The short version
A long B2B sales cycle breaks most attribution because the tools forget faster than buyers decide. Google Ads and GA4 look back 90 days at most, while 6sense puts the average B2B buying cycle at about 10 months.
For a service business with a 90 to 180 day deal, the fix isn't a smarter model. It's recording the right facts at each stage in the CRM, where they don't expire.
Stamp the original source at first conversion, record the source again when the opportunity is created, add every buyer to the deal and ask how they heard about you.
Report by the month a lead or opportunity was created, not the month it closed, and send stage milestones back to ad platforms long before the deal is won.
A founder closes a $90,000 deal in March. The CRM says the source is "direct". The first form fill was in October, after the buyer clicked a LinkedIn ad. By the time the contract was signed, every ad platform had long since stopped counting that click, and the original source had been overwritten when a sales rep re-created the contact.
This is the normal state of attribution across a long B2B sales cycle. This guide covers what a typical B2B sales cycle looks like for a service business, why tools lose track of long deals, the seven practices that keep attribution intact for 90 to 180 days and longer, how to report on deals that close next quarter and a 60-day setup plan.
What the B2B sales cycle looks like for a service business
The sales cycle length is the time from first meaningful contact to signed contract. For a B2B services or SaaS firm selling five- and six-figure contracts, it commonly runs three to six months, and longer when several people have to sign off. 6sense's 2025 Buyer Experience Report found that "average buying cycles shortened from about 11 months in 2024 to 10 months in 2025" (6sense via Business Wire). That figure includes the research buyers do before they ever contact you.
Stage | Typical timing in a 6-month deal | What usually gets recorded |
|---|---|---|
Research and first exposure | Months before first contact | Little or nothing: podcasts, peers, LinkedIn posts, AI answers |
First conversion | Day 0: form, demo request or reply | Original source, if tracking is set up properly |
Qualification | Weeks 1 to 3 | Lead status, sometimes a sales-qualified lead date |
Discovery and proposal | Weeks 3 to 12 | Opportunity created, calls and meetings, if logged |
Evaluation and negotiation | Months 3 to 5 | New contacts from the buying committee, often added by hand |
Closed won or lost | Month 6 | Amount and close date |
Most attribution problems start in the gaps of that table: early research that's never recorded, and the months between first conversion and close, where sources get lost or overwritten.
Why attribution tools lose track of long deals
Lookback windows are short. Google Ads lets you set a click-through conversion window "anywhere from 1 to 30, 60, or 90 days", with 30 days as the default (Google Ads Help). GA4's window for key events is 90 days by default, with 30 and 60 as the alternatives (Google Analytics Help). A deal that closes in month six is outside every one of them.
Offline imports have limits too. Google's offline conversion import FAQ says "you can upload any conversion with GCLID so long as it's no longer than 90 days" (Google Ads Help). Wait until closed-won to upload, and many deals won't count.
Cookies and devices change. Six months is long enough for cookies to expire, browsers to change and the buyer to switch laptops.
The buyer changes. The person who filled in the form is often not the person who signs. New members of the buying committee arrive through email introductions and get added to the CRM manually, with no source at all.
CRM records get overwritten. Duplicate contacts, merges, imports and manual edits can reset the original source months after it was set.
The answer is to stop relying on tools that forget, and store the facts in the CRM at the moment they happen. Our guide to first-party data attribution explains why the CRM, not the ad platform, should be the record of truth.
Seven practices that keep a 6-month deal attributed
1. Stamp the original source at first conversion, and lock it
Capture source, medium, campaign, landing page and click ID in hidden form fields using consistent UTM parameters, and write them to a contact field that nothing can overwrite. Restrict who can edit it. This one field survives any lookback window.
2. Record the source again when the opportunity is created
Copy the contact's original source onto the deal when it's created, plus a second field for what triggered the opportunity, such as a webinar, an outbound sequence or a referral. In a long cycle, the touch that turned interest into a sales conversation is often different from the first one.
3. Save the date of every stage
Record when each deal entered each stage: lead created, qualified, opportunity created, proposal sent and closed. Most CRMs can do this automatically. Stage dates let you see which sources produce fast deals and which stall, and they make milestone-based models possible later.
4. Add the whole buying committee to the deal
Associate every contact involved with the deal, each with a role such as decision-maker, champion or finance. When the CFO joins in month four, their source won't be on the original form, but the deal will still carry the champion's source.
