# Lead attribution statistics 2026

Canonical: https://sourcetag.io/blog/lead-attribution-statistics

Most marketing teams cannot say which channel produced their leads. In Funnel's 2026 Marketing Intelligence Report, [86% of in-house marketers and 79% of agency marketers](https://page.funnel.io/hubfs/Gated%20content/2026-Marketing-Intelligence-Report_Digital.pdf) said they struggle to determine the impact of each channel. Funnel's later research found only [18% can clearly determine it, and 10% highly trust their attribution model](https://ppc.land/36-of-marketers-keep-funding-channels-they-think-dont-work-funnel-finds/). The data they do have is shrinking. [29.5% of internet users](https://backlinko.com/ad-blockers-users) use an ad blocker, Safari [deletes script-set cookies after 7 days](https://webkit.org/tracking-prevention/), and a compliant consent banner loses [an average of 60% of visit data](https://www.etracker.com/en/cookie-consent-benchmarks/) where consent is required. Meanwhile the average B2B journey runs [272 days and 88 touchpoints](https://www.prnewswire.com/il/news-releases/dreamdata-linkedin-ads-benchmarks-report-2026-linkedin-outperforms-all-platforms-delivering-positive-return-on-ad-spend-302709177.html).

Every number below links to the page it came from. Where a source does not publish its sample size, that is noted.

## Can marketers attribute leads to a channel?

Mostly not, and the surveys agree on that even when the exact figures differ.

| Statistic | Figure | Source |
|---|---|---|
| In-house marketers who struggle to determine each channel's impact | 86% | [Funnel, 2026 Marketing Intelligence Report](https://page.funnel.io/hubfs/Gated%20content/2026-Marketing-Intelligence-Report_Digital.pdf) |
| Agency marketers who struggle to determine each channel's impact | 79% | [Funnel, 2026 Marketing Intelligence Report](https://page.funnel.io/hubfs/Gated%20content/2026-Marketing-Intelligence-Report_Digital.pdf) |
| In-house marketers without up-to-date visibility of performance across channels | 68% | [Funnel, 2026 Marketing Intelligence Report](https://page.funnel.io/hubfs/Gated%20content/2026-Marketing-Intelligence-Report_Digital.pdf) