A new client signs a six-figure engagement. Everyone wants credit.
Your CRM says the source was "organic search," because the first thing they did was land on a blog post 14 months ago. Your LinkedIn analytics say they clicked a sponsored post in June. Your email platform says they opened 23 newsletters. Your founder says, "I met their COO at a conference in 2023."
The client says, when you finally ask, "Oh, Dana at Hartwell mentioned you. Then I listened to your podcast episode with the logistics guy."
None of your software knew about Dana. None of it knew about the podcast.
That gap, between what your tools recorded and what actually happened, is the whole problem with B2B attribution. This post is part of our guide to measuring B2B marketing, and it covers the common models, where each one breaks, and the honest setup we recommend for founder-led firms.
What is pipeline attribution in B2B?
Pipeline attribution is the practice of assigning credit for a sales opportunity to the marketing and sales activities that produced it. In B2B it answers one practical question: where did this deal come from, so you know where to put more time and money. Every model gets it partly wrong. The goal is a version that's wrong in ways you understand.
Why attribution is harder in B2B than anywhere else
Three things break the simple models.
The journey is long. 6sense's 2025 B2B Buyer Experience Report puts the average buying cycle at about 10 months globally and 11 months in North America. Cookies expire. People switch devices. The person who read your article in January isn't the person who books the call in October.
Buyers do most of the work before you see them. The same 6sense study found 79% of buyers reach out first, and they do it after about 61% of the buying process is done. By the time someone books a call, they've usually already ranked you. They put their top-ranked vendor in the lead early and bought from that vendor 77% of the time.
The most important touches are invisible. Peer recommendations, podcasts, private LinkedIn messages, a mention in a Slack group, a partner who forwards your article. Tracking software sees none of it. In service firms, where trust drives the buy, those invisible touches often carry the most weight.
The common models, and where each one breaks
First-touch attribution
How it works: All the credit goes to the first recorded interaction.
Where it helps: It tells you which channels introduce new people to your firm. Good for judging awareness work.
Where it breaks: "First recorded" is not "first." The real first touch is often a referral or a conversation your software never saw. First-touch also ignores everything that actually convinced the buyer.
Last-touch attribution
How it works: All the credit goes to the final interaction before the opportunity opened.
Where it helps: It tells you which channels convert people who are already interested.
Where it breaks: It rewards whatever sits closest to the booking button. Your "book a call" page and branded search will look like heroes. The article, newsletter, and webinar that made the person want to book get nothing. Cut those, and next quarter the booking page has nobody to convert.
Multi-touch attribution
How it works: Credit is spread across several recorded touches using a rule. Equal shares (linear). More weight on the first and last touch (U-shaped). More weight on recent touches (time decay). Or a statistical model that assigns weights for you.
Where it helps: It admits that several things contributed. For companies with thousands of deals a year and clean tracking, the patterns can be useful.
Where it breaks: It only divides credit among the touches it can see. If the referral and the podcast did the real work, multi-touch spreads 100% of the credit across the 30% of the journey that was visible. It looks rigorous, which makes it more dangerous. And most founder-led firms close dozens of engagements a year, not thousands. That's far too few deals for a weighting model to tell you anything reliable.
Platform-reported attribution
How it works: Each ad and email platform reports its own conversions.
Where it breaks: Every platform grades its own homework. Each one claims the deal. Add up the claims and you'll often find you sourced well over 100% of your pipeline. We cover this in why most B2B marketing dashboards lie.
Reality: the hybrid setup we recommend
You don't need a perfect model. You need an honest one. For founder-led B2B firms, that means combining two imperfect sources and comparing them.
Source 1: the original source field
Every opportunity in your CRM gets one original source from a short, fixed list:
- Marketing (with the specific asset or channel noted)
- Referral (with the referrer's name)
- Founder or partner network
- Existing client (expansion or renewal)
- Strategic partner
One source per opportunity. No splitting. This is what feeds Pipeline Sourced in The 5-Metric Ceiling, our five-number framework for measuring B2B marketing.
Source 2: what the buyer says
Add a required open-text question to every booking form and every intake call: "How did you hear about us?" Open text, not a dropdown. Dropdowns get you "Other" and "Google." Open text gets you "Dana at Hartwell mentioned you, then I listened to your podcast."
Write the answer into the CRM in the buyer's own words.
