Short answer
First-touch attribution helps examine discovery, while last-touch examines the interaction close to conversion. Compare models on the same contact set with stable totals and criteria; allocating credit does not establish causality or create additional sales.
Key verification: Reconcile total sales before interpreting differences between channels.
Sources and limitationsChoose the attribution model based on the decision you need to make and keep it stable during the comparison. First contact helps examine discovery; last contact looks at the interaction closest to conversion. None alone proves which channel caused the sale.
The same sale can appear in different channels
Fictional example: a person arrives through a campaign, returns through organic search and ends up contacting after a direct visit. The first contact assigns credit to the initial campaign; the last contact, to the final visit. Income has not changed: the distribution rule has changed.
If a previous report used another rule, do not interpret the difference as a loss or business growth without reviewing the history on a comparable basis.
What WhatConverts offers
Your documentation places the model selection on Elite plans and describes the person's first contact, last contact, last non-direct, last paid, and last contact, even after converting. Warn that changing the model can alter source and medium in leads and reports. Official attribution models .
Do not confuse this configuration with the model of a conversion action in Google Ads or with that of another system. Two platforms can apply different windows, identities, and rules even though the labels look similar.
Table to formulate the decision
| Business question | Model to explore | Interpretive risk |
|---|---|---|
| Which channels discover new contacts? | First contact | May leave subsequent interactions without credit |
| What interaction appears before contacting? | Last Contact | May overrepresent direct returns |
| What was the last known channel before direct return? | Last non-direct | Depends on observable history |
| Which paid campaign did you participate in most recently? | Last paid | Does not represent the entire organic contribution |
This is an editorial interpretation of questions, not a universal investment recommendation. Define beforehand which events make up the tour and what information is missing due to consent, devices or external systems.
Protocol to compare models
- Select a closed period and a fixed list of eligible contacts.
- Save model, available window, filters and amount used.
- Calculates the current distribution and keeps the totals as a control.
- Examines the same set with an alternative rule, without mixing other changes.
- List contacts whose channel changes and review their routes.
- Document what decision changes and why, before adopting the new rule.
If the interface modifies the global report when the model changes, it agrees to the test with the people who use it and restores the expected configuration. Do not alter a client report without recording the date and reason.
What you should not add
If a sale of 500 euros appears under first contact in one report and under last in another, it is still a sale of 500 euros. Do not add both credits as if they were additional income. Maintain the business identifier and verify that the observed sales total is consistent before analyzing channels.
To prove that a conversation ended in a closed transaction, use sales call attribution . To decide how much temporal information is missing, review campaign maturation by cohorts .
Before paying for more models
WhatConverts may fit if you need to compare tours and models with the right plan. First ensure capture and relationship with contacts. More modeling on incomplete data does not eliminate uncertainty or justify attributing causality to a campaign.
Sources and limitations
Documentary review: . Content type: Documentation-based guide.
Sources describe terms and capabilities stated by their owners. Proposed protocols and fictional examples do not establish product tests performed by CallsIQ.
- Official attribution modelswww.whatconverts.com
Check current terms
Consider these options if they solve the problem described. Confirm features, limits and availability in your country.
WhatConvertsAffiliate linkThe labelled commercial links may earn CallsIQ a commission or referral reward. Our commercial policy.
Official sources, explained calculations and clearly labelled examples. Read about our methodology and use of AI in writing.