5. Ask how they heard about you, and confirm it on the first call
Add a required, open-text "how did you hear about us?" field to forms and confirm the answer in discovery. Self-reported attribution captures the research months that no tracking sees. Our guide to dark social attribution covers how to run it.
6. Send stage milestones to ad platforms early
Don't wait for closed-won. Upload qualified leads and opportunities as offline conversions within days of them happening, so they fall inside the platform's window and the algorithms learn from real pipeline. Our guide to attribution in GA4 and Google Ads walks through the setup.
7. Log offline activity on the deal
Meetings, events, proposal reviews and reference calls only count if they're logged and associated with the deal. In a services sale, these touches often decide the outcome.
// The Rule
Ad platforms forget after 90 days. Your CRM doesn't have to. Record the source at every stage when it happens, not when the deal closes.
How to report on deals that close next quarter
Long cycles also break the way most teams read reports. If you judge last month's marketing by last month's closed revenue, you're really judging what you did six months ago.
Report by cohort, not close date. Group leads and opportunities by the month they were created, then track how much pipeline and revenue each monthly cohort produces over time. Cohort analysis shows whether a channel is improving long before its deals close.
Use leading indicators. Qualified pipeline created by source in the last 30 to 60 days tells you far sooner than revenue whether a channel is working. Our guide to last-touch attribution covers the four metrics that predict revenue.
Judge on rolling quarters. One month of a six-month cycle is noise. Compare rolling three- or six-month periods.
Watch speed by source. Compare cycle length and win rate by original source. Referral deals often close faster than paid ones, which changes what each is worth.
Question | Report to build | Review |
|---|---|---|
Which channels start real sales conversations? | Opportunities created by original source, by month created | Monthly |
Which channels produce revenue? | Closed-won revenue by original source, by cohort month | Quarterly |
Which channels produce fast, likely deals? | Cycle length and win rate by original source | Quarterly |
What is tracking missing? | Self-reported vs tracked source on opportunities | Monthly |
Once these are reliable, a milestone model such as W-shaped, which credits the first touch, lead creation and opportunity creation, can show how channels work together. Our guide to choosing a B2B attribution model compares the options.
A 60-day setup plan
Days | Focus | Done when |
|---|---|---|
1 to 15 | Audit forms, hidden fields, UTMs and the original source field | Every new lead arrives with a locked source and click ID |
16 to 30 | Add deal source fields, stage dates, contact roles and the self-reported field | Every new opportunity carries both sources and its buying committee |
31 to 45 | Send qualified leads and opportunities to ad platforms as offline conversions | Uploads run weekly or faster |
46 to 60 | Build the four cohort reports and a monthly review | Budget decisions use pipeline by cohort, not last month's closed deals |
If you run HubSpot, our guide to HubSpot attribution shows how to set most of this up without a data team.
How mrktrs handles long-cycle attribution
We build long-cycle attribution into our pipeline and Sales Helper work for B2B services and SaaS firms between $1M and $20M in revenue: locked source fields, deal-level source and stage dates, buying committee roles, a self-reported field, offline conversion uploads at each milestone and cohort reporting your leadership team can read in a few minutes.
Losing track of where your six-month deals really started? Book a 30-minute call and we'll show you where the trail goes cold.
Frequently asked questions
How long is a typical B2B sales cycle?
It varies widely by deal size and number of approvers. Smaller B2B purchases may close in weeks, while services and software contracts involving several stakeholders often take three to six months or more once research time is included.
What are the stages of a B2B sales cycle?
Most follow a similar path: research, first contact, qualification, discovery, proposal, evaluation and negotiation, then a won or lost decision. Firms often add their own steps, such as a pilot or a security review.
What attribution model suits a long sales cycle?
Milestone-based models, which give credit to the first interaction, lead creation and opportunity creation, generally fit long cycles better than single-touch models, provided the CRM reliably records those milestones.
Why does closed-won data not appear in Google Ads?
Usually because the deal closed after the platform's conversion window ended, the click ID was never stored on the lead, or the conversion was uploaded too late. Sending earlier pipeline stages avoids most of these failures.
How can marketing be judged before long deals close?
By leading indicators such as qualified opportunities created per channel, the value of that pipeline and early conversion rates between stages, reviewed on rolling periods rather than single months.