Revenue attribution by source

Revenue attribution by source is the practice of assigning closed revenue to the channels, campaigns, and touchpoints that influenced each deal. Instead of stopping at leads or pipeline, it follows the money all the way to booked revenue and asks a single question for every pound: which source should get the credit. A deal that started with a paid search click, continued through a webinar, and closed after a sales call has three plausible owners, and the attribution model decides how the revenue is split between them.

The metric matters because most teams already know their channels generate traffic. What they often cannot answer is which channel generates revenue that sticks. Paid social might produce the most leads while organic search produces the most expansion. Without attribution by source, marketing spend gets allocated on cost per lead conversion rate rather than on revenue produced, and the cheapest leads quietly starve the channels that close.

Attribution is always a model, not a fact. First-touch credits the source that started the journey, last-touch credits the source that closed it, and multi-touch spreads credit across the path. Each answers a different question, so the model you choose shapes the conclusion. The discipline is picking one model deliberately and applying it consistently rather than switching to whichever view flatters a given channel.

For each source, divide the revenue credited to it by the total attributed revenue for the period, then multiply by 100 to get a percentage share. If paid search is credited with 120,000 pounds out of 600,000 pounds in total attributed revenue, paid search holds a 20% revenue share. Repeat for every source and the shares sum to 100%.

The hard part is not the division, it is deciding how much revenue each source is credited with in the first place. That depends on the attribution model and on clean, joined data linking touchpoints to closed deals. Get those two inputs right and the percentage falls out cleanly.

1. 1

   Define the revenue you are attributing

   Choose new bookings, total contract value, or recurring revenue, and use the same definition across every source. Mixing one-time fees into some sources and not others distorts every share.

2. 2

   Choose an attribution model

   First-touch, last-touch, linear, or position-based. Each splits credit differently. Document the choice so everyone reads the numbers the same way and nobody re-litigates the model when their channel looks weak.

3. 3

   Stitch touchpoints to closed deals

   Join marketing touch data to the CRM record that booked the revenue. Without a reliable identity link from anonymous touch to closed account, attribution is guesswork dressed up as data.

4. 4

   Credit revenue to each source per the model

   Apply the model to allocate the deal value across the touchpoints on its path, then sum the credited revenue for every source across the period.

5. 5

   Convert to a percentage share

   Divide each source total by total attributed revenue and multiply by 100. Track the shares over time, because a rising share matters more than a single snapshot.

Revenue attribution by source is a headline number that hides a chain of causes. A channel share is the product of how much traffic the source sends, how well that traffic converts, the deal sizes it produces, and the rate at which those deals close. A metric tree pulls these drivers apart so a falling share has a diagnosable cause rather than a shrug.

Decomposing the metric this way shows you where to act. If paid search revenue drops, the tree tells you whether spend fell, click conversion rate softened, or average deal size from that channel declined. Each leaf points at a different team and a different fix.

There is no universal target share for any one source, because a healthy mix depends on your motion, market, and stage. What benchmarks usefully describe is the shape of a healthy portfolio: enough diversity that no single source can sink you, and a clear sense of which channels are scaling versus saturating. Use the ranges below as a sanity check, not a goal.

You do not improve the metric by gaming the model, you improve it by making the underlying decisions better and the data cleaner. The aim is twofold: shift spend toward sources that produce revenue, and trust the numbers enough to act on them.

1. 1

   Treating the model as objective truth

   Every model encodes an assumption about which touch deserves credit. Presenting one model as the answer hides that choice and invites endless argument when a channel looks bad.

2. 2

   Switching models to win the argument

   Quoting last-touch when it favours your channel and first-touch when it does not destroys trust in the whole system. Pick a model and hold it across reporting periods.

3. 3

   Ignoring deals with no tracked touches

   Dark social, word of mouth, and offline conversations are real revenue with no clean source. Bucketing them as direct or dropping them silently overstates the channels you can track.

4. 4

   Confusing attribution with incrementality

   A source can be credited with revenue it did not cause. Without holdout testing on major channels, you may pour spend into a source that was simply present when deals that were going to close anyway closed.