add together?
No. They are alternative allocations of the same eligible revenue. In the $120 example, each view allocates $120; combining them would count the same sale twice. Compare the allocations rather than adding the views.
Why can a source receive zero under both models?
It can sit in the middle of the eligible recorded path. First-touch credits only the beginning, and last-touch credits only the end. Inspect the route or another appropriate allocation model before concluding that a middle touch had no role.
What happens if there is only one recorded touch?
If that touch is eligible and linked to the sale, both rules give it the full credit. Agreement between the models describes the available path; it does not prove that every earlier influence was captured.
Is last-touch the same as last non-direct click?
Not automatically. Last non-direct click applies an additional rule about ignoring Direct traffic. Platforms can differ in eligible interactions and Direct handling. Read the exact tool’s model definition before comparing its numbers with another report.
Should an agency report both models every month?
Keep one agreed lead model for consistency. Include the second view when it explains a material difference or helps answer the client’s question. Repeating two large tables without an explanation is less useful than showing one difference and its next action.
Ten years in SEO, building products and helping clients connect marketing work to business results. Founder of Attriboo.
Get to know Jay · Follow on LinkedIn
