Last-click attribution is the model that hands 100% of a conversion’s credit to the very last touchpoint a user interacted with before converting — the final ad click, organic visit, or email link — and zero to everything that came before it. It’s the default in most reporting tools because it’s dead simple to compute and explain. It’s also the model most likely to make you defund the channels that actually built the demand it takes credit for.
Last-Click Attribution
Last-click attribution is a single-touch attribution model that assigns full credit for a conversion to the final marketing touchpoint a user clicked before converting, ignoring every earlier interaction in the journey.
Why everyone defaults to last-click
For years, “Last Click” was the literal default in Google Analytics, Google Ads, and most ad platforms — and old habits die hard. The appeal is obvious: it’s the cleanest possible story. One conversion, one touch, one channel gets the trophy. No fractional math, no journey stitching, no modeling assumptions to defend in a meeting.
That simplicity is exactly the trap. The model answers a narrow question — which channel was last in line? — and people read it as a much bigger one: which channel caused the sale? Those are not the same question, and the gap between them is where budgets go to die.
Last-click doesn’t measure influence. It measures who happened to be standing closest to the door when the buyer walked through it.
How last-click attribution works
The mechanics are straightforward, which is part of why it survives:
- Touchpoint tracking. Each interaction — a paid click, an organic landing, an email click, a referral — is logged by analytics and ad platforms via first-party cookies, click IDs (gclid, fbclid), and UTM parameters, each with a timestamp and source.
- Conversion window. Interactions are evaluated inside a defined lookback window (commonly 7, 30, or 90 days). Any conversion inside that window is eligible for attribution.
- Find the final touch. When the conversion fires, the model identifies the last recorded, eligible touchpoint before it.
- Assign 100% credit. That final touch’s channel takes the entire conversion. Every prior interaction gets nothing.
- Roll up reporting. Credit is aggregated by channel, campaign, or keyword — the basis for the ROAS and CPA numbers most teams optimize against.
A wrinkle worth knowing: GA4 retired classic last-click in favor of last-click excluding direct. Direct traffic only wins credit if it’s the only touch in the path — otherwise credit passes to the last non-direct channel. That single rule quietly reshapes which channel “closes” most of your conversions, so know which variant your reports are running before you trust a number.
Where it lies: the assisting-channel problem
Take a realistic B2B path. A prospect discovers you through a content marketing piece via organic search, comes back a week later from a retargeting ad, reads two more articles, then finally Googles your brand name and clicks a paid brand-search ad to convert.
Under last-click, paid brand search gets 100% of that conversion. Organic content — the channel that created the demand — gets 0%. Retargeting gets 0%. So the team “optimizes” by pouring budget into branded paid search (which mostly recaptures demand it didn’t create) and trims content and upper-funnel spend because the dashboard says they “don’t convert.”
That’s the core failure mode of any single-touch model, and it’s why an attribution model choice is a strategic decision, not a reporting setting. Last-click systematically over-credits bottom-of-funnel tactics and starves the top of the conversion funnel that feeds them.
Last-click vs. other attribution models
| Model | Credit logic | Best for | Main weakness |
|---|---|---|---|
| Last-click | 100% to final touch | Short, single-channel paths; quick ROI sanity checks | Ignores all assisting channels |
| First-click | 100% to first touch | Pure demand-gen / awareness questions | Ignores everything that closes the deal |
| Linear | Equal split across all touches | A flat, “every touch counted” baseline | Treats a throwaway touch like a decisive one |
| Time-decay | More credit to touches nearer the conversion | Longer sales cycles with nurture | Still under-credits the top of funnel |
| Position-based (U-shaped) | 40% first, 40% last, 20% middle | Balancing demand-gen and closing | The 40/40/20 weights are arbitrary |
| Data-driven (DDA) | Algorithmic credit from observed lift | High-volume accounts wanting nuance | Black-box; needs conversion volume and clean data |
For most accounts the honest upgrade path is from last-click toward multi-touch attribution — and ultimately toward incrementality testing, which sidesteps the credit-splitting argument entirely.
The privacy era broke the data underneath it
Even if you accept last-click’s logic, the data feeding it has degraded badly. The model assumes a clean, fully observed click path — and that assumption no longer holds.
- Third-party cookie deprecation and aggressive browser restrictions (Safari ITP, Firefox ETP) shorten or sever the trails between touchpoints, so the “first” touches often vanish and the “last” one is over-represented.
- iOS App Tracking Transparency (ATT) strips the click-level signal from a large share of mobile users, leaving modeled gaps where the platform guesses.
- Consent Mode means a meaningful slice of conversions are now consent-denied and modeled, not directly observed — so the “last click” can be a statistical estimate dressed up as a fact.
- AI Overviews and zero-click search add influence that never produces a click at all: a buyer can read your brand inside an AI answer, form intent, and arrive later via “direct” — a touch last-click can’t see, let alone credit.
In other words, last-click was always a simplification. In the privacy era it’s a simplification built on increasingly incomplete data, which makes naive channel-cutting even riskier.
When last-click is actually fine
It isn’t useless — it’s just narrow. Reach for last-click when:
- The path is genuinely short and single-channel — a transactional, high-intent purchase with one obvious touch.
- You need a fast, consistent sanity check rather than a strategic budget verdict.
- Data volume is too thin for a data-driven model to be trustworthy.
- You want a stable baseline to compare other models against — the differences between models are often more instructive than any single model’s output.
Use it as one lens inside a broader digital marketing analytics practice — never as the sole arbiter of where money goes. We run last-click alongside an assisting-conversions view and at least one incrementality read before touching a budget line. If you want help wiring attribution into decisions that hold up, that’s the core of our growth program.
Frequently Asked Questions
What is last-click attribution in simple terms?
Last-click attribution gives all the credit for a sale to the final marketing touch a person clicked before buying. If someone saw a Facebook ad, read a blog post, then clicked a Google ad and converted, the Google ad gets 100% of the credit and the other two get nothing. It rewards whoever was last, not whoever was most influential.
Is last-click attribution still the default in GA4?
No. GA4 moved to last-click excluding direct (now branded “Paid and organic last click”). Direct traffic only earns credit when it’s the only touch in the path; otherwise the last non-direct channel gets it. Universal Analytics’ classic last-click is gone, so check which variant your current reports actually use.
Why is last-click attribution considered bad or misleading?
It’s misleading because it credits only the closing touch and ignores every channel that built awareness and intent earlier. Teams then defund top-of-funnel work — content, social, display — that looks unproductive but actually feeds conversions. Privacy changes like cookie deprecation and Consent Mode make its underlying data even less complete.
Last-click vs. first-click attribution — what’s the difference?
Both are single-touch models that credit one interaction completely. Last-click gives 100% to the final touch before conversion (good for spotting closers); first-click gives 100% to the first touch (good for spotting demand sources). Each tells half the story, which is why most teams move to multi-touch or incrementality testing for real budget decisions.
Should I use last-click attribution for SEO performance?
Cautiously. Last-click undervalues SEO because organic search often introduces people early, then they convert later via brand search or direct. Judging organic on last-click alone will make it look weaker than it is. Pair it with assisted-conversion data and incrementality tests to see SEO’s real contribution to the conversion funnel.