Most B2B marketing attribution was built for journeys that finish inside a month. The average B2B purchase now takes about ten months, and most of it happens before the buyer ever talks to you.
B2B marketing attribution is the work of deciding which marketing activity deserves credit for a deal. It is easy when a buyer clicks an ad and signs the same week. It breaks when the deal takes most of a year, involves ten people and starts in a conversation nobody logged.
That is now the normal case. 6sense’s 2025 B2B Buyer Experience Report, a survey of nearly 4,000 B2B buyers, puts the average buying cycle at 10.1 months, down from 11.3 months in 2024. Buyers first contacted a seller about 61% of the way through that journey, and they started that contact themselves 79% of the time.
The short answer: ad platforms can only credit what they saw in roughly the last 90 days, so pair their reports with asking buyers how they found you, and judge both against pipeline over your real sales cycle.
This piece is for the B2B SaaS or fintech growth lead who has to defend a budget with numbers that do not add up. It covers what attribution can and cannot tell you over a long cycle, where multi-touch models fail, what asking buyers adds, and the short list of numbers worth reporting instead.
The short version
- Google Ads and GA4 look back 90 days at most. A ten-month sales cycle sits mostly outside every window.
- Multi-touch attribution can only split credit among the touches it recorded. Everything else gets nothing.
- Asking buyers “How did you hear about us?” finds sources that software misses, and it costs one form field.
- Report pipeline by source, time lag and CAC payback over a window that matches your sales cycle.
Table of Contents
What B2B Marketing Attribution Can and Can’t Tell You Over a Long Cycle
Software attribution answers a narrow question well: which tracked touch came before a tracked conversion, inside a set window. The trouble is the window.
Google Ads lets you record a click-through conversion for anywhere from 1 to 90 days after an ad interaction, depending on the conversion source, and applies 30 days on Search and Display if you leave the setting alone (Google Ads Help, About conversion windows, checked 29 October 2026). GA4’s key event lookback window defaults to 90 days for most key events, with 30 and 60 days as the alternatives (Analytics Help, Select attribution settings, checked 29 October 2026). Neither goes past 90.
Set that against 10.1 months. If the conversion you track is a signed deal, only ad clicks from the final 90 days can receive any credit. The conference, the podcast, the analyst call and the peer recommendation from six months earlier are invisible by design.
Now look at where the sales conversation starts. 6sense puts the average point of first contact in 2025 at about 27 weeks, just over six months in. Everything before it is the stretch where the buying group forms its shortlist, and it is the part attribution software sees least.
That stretch decides most deals. 6sense found that four out of five deals are won by the vendor buyers already favored before first contact, and that the winner was on the Day One shortlist 95% of the time.
So what can b2b attribution tell you? Which channels capture demand that already exists, and roughly how fast they turn it into conversations. What it cannot tell you is where the preference was formed. Treat it as a demand-capture report, not a demand-creation report.
One caution on the benchmark. 6sense’s buyers reported a median purchase of $200,000 to $300,000, and the study required at least $25,000 of spend in the past two years to take part. Don’t assume a smaller deal means a shorter cycle, either: 6sense found buying-group size explained nearly a quarter of the variation in cycle length, while purchase cost accounted for less than 1%. Your own cycle is the number to use, and your CRM already holds it.
Where Multi-Touch Attribution Breaks
Multi-touch attribution spreads credit across several touchpoints instead of giving all of it to the last click. In principle that suits a long B2B journey. In practice it runs into four problems.
It can only divide what it saw. A multi-touch model is still a model of tracked touches. Google says as much about its own interactions-to-conversion view: it reflects only the keywords and ads in your Google Ads account, so paths can look shorter than you expect (Google Ads Help, About attribution reports, checked 29 October 2026). A buying group that met you at an event, heard you on a podcast and then searched your name shows up as one branded search.
The rules-based versions are gone from Google’s tools. In April 2023 Google announced that first click, linear, time decay and position-based models were going away across Google Ads and GA4, leaving data-driven attribution, last click and external attribution. GA4’s help center says they have been unavailable since November 2023. If a dashboard still shows a linear or U-shaped split, it comes from another tool, and it is worth knowing which.
Data-driven does not fix the visibility problem. Google describes data-driven attribution as using its AI to estimate the impact of each touchpoint. It weighs the touches it can see, inside its window and its network. It cannot give credit to a conversation it never logged.
It counts people, not buying groups. 6sense’s 2023 and 2024 research found that B2B buying groups average more than ten members. A form fill comes from one of them. The touches that moved the rest of the group sit in other browsers and other devices, often in no ad account at all.
Last Touch vs Multi Touch Attribution
Last-touch attribution gives all the credit to the final interaction before a conversion. Multi-touch splits it. Over a long cycle, neither tells you what started the deal. They only disagree about how to divide the tracked end of it.
| Method | What it credits | What it cannot see | Use it to answer |
|---|---|---|---|
| Last touch | The final tracked interaction before the conversion | Every earlier touch, tracked or not | Which campaign or keyword turned existing intent into a conversation |
| Multi-touch or data-driven | Tracked touches inside the lookback window | Anything off-platform or older than 90 days | Which keywords and campaigns assist without closing |
| Self-reported attribution | Whatever the buyer remembers and names | Touches the buyer forgot or never noticed | Where demand is being created |
Use last touch for what it does well. Use the Assisted conversions report in Google Ads to find keywords that rarely close but often sit on the path, which is what Google built it to show. Don’t use either one to decide whether brand, content or events are “working.”
Self-Reported Attribution: What Buyers Tell You
Self-reported attribution means asking buyers directly how they found you, usually with an open text field on the demo or contact form: “How did you hear about us?” It costs one field, and it reaches the part of the journey software cannot.
