Last-Touch Attribution: credit the channel that closed the deal.

What last-touch attribution is

Last-touch attribution gives 100 percent of conversion credit to the final marketing touchpoint before conversion. The customer saw a Meta ad, searched Google, read a review, clicked an email, then converted. Last-touch credits the email with the whole conversion. The model was the default in Google Analytics, Adobe Analytics, and most ad platforms for 25 years. Google made data-driven attribution the GA4 default in October 2023, but last-touch still drives the majority of ad-platform-reported numbers in 2026.

The model is honest about what it measures: closing. The channel that gets last-touch credit is the channel that pushed the customer over the line. That information is valuable for closing-stage planning and for direct-response campaign optimization. The trap is using last-touch for awareness-stage decisions, because it gives no credit to the channels that introduced the customer in the first place.

How it works

Track every touchpoint. When a conversion happens, look up the last touchpoint chronologically within the lookback window. Assign 100 percent of conversion value to that channel. Sum across the period to get channel performance.

The standard implementation in ad platforms: a 30-day click + 1-day view lookback window (Meta), a 30-day post-click window (Google Ads), and similar variations across other networks. Each platform reports the conversions where it was the last paid touch, leading to systematic over-counting when platforms are combined.

When last-touch attribution fits

Last-touch is the right model for closing-stage channels (paid search on branded keywords, retargeting, lifecycle email) and for direct-response campaigns where the goal is to drive a specific immediate action. Last-touch is the wrong model when you have a multi-touch funnel and want to allocate budget across the whole journey — it will systematically over-credit closing channels.

Common failure modes

Three failures appear constantly. Letting platforms report their own last-touch numbers without deduplication. Cutting top-of-funnel channels based on poor last-touch ROAS. Treating retargeting and branded search as incremental when they are usually closing already-decided customers.

Platform self-reporting. Every ad platform reports last-touch conversions from its own clicks. Summed across platforms, the total exceeds 100 percent of conversions because the same customer is counted by multiple platforms. The fix is server-side deduplication or a unified attribution layer.

Cutting awareness too soon. Last-touch attribution under-credits awareness channels. Cutting them produces a six-month delayed collapse in total conversions as the funnel runs dry. The fix is to pair last-touch with first-touch and incrementality.

Over-crediting closing channels. Branded search and retargeting are almost always closing already-decided customers. Incrementality tests on these channels typically show 30-60 percent non-incremental conversions. The fix is to test, not to scale based on attribution alone.

Quick answers

What is last-touch attribution?

Give 100 percent of conversion credit to whichever channel touched the customer last before conversion. The default attribution model in Google Analytics until October 2023 when GA4 switched to data-driven.

Why is last-touch criticized?

It over-credits closing channels and under-credits awareness channels. Branded search and retargeting almost always close already-decided customers, but last-touch credits them as if they drove the conversion.

Is last-touch still useful?

Yes, for closing-stage analysis. Honest about what it measures. The problem is using it as the only model for whole-funnel decisions.

What is the typical lookback window?

Meta uses 30-day click + 1-day view by default. Google Ads uses 30-day post-click. Each platform has its own default.

How does last-touch survive iOS ATT?

Partially. Last-touch within the ad platform's own ecosystem still works (Meta CAPI sees the last Meta touch). Cross-platform last-touch is degraded because cookies and IDFA are restricted.

Should I use last-touch or first-touch?

Neither alone. Use both plus data-driven attribution and validate with incrementality. Different models answer different questions.

Frequently asked

What is last-touch attribution?

An attribution model that gives 100 percent of conversion credit to the final marketing touchpoint before conversion. The default in Google Analytics, Adobe Analytics, and most ad platforms for 25 years until Google switched GA4 to data-driven in October 2023.

How does last-touch attribution work?

Track every touchpoint. When a conversion happens, look up the last touchpoint chronologically within the lookback window. Assign 100 percent of conversion value to that channel.

What is the typical last-touch lookback window?

Meta: 30-day click plus 1-day view. Google Ads: 30-day post-click. GA4 (before October 2023 default change): 6-month last-click. Each platform has its own default; most teams should override to match their actual customer journey.

Why did Google switch from last-touch to data-driven attribution?

Google made data-driven attribution the default in GA4 in October 2023 because last-touch systematically over-credits closing channels and undercounts the contribution of awareness and consideration touches. Data-driven uses observed paths to allocate credit probabilistically.

When does last-touch attribution work well?

For closing-stage channels and direct-response campaigns where the goal is the immediate action. Branded search, retargeting, and email all have honest last-touch signal. The model is also fine for single-touch journeys where there is only one channel.

When does last-touch break down?

When the customer journey has multiple meaningful touches. Last-touch ignores everything before the last interaction and over-credits the closing channel. Cutting awareness channels based on weak last-touch ROAS often produces a delayed funnel collapse.

How is last-touch different from data-driven attribution?

Last-touch uses a fixed rule (100 percent to last touch). Data-driven attribution uses observed user paths and Markov-chain modeling to estimate each channel's incremental contribution. Data-driven is closer to causal.

How does iOS ATT affect last-touch?

Last-touch within a single ad platform still works (Meta CAPI sees Meta's last touch). Cross-platform last-touch attribution is materially degraded because cookies and IDFA restrict the visibility of last-touch events across different platforms.