The definition
Lead attribution is the practice of recording where a lead came from, so that revenue can later be traced back to the channel that produced it. The unit of attribution is the touch: an interaction between the future customer and you. A model decides which touch gets the credit. First touch credits the channel that started the relationship. Last touch credits the channel closest to the enquiry. Multi-touch spreads credit across several interactions.
In a CRM context the definition is narrower and more useful: attribution is a source value stamped on every lead at creation, which survives conversion to a deal and shows up next to won revenue in reporting. That single stable field is what this page is about. Everything more sophisticated depends on it existing first.
Why it matters
Attribution is how marketing spend stops being a matter of opinion. Work an example. A team spends €3,000 per month on paid ads and €1,500 on event sponsorships. Over two quarters, the source field shows paid ads produced 90 leads, 12 deals and €48,000 won. Events produced 20 leads, 8 deals and €96,000 won. Cost per euro won: €0.38 for ads, €0.09 for events. The channel that looked small in lead count is four times more efficient in revenue, and the next budget round has an actual basis.
Without the field, the same team argues from anecdote, and the loudest channel wins. Attribution also sharpens sales itself: win rate split by source typically spreads from single digits on cold outbound to 40 percent or more on referrals, which changes where qualification effort should go, and gives the MQL and SQL handoff its scoreboard.
How to set it up in Pipedrive
The core is one single-option field named Source, on the lead or deal, required at creation, with six to ten honest options. Add an optional Source detail text field for the campaign or referrer name. Because leads convert into deals with their fields intact, stamp the source as early as possible, which in practice means at the lead inbox, not at deal creation. Web forms and integrations should write the field automatically, so the human obligation only covers what humans create. The field design rules, including why required fields must stay few, are in custom fields.
Reporting is then two Insights views. Won deal value split by source answers the budget question. Deal conversion split by source answers the quality question. Read them quarterly with marketing at the table, and track the share of unknown as a hygiene number with a target below ten percent.
Where teams get it wrong
The classic failure is the optional field: three years of data, 60 percent empty, and every report ends in we cannot really tell. Required at creation, short list, done. The second failure is option sprawl, where the list grows a new value per campaign until reps face 40 choices and pick the top one. Campaigns belong in the detail field, channels in the list. The third is retroactive rewriting: someone updates old deals from memory before a budget meeting, which converts a data gap into an outright fiction.
And keep the model constant. Switching from first touch to last touch mid-year makes every historical comparison invalid, usually right when someone needs the trend. Pick first touch, write it down where reporting definitions live, and let an analytics tool handle any multi-touch curiosity outside the CRM.
Questions
First touch or last touch: which should we use?
For a sales team in Pipedrive, first touch, held constant. It answers the budget question of which channels start relationships that end in revenue. Last touch flatters whatever happens near the close, usually a sales conversation, and tells marketing nothing. Larger setups run both in an analytics tool, but the CRM field should carry one stable answer.
How many source options should the field have?
Between six and ten. Referral, outbound, website, paid ads, event, partner and existing customer cover most B2B teams. Fewer and everything lands in other. More and reps guess, which is worse than a short honest list. Add a free-text detail field next to it for campaign specifics.
Can Pipedrive track attribution automatically?
Partly. Web forms and integrations can stamp the source on leads they create, and deals created from those leads keep it. Anything a human creates by hand relies on the human, which is why the source field must be required at creation and short enough to answer honestly.
What do we do with deals that have source unknown?
Treat the share of unknown as a data quality metric in itself and drive it under ten percent. Above that, channel comparisons become fiction. Backfilling old deals from memory is usually wasted effort. Fix the intake so new records arrive attributed, and let the clean data accumulate.