U-Shaped Attribution
U-shaped attribution is a multi-touch model that assigns the most credit to the first touch and the last touch before conversion, splitting the remaining credit among middle interactions.
Key takeaways
- Credit plotted by position forms two peaks at entry and conversion with a shallow trough.
- The shape asserts that earning attention and closing a decided buyer are the hardest jobs.
- In vendor-comparison markets the trough understates the middle where the choice is made.
- Organisations with a formal sales handover usually graduate to the three-peak W-shaped model.
- A deep trough plus weak first-touch tracking produces the least trustworthy report of all.
In depth
Plot credit against position along the path and the U is literal: a tall bar at the entry point, a tall bar at the conversion, and a shallow trough stretched between them. The standard configuration puts forty percent under each peak and leaves twenty percent to be shared across the trough, however many touches it contains. What makes the curve a U rather than any other shape is a claim about buying: that earning a stranger's attention and converting a decided buyer are the two hardest jobs, and the rest is supporting work.
How well the shape fits depends on where persuasion actually happens in your category. Where the product is understood and the real choice is between vendors, the middle carries the comparison work and a shallow trough understates it badly. Where almost nobody knows the category exists, the entry peak genuinely is the expensive part and the shape earns its keep. Deepening the trough sharpens the story for the two teams that own the ends while removing the evidence a middle team could use to defend its budget.
Because the two peaks map neatly onto two organisational functions, U-shaped reporting is often adopted to settle an argument rather than to describe data: demand generation owns the first peak, demand capture owns the last, and both leave the meeting holding a number. Used deliberately that is defensible, provided the team responsible for the middle gets its own metric, such as stage progression. In a quiz funnel the scorecard usually sits under the closing peak, so pair it with a first-touch view before concluding it drives acquisition.
The U assumes exactly two moments deserve emphasis, which is why organisations with a formal handover to sales usually outgrow it and move to a three-peak W-shaped model. It also asks you to trust the first and last touch more than everything between them, so it inherits every weakness in entry tracking: when a third of journeys begin as direct traffic, a third of the entry peak is being credited to nothing in particular. A deep trough combined with weak first-touch data is the least reliable configuration.
Example in practice
How to measure it
Read the peaks against the trough as a ratio: total credit at the two ends divided by total credit in the middle. Compare that ratio with where your own team believes persuasion happens. A funnel where sales insists the demo decides everything, reported with a twenty percent trough, is describing a disagreement rather than a measurement, and the disagreement is the thing worth resolving first.
Check the health of each peak separately. For the entry peak, track the share of paths whose first touch carries a real source rather than direct or unknown. For the closing peak, track how often the last touch is a transactional or self-initiated visit. When either share runs high, the corresponding forty percent is being handed to a placeholder rather than to a channel.
Common mistakes
The most common misuse is adopting the U because it is politically convenient and then treating its output as a measurement. Two teams each receive a headline number, the middle of the funnel disappears from the conversation, and nobody revisits the split for a year. Set a review date when you adopt the shape, and require anyone quoting the forty percent figure to name it as a configured weight rather than as a finding.
The second is applying the shape to a conversion event that sits early in the journey. If the modelled conversion is a newsletter signup, the closing peak lands on whichever page held the form, and the report concludes that the blog closes deals. Model against the outcome you actually care about, an opportunity or a closed deal, and treat earlier events as intermediate steps rather than as endpoints.
Frequently asked questions
How does U-shaped attribution differ from last-click?
Last-click gives 100% of the credit to the final touchpoint, ignoring how the lead was discovered. U-shaped splits credit so the introducing channel also gets recognized, giving a fairer view of the full journey.
What weighting does U-shaped attribution use?
The standard split is 40% to the first touch, 40% to the last touch before conversion, and 20% shared among the middle interactions. Many analytics tools let you adjust these percentages.
When should I use U-shaped attribution for a quiz funnel?
Use it when both your acquisition channel and your conversion experience deserve credit, such as an ad that drives discovery and a quiz that captures the lead. It is ideal for funnels with a clear entry and a clear conversion event.
What does the U in U-shaped attribution refer to?
The shape of the credit distribution when you chart it by position. Credit is high at the first touch, drops through the middle of the journey, and rises again at the converting touch, tracing a U. It is a description of the weighting curve, not of the customer journey itself, which rarely looks symmetrical in practice.
Should I use U-shaped or W-shaped attribution?
Use the U when marketing owns the journey end to end and there is no clean mid-point milestone. Move to the W when sales takes over at a defined moment, such as opportunity creation, and both teams need credit for the stage they influenced. The W only works if that milestone is stamped reliably in your CRM.
How much credit does the middle of the journey get?
Twenty percent in the standard configuration, shared across every middle touch. On a path with six middle interactions that is a little over three percent each, so individual nurture steps look negligible even when the sequence as a whole matters. Report the middle as a combined total as well as per touch to avoid that illusion.
Does U-shaped attribution suit B2B lead generation?
It fits reasonably when the lead conversion is the outcome you are modelling, because entry and conversion really are the two events marketing controls. It fits poorly when the outcome is revenue, since the deal is usually decided in conversations the model treats as trough. Match the shape to the outcome before adopting it.
What breaks U-shaped attribution most often?
Unreliable first-touch data. The entry peak carries forty percent of every conversion, so any gap in source capture is amplified fourfold compared with a middle touch. Direct traffic, stripped referrers and expired cookies all land on that peak. Audit entry sources before you trust anything the model says about acquisition channels.
How do I explain U-shaped attribution to a finance stakeholder?
Present it as an agreed accounting rule rather than a measurement: each sale is split forty, twenty, forty between the channel that introduced the buyer, everything in between, and the channel present at the close. Say plainly that the percentages are chosen, not observed, and show what the ranking looks like under a second split.