Measurement
Part of How to measure prospecting software performance
How first-touch, last-touch, linear and position-based attribution differ for prospecting
Explains first-touch, last-touch, linear and position-based attribution for prospecting, with a decision table, a worked credit example and sales cycle guidance.
What to take away
- Attribution methods split into two familiessingle-touch models that credit one interaction, and multi-touch models that spread credit across several. The right choice follows the buying cycle, not the vendor demo.
- First-touch suits demand creation reporting, last-touch suits closing teams, and linear or position-based models suit longer cycles with several stakeholders.
- The wider prospecting software measurement and reporting guide 2027 sets out how attribution fits into a full reporting stack.
How the main attribution methods differ
Every model is a convention for sharing credit rather than a measure of truth. The difference lies in how much weight each one places on the first interaction and the last.
Attribution Models Compared
First-touch
- Credit
- First interaction
- Answers
- New names source
- Ignores
- Middle touches
- Best for
- Outbound prospecting
Last-touch
- Credit
- Final interaction
- Answers
- What closed deal
- Ignores
- Starting content
- Best for
- Sales-owned numbers
Linear
- Credit
- Evenly split
- Answers
- Every touch equal
- Ignores
- Touch importance
- Best for
- Simple explanation
Position-based
- Credit
- First and last heavy
- Answers
- Long cycle influence
- Ignores
- Middle detail
- Best for
- Multiple stakeholders
Single-touch models
First-touch gives all credit to the interaction that created the contact. Teams running heavy outbound prospecting often start here because the answer is easy to act on.
Last-touch gives all credit to the final interaction before a deal closes.
Multi-touch models
Linear splits credit evenly across every recorded touch.
Position-based models weight the first and last touches more heavily and spread the remainder across the middle. They suit longer cycles where several people influence the decision.
Prospecting tools such as Apollo.io and ZoomInfo record outbound touches; attribution tools such as Dreamdata and HockeyStack join those touches to CRM opportunities.
Decision table
Choose
- Short cycle, one decision maker
- Last-touch, with first-touch as a secondary view
- Long cycle, several stakeholders
- Position-based or linear
- New outbound programme, unknown channel mix
- First-touch
- Small team, limited tracking
- One documented model
Avoid
- Short cycle, one decision maker
- Linear, which adds noise without insight
- Long cycle, several stakeholders
- Last-touch alone
- New outbound programme, unknown channel mix
- Position-based before you have touch data
- Small team, limited tracking
- Running several models in parallel
Worked example: three deals, three rankings
The figures below are illustrative, not a benchmark.
| Deal | Value | Touch sequence |
|---|---|---|
| A | £20,000 | Paid search, webinar, demo |
| B | £10,000 | LinkedIn, webinar, demo |
| C | £30,000 | Organic search, email, demo |
First-touch credits the first channel in each deal: organic search £30,000, paid search £20,000, LinkedIn £10,000. Last-touch credits demo with all £60,000. Position-based with 40% first, 20% middle, 40% last gives demo £24,000, organic search £12,000, paid search £8,000, email £6,000, webinar £6,000, LinkedIn £4,000.
The ranking changes: first-touch puts organic search first; last-touch puts demo first; position-based puts demo first but keeps organic search and paid search in the running.
Whichever method you choose, the data feeding it must be lawful. The ICO guidance on direct marketing using electronic mail explains consent and the soft opt-in for business contacts, which affects what you can record and reuse.
The table is a starting point, not a rule. As a rule of thumb, below 200 tracked touches a month, a single model is all the volume supports; above 200, position-based starts to pay for the extra setup.
Matching the method to your sales cycle
Short cycles with one decision maker
Where one person researches and buys within days, first-touch and last-touch usually tell the same story. Pick last-touch if the sales team owns the number, and keep first-touch as a secondary view for marketing.
Long cycles with several stakeholders
Where a buying group forms over months, single-touch models hide the work done in the middle. Position-based or linear models give a fairer picture, provided every touch is captured in one place.
Attribution also depends on what you can lawfully store. The retained UK GDPR text sets out retention and individual rights, and those rules shape which contact records you may keep for reporting.
For teams weighing evidence quality against cost, the prospecting software benchmark research in England shows how other buyers have approached the same trade-off.
Political and public affairs teams should check published election data from the ONS elections statistics when sizing audiences, since those figures are more dependable than vendor estimates. That work sits outside this lead-generation guide; the ONS figures do not set attribution rules for prospecting pipelines.
Common questions
Which attribution method is most accurate?
None is accurate in an absolute sense. Each model is a convention for sharing credit. Accuracy comes from consistent recording and a written definition everyone accepts, not from the model itself.
Can we run two models at once?
You can, but report one as primary. Running several models without a stated primary view invites arguments about which number is real and slows every review meeting.
How often should we revisit the model?
Quarterly is a sensible rhythm. Revisit sooner if the channel mix changes sharply, because a model tuned to one mix can mislead once the mix shifts.
Does attribution change what we may store?
Yes. Retention rules limit how long you keep contact records, so your model must work within those limits rather than assume unlimited history.
