Quick answer
A linear attribution model is a rules-based multi-touch model that gives every touchpoint on a conversion path the same share of credit: 1 divided by the number of touchpoints, so four touches get 25% each and three get a third each. It shows who took part, not who should get the budget.
Formula: credit per touchpoint = 1 ÷ number of touchpoints
What is a linear attribution model?
A linear attribution model is a rules-based attribution model that gives every touchpoint on a conversion path the same share of credit: 1 ÷ n, where n is the number of touchpoints in that path. Three touches means a third each; four means 25% each. Position, recency and channel do not change the share.
That one rule is both the appeal and the limit. Linear refuses to play favourites, so it answers "who took part?" cleanly. It cannot answer "who should get the next dollar?", because it scores a decisive pricing-page visit and a passing display impression identically.
Also called equal-credit or even-split attribution, it is one rule in the multi-touch attribution family.
Formula and a worked example
Credit per touchpoint = 1 ÷ number of touchpoints in the path. Add the fractions per channel; the totals must equal the number of conversions — that is your arithmetic check.
Sample data: six conversions, C1–C6, across four channels; each path converts once.
| Path | Touchpoints in order | n | Credit per touch | Last click |
|---|---|---|---|---|
| C1 | Paid social → Organic search → Email | 3 | 1/3 | |
| C2 | Organic search → Email | 2 | 1/2 | |
| C3 | 1 | 1 | ||
| C4 | Paid social → Referral → Organic search → Email | 4 | 1/4 | |
| C5 | Referral → Paid social | 2 | 1/2 | Paid social |
| C6 | Organic search → Paid social → Referral | 3 | 1/3 | Referral |
Twelfths keep the fractions exact (1/3 = 4/12, 1/4 = 3/12, 1/2 = 6/12):
| Channel | Linear credit | Linear (decimal) | Last-click conversions |
|---|---|---|---|
| Paid social | 1/3 + 1/4 + 1/2 + 1/3 = 17/12 | 1.42 | 1 |
| Organic search | 1/3 + 1/2 + 1/4 + 1/3 = 17/12 | 1.42 | 0 |
| 1/3 + 1/2 + 1 + 1/4 = 25/12 | 2.08 | 4 | |
| Referral | 1/4 + 1/2 + 1/3 = 13/12 | 1.08 | 1 |
| Total | 72/12 | 6.00 | 6 |
The check passes: 17 + 17 + 25 + 13 = 72 twelfths = 6 conversions, one per path.
Now read the two credit columns against each other. Email closes four of the six paths, so last click hands it 4 while linear hands it 2.08 — not a penalty, just an acknowledgement that email rarely acted alone. Organic search closes nothing: 0 under last click, 17/12 ≈ 1.42 under linear, tied with Paid social. It is an assist-only channel, invisible in a last-click report. That is the case for linear: participation becomes legible. It is not a case for moving budget — 1.42 says nothing about whether those organic visits changed anyone's mind.
Linear vs position-based vs time-decay
Same path, different rules: C4 (Paid social → Referral → Organic search → Email), worth $120, touches 21, 14, 7 and 0 days before conversion. Standard rules: position-based puts 40% on the first and last touch and splits 20% across the middle; time decay weights recent touches exponentially with a one-week half-life (Google Ads API enum).
| Touchpoint (days before conversion) | Linear | Position-based | Time-decay (7-day half-life) |
|---|---|---|---|
| Paid social (−21) | 25% = $30 | 40% = $48 | 1/15 = 6.7% = $8 |
| Referral (−14) | 25% = $30 | 10% = $12 | 2/15 = 13.3% = $16 |
| Organic search (−7) | 25% = $30 | 10% = $12 | 4/15 = 26.7% = $32 |
| Email (0) | 25% = $30 | 40% = $48 | 8/15 = 53.3% = $64 |
| Total | $120 | $120 | $120 |
Recomputable: weight = 2^(−days ÷ 7) gives 1/8, 1/4, 1/2 and 1, summing to 15/8. Divide each weight by 15/8 to get 1/15, 2/15, 4/15 and 8/15; times $120, that is $8, $16, $32 and $64. This is the textbook exponential rule on sample data, not one tool's recorded output.
Edge rules to know before you argue about models:
- One touchpoint: every rule gives it all the credit; the three models agree.
- Two touchpoints: linear gives 50/50, and position-based also gives 50/50 because there is no middle to fund.
- Three or more: the models diverge, and the gap widens with path length.