Then compare them every quarter
When tracking and the buyer disagree, the buyer usually wins. They don't remember everything, but they remember what mattered to them.
The disagreements are where you learn the most. If your CRM says "organic search" and buyers keep saying "my peer group recommended you," you have a referral engine that happens to include a Google search step. If buyers keep naming your podcast and your software can't see the podcast at all, you've found a channel your dashboard was hiding.
Track influence separately
Referrals stay referrals. But if the referred buyer read your case study, attended your webinar, or replied to your newsletter in the 90 days before the opportunity opened, flag it as marketing-influenced. That feeds Pipeline Influenced, the second number in the ceiling.
This matters because in most service firms, marketing's biggest job is making referrals land. A referral to a firm with no visible point of view is a coin flip. A referral to a firm whose thinking the buyer has already read is close to a sure thing.
What this looked like at TRANSEARCH USA
Before working with Rockstarr & Moon, TRANSEARCH USA had marketing that ran in bursts and wasn't tied to sales goals, and operations couldn't see pipeline stages or which campaigns worked. Nobody had one view of results.
The fix included shared reporting across sales, operations, and marketing, so leadership could see in one place which efforts were producing conversations. Once the three teams were looking at the same numbers, they started working as one revenue workflow. Booked calls rose 969%, and clicks to book a call rose 115.6%. See the TRANSEARCH USA case study.
Shared visibility is what changed. You can't double down on what's working until everyone agrees on what's working.
A worked example: the quarter where search "won"
Say you run a boutique consulting firm. Your Q2 attribution report, built from website tracking, says organic search sourced 9 of your 14 new opportunities. Your instinct is to pour money into SEO.
Then you read the "How did you hear about us?" answers on those 9:
- 4 said a former client recommended you, then they searched your name.
- 2 said they heard you on an industry podcast, then searched your name.
- 2 said they found an article while researching a problem.
- 1 said, "My CFO forwarded your newsletter."
So search was the doorway for all 9. It was the reason for only 2.
The real lessons: your referral engine is strong, and a podcast appearance did more than your ad budget. The decisions change. Make it easier for happy clients to refer you. Book more podcast spots. Keep writing the articles that pulled in those 2 cold searchers. And stop counting searches for your own firm's name as proof SEO is working.
The attribution checklist
- Pick one original source per opportunity from a fixed list of five.
- Add a required open-text "How did you hear about us?" to every booking form and intake call.
- Record the answer in the buyer's words.
- Flag referral opportunities that engaged with marketing in the prior 90 days as influenced.
- Each quarter, compare tracked sources with what buyers said, and act on the disagreements.
- Separate searches for your firm's name from non-branded search in every report.
Related reading
Attribution feeds two of the five numbers. For how to instrument all five, read the only 5 B2B marketing KPIs that matter. For the content side, where attribution gets hardest, read how to measure content marketing in B2B. For an older look at funnel math, see how to measure and grow funnel revenue.
And for the full framework, read our guide on how to measure B2B marketing.
Frequently asked questions
What is pipeline attribution in B2B? Pipeline attribution is the practice of assigning credit for a sales opportunity to the marketing and sales activities that produced it. In B2B, it answers the question "where did this deal come from?" so you know which channels deserve more time and money.
What's the difference between first-touch and multi-touch attribution? First-touch gives all the credit to the first interaction a buyer had with you. Multi-touch spreads credit across several interactions using a rule, like equal shares or extra weight on the first and last touch. First-touch is simple and biased toward awareness channels. Multi-touch looks more complete but only counts touches your software could see.
What is self-reported attribution? Self-reported attribution means asking buyers directly how they found you, usually with a required open-text "How did you hear about us?" field on booking forms and intake calls. It captures the podcasts, peer conversations, and private messages that tracking software can't see.
Which attribution model is best for a small B2B service firm? A hybrid: record the original source on every opportunity, ask every buyer how they heard about you in their own words, and compare the two each quarter. Small firms rarely have enough deal volume for algorithmic multi-touch models to mean anything.
Should referrals count as marketing-sourced? No. Count referrals as their own source. If the referred buyer engaged with your marketing before the deal opened, count that opportunity as marketing-influenced. Keeping the two separate stops marketing from taking credit for the founder's relationships, and stops the founder from ignoring marketing's role in making referrals land.
If you can't say with confidence where your last five clients came from, let's figure it out.