One published test of the gap comes from Refine Labs, a B2B demand generation agency, which ran a 12-month comparison on its own “Book a Strategy Call” form. It made the free-text question mandatory and compared the answers for 620 conversions with HubSpot’s software attribution.
- Software credited direct and organic web search for 78% of conversions. Buyers named web search 12% of the time.
- For closed-won deals, software credited web search, paid included, for 79%. Buyers named it 3% of the time.
- Buyers named dark social sources, meaning social media, podcasts, word of mouth and communities, for 98% of closed-won revenue.
Read that as one company’s evidence, not an industry benchmark. Refine Labs runs its own go-to-market on social media and a podcast and studied its own pipeline, and the article, published in October 2024, labels its charts as a 2023 analysis. What generalizes is the mechanism: where a buyer first heard of you and where they first clicked are different events, and software only records the second.
To make it usable:
- Keep the field open text. A dropdown can steer the answer toward the options you thought of.
- Require it on high-intent forms only, such as demo, contact and quote requests, not newsletter sign-ups.
- Code the answers monthly into a few buckets: search, peer or referral, event, podcast or social, partner, other.
- Store the raw answer in your CRM next to the software source, so you can compare the two deal by deal.
- Ask again on the first sales call and log both answers, so you can see whether they match.
What to Measure Instead of a Perfect B2B Marketing Attribution Model
No model gives a long cycle a clean answer. A short set of numbers, read over the right window, gives you enough to move budget with confidence.
1. Pipeline by source, in two columns. Report marketing-sourced pipeline by software source and by self-reported source, side by side.
Where the columns agree, act. Where they disagree, the gap is the finding: it shows how much of what software files under branded search or direct traffic started somewhere else. Count qualified opportunities, not leads; where the MQL and SQL line sits decides what that number means.
2. Time lag, from two places. In Google Ads, the attribution reports show average days to conversion, measured from the first or last ad interaction, and Google suggests using them to pick a conversion window that fits your cycle.
Its own example: if 75% of conversions arrive between day 25 and day 30, a shorter window misses them. The time lag view covers ad interactions on google.com only.
Then take the lag from your CRM: days from first recorded touch to opportunity, and from opportunity to closed-won. That second figure is your real sales cycle, and it is the window your marketing should be judged over.
3. Revenue sent back to the ad platform. Optimizing on form fills teaches Google to find more people who fill in forms. Uploading qualified opportunities and closed deals as offline conversions teaches it which clicks became revenue.
Google records uploaded click conversions inside each conversion action’s window, up to 90 days for GCLID-based uploads (About attribution reports). On a ten-month cycle many closed deals will fall outside that, so upload the opportunity stage as well, because it arrives earlier. Upload within seven days of the event, too: Google’s attribution reports skip offline conversions uploaded later than that, even though standard reporting still counts them.
4. CAC payback period, at the top. CAC payback is how many months of gross margin it takes to earn back what you spent to win a customer.
It is a number finance already understands, and it survives the attribution argument because it is calculated on total spend. In the 2026 Aleph × Benchmarkit study of 198 companies, median SaaS CAC payback is about 16 months, and closer to 22 at $50K to $100K ACV. The detail is in our CAC payback period benchmarks for B2B SaaS and fintech, and you can work out your own figure with the free CAC calculator.
| Number | Where it comes from | Window | The decision it informs |
|---|---|---|---|
| Pipeline by software source | CRM, fed by UTMs and the ad platforms | Your sales cycle | Which channels capture demand |
| Pipeline by self-reported source | The open form field, coded monthly | Your sales cycle | Which channels create demand |
| Time lag | Google Ads attribution reports and the CRM | Up to 90 days in Ads; full cycle in the CRM | How long to wait before judging a channel |
| Offline conversions | CRM uploads to Google Ads | Up to 90 days for GCLID | What the bidding learns from |
| CAC payback period | Finance: spend and gross margin | Trailing quarters | Whether acquisition pays for itself |
5. All of it over a window that matches the cycle. A 30-day read on a ten-month cycle mostly measures deals that were already moving.
That is why our reporting runs over a window matching the sales cycle, and why we do not report any 30-day performance window or blended ROAS. Why ROAS misleads in B2B SaaS covers the second point. If you would rather have someone find the gaps for you, the tracking and attribution section of the Acquisition Audit covers what you are measuring, what it hides and the window it should be measured over.
FAQ: B2B Marketing Attribution
What is B2B marketing attribution?
It is the process of giving credit for pipeline and revenue to the marketing touches that came before them. In B2B the conversion that matters is usually an opportunity or a signed contract, so the attribution that counts lives in the CRM, not only in the ad platform.
What is the best attribution model for B2B?
Inside Google’s own tools the choice is now narrow: data-driven attribution is the default in Google Ads and GA4, and last click is the alternative. Neither works alone over a long cycle, so pair one of them, to see what captures demand, with self-reported attribution to see what creates it, both judged over your real sales cycle.
Is multi-touch attribution worth it for B2B?
It is useful for spotting keywords and campaigns that assist inside your ad accounts. It is not worth buying as an answer to which channels create demand. If you bring CRM deals back into Google Ads, upload them within seven days of the event, or the attribution reports will not see them.
How long should a B2B attribution window be?
As long as the platform allows, which is up to 90 days in Google Ads and GA4, plus a CRM view that runs the full length of your sales cycle. Check the time lag report first: if conversions cluster late, a 30-day window undercounts them. Change it early: a new window applies only to conversions from that day forward, and Google does not go back and count the ones the old window missed.
B2B Marketing Attribution: What to Do Next
Pull your average sales cycle from the CRM and set it beside your conversion windows. Add an open “How did you hear about us?” field to your demo form this week. In a quarter you will have two columns of pipeline by source, and the gap between them will tell you more about your B2B marketing attribution than any new model.