If most paths are one or two sessions, changing models barely moves the numbers; check path length first.
When to use linear attribution (and when not to)
Decide by the question you are asking, not by habit:
- If you are auditing channel participation — which channels appear on converting paths at all — use linear. Every participant gets a non-zero share; nothing hides behind the closing channel.
- If the project is new, with no conversion history to train a data-driven model, use linear as the starting map; it needs only tagged sessions.
- If you have to explain credit to finance or channel owners, use linear. Equal shares are easy to defend in a meeting; no hidden weights to argue about.
- If you are about to raise or cut a budget on this number alone, do not. Linear cannot separate a decisive touch from a passenger; read it beside last click or position-based, which show who closes.
- If paths are long or repetitive — 20 touches means 5% each, and a retargeting channel can sit on every path — do not read linear credit as influence. Frequency inflates it.
The point of this page: linear does not play favourites, so use it to see who participated — and read a closing-weighted model before deciding who gets the money.
Linear attribution in GA4 and Google Ads
Cannot find linear in your Google stack? That is expected, not a settings problem:
- 6 April 2023: Google announces first click, linear, time decay and position-based are going away in Google Ads and GA4; data-driven, last click and external attribution are unaffected. The note reports fewer than 3% of Google Ads web conversions were attributed with those four rules (Google Data, February–March 2023).
- June 2023: conversion actions that do not already use these models can no longer select them.
- Mid-July 2023 (update of 21 June 2023): selectability is removed for all conversions, including switching existing ones.
- September 2023: remaining conversion actions switch to data-driven, with last click available instead; the rules also leave reporting, including the Model comparison report.
- November 2023: GA4 removes the four models. Attribution reports keep data-driven, paid and organic last click, and Google paid channels last click.
Guides that still walk you through picking Linear in GA4's model comparison are out of date. The rule survives outside Google's interface: the Google Ads API still defines a linear value splitting credit equally across clicks, used by external attribution uploads.
Reading linear attribution in TapCub
Back to the six-path example: Organic search shows 0 in the last-click column and a budget cut is on the table. TapCub puts the models side by side, so the review does not require picking one first.
- Set the conversion and the touchpoint. In All-platform analytics › Behavior › Behavior analysis, choose attribution and pick the conversion event as the goal. Credit goes to the sessions before the conversion, including the converting session; the lookback window defaults to 30 days (1–90). The touchpoint dimension defaults to channel, falling back to referrer channel when empty, and can be switched to referrer channel, UTM source, UTM campaign, referrer domain or landing page.
- Select all five models. First touch, last touch, linear, U-shaped (40-20-40) and time decay each become a column. Linear splits a conversion equally across its sessions in the window; U-shaped gives one touch all the credit, two touches half each, and 40/20/40 beyond that; time decay halves a session's weight for every half-life between its start and the conversion (default 7 days, 1–90) — the same 2^(−days ÷ half-life) curve as the $120 example above.
- Find the assist-only rows. Zero in the last-touch column plus a positive linear number means a channel took part but never closed: Organic search here, 0 against about 1.42. That is the row to protect from a last-click-only cut.
- Decide with two columns. Use linear for participation and channel-mix audits, and last touch or U-shaped for closing credit. Each row also shows touchpoint conversion = paths ÷ touchpoint sessions; up to 50,000 conversions are scored per query. See compare attribution models.
Boundaries: session-level credit, not a mobile measurement partner — no click-to-install matching, no ad buying or bidding, and no budget changes; TapCub measures. Bots are counted separately, so crawlers do not inflate the session counts.
Sources
- Google Ads API reference, AttributionModel enum (v17 path, now served as the current version): linear equal credit; position 40/20/40; time decay one-week half-life. Checked 2026-11-15.
- Google Analytics Help, Get started with attribution: four rules-based models unavailable as of November 2023, three remain. Checked 2026-11-15.
- Google Ads Help, First click, linear, time decay, and position-based attribution models are going away (6 April 2023): June and September 2023 steps, reporting removal, under-3% figure. Checked 2026-11-15.
- Google Ads announcements, Timing update (21 June 2023): selectability removed from mid-July 2023. Checked 2026-11-15.
Was this explanation clear?
We use your answers to fix and expand this page.
Couldn’t send. Please try again in a minute.
Thanks for the feedback. We’ll use it to improve this page.You’ve already sent feedback on this page. Thanks.

Product Analyst
Focuses on event analysis, conversion funnels, and retention, using clear metric definitions and structured analysis to explain product performance.