Key takeaways
- For B2B marketing attribution with a long sales cycle, don't pick one model. Report first-touch, influence and click-ID proof side by side, and say which claim each one supports.
- At Pipe17 the same CRM said paid search was 0.5% of revenue on first touch and that marketing touched 55% of revenue on a 180-day influence field. Both were true.
- When deal source fields are blank or wrong, work from contacts. A contact's first-touch UTM and its link to a deal are more reliable than the deal's own source field.
- Judge paid search on closed deals over a window as long as the sales cycle. A six-month keyword-level join showed a 4.0x return, one keyword at 38.2x and five that had produced nothing.
01B2B marketing attribution for long sales cycles, in short
When deals take six months or more and involve several people, no single attribution model gives a number you can defend. Report three views side by side, first-touch, influence and CRM proof, and match the lookback window to the length of the sales cycle.
The reason is in the data itself. First-touch credits one contact's first visit, usually months before the deal and often not the person who signs. Influence counts any marketing touch on the buying group, which makes it generous. Each is right about its own question. Trouble starts when one of them is shown alone on a leadership dashboard.
02Three views, never merged
At Pipe17, which sells an OrderOps platform on a sales cycle of six months or more, the CRM held two answers to "what did marketing do?" and they were more than 100x apart. First-touch source said paid search was 0.5% of revenue. A 180-day influence field said marketing had touched 55% of revenue, and 79% in the most recent quarter. About three quarters of deals listed their source as offline. Leadership dashboards showed the first number.
I stopped trying to pick one. Leadership reporting moved to three views, each with the claim it can support.
| View | Question it answers | One quarter of new business at Pipe17 | What it can't tell you |
|---|---|---|---|
| First-touch paid search | Where did the contact first come from? | 6 deals | What happened across the rest of the buying group |
| Influenced, 180-day lookback | Did marketing touch the buying group before the deal opened? | 29 deals | Whether the touch changed anything |
| CRM proof | Is there a Google click ID on a contact tied to the deal? | 6 of 6 paid deals | Anything about channels that leave no click ID |
The same rule applied to cost numbers. Cost per sales-qualified opportunity fell from $14,450 to $3,700 in two quarters, and that figure counts opportunities paid media influenced. Strict first-touch gives a higher cost. I reported both and said which was which.
03How to get data that holds up
- Define the stage you count.At Pipe17 a deal became a sales-qualified opportunity when it entered the first sales stage. I rebuilt the quarterly count from stage-entry dates because the default report undercounted. One quarter read 18 when the real number was 21.
- Work from contacts when deal fields break.A contact's first-touch UTM is reliable, and so is its association to a deal, even when the deal's own source field is blank. Start from the contacts on each deal and read their sources.
- Make every form keep the click.Forms had been routed to the CRM through Segment, with Store Leads adding enrichment. I replaced that with a small server-side snippet that posts every form straight to HubSpot, and a first-party cookie that stores the visitor's first-touch UTMs, referrer, landing page and Google click ID.
- Test the pipeline before you trust it.I compared a week on each side of the switch: 14 form submissions with 3 paid search and 3 click IDs before, 12 submissions with 5 paid search and 5 click IDs after. Nothing was lost, and all 5 paid search submissions kept their click ID.
- Fix the UTMs at the source.An audit of about 8,000 contacts found that roughly one in ten campaign names followed any convention. The standard I wrote is in my note on a UTM naming convention.
04The six-month deal-level join
With a six-month sales cycle you can't judge paid search in 90 days. So I joined six months of Google Ads spend to HubSpot deals through the click ID, keyword by keyword.
- The window held 90 paid search contacts.
- The account returned 4.0x on closed deals.
- One keyword returned 38.2x.
- Five keywords had spent real money and produced nothing. I cut them.
The join ran in the other direction too. HubSpot held 731+ contacts with a Google click ID, and opportunity and closed-won stages went back to Google Ads as offline conversions with a value on each. That setup is in my note on value-based bidding.
The same method judged another channel. I traced a full year of LinkedIn contacts into HubSpot and found the channel had spent 88% more than Google, created one deal that closed lost, and produced no closed revenue. Google created ten deals in the same period. Nobody had checked the channel against deals before.
05What all three views still miss
- Where brand searchers heard your name. 14 of 15 paid-search sales-qualified opportunities had first arrived by searching for the company's own name. Asking buyers how they heard about us showed AI assistants were behind $960K in closed-won revenue that the CRM had credited to paid search. That's in how to track ChatGPT and AI traffic in GA4.
- Whether the spend caused the deal. Attribution describes paths. It doesn't tell you what would have happened without the ad. For that I paused the brand campaign for 22 days and compared total lead flow with the 22 days before, which barely moved. The method is in how I run incrementality tests.
- Leads nobody worked. A channel can look bad in every view because its leads were never called. On LinkedIn, 42% of leads never had an owner assigned. Check routing before you read the attribution.
If your CRM gives two answers that don't agree, sorting that out is usually where I start as a B2B demand generation consultant.
FAQQuestions people ask
- What is the best attribution model for B2B with a long sales cycle?
- No single model holds up, so report more than one. At Pipe17, where deals took six months or more, Farhan Munshi reported three views side by side: first-touch source, influence with a 180-day lookback and CRM proof by Google click ID. Each view came with the one claim it could support, and the three were never blended into one number.
- Why do first-touch and influence attribution give such different numbers?
- They answer different questions. First-touch asks where a contact first came from. Influence asks whether marketing touched the buying group before the deal opened. At Pipe17, Farhan Munshi found first-touch credited paid search with 0.5% of revenue while a 180-day influence field said marketing touched 55%. Both were true.
- How do you attribute revenue when the deal source field is blank or wrong?
- Work from contacts. A contact's first-touch UTM is reliable, and so is the contact's association to a deal, even when the deal's own source field is empty. Farhan Munshi used this method at Pipe17, where about three quarters of deals listed their source as offline.
- How long should the attribution window be for B2B paid search?
- At least as long as the sales cycle. At Pipe17, Farhan Munshi joined six months of Google Ads spend to HubSpot closed deals through the Google click ID, keyword by keyword. The account returned 4.0x on closed deals, one keyword returned 38.2x and five keywords that had produced nothing were cut. A 90-day view would have missed all of it.
- How do you keep the Google click ID from the ad click to the closed deal?
- Capture it on the form and store it on the contact. At Pipe17, Farhan Munshi replaced a customer data platform with a server-side snippet that posts every form to HubSpot, plus a first-party cookie holding first-touch UTMs, referrer, landing page and click ID. In the week after the switch, all 5 paid search submissions kept their click ID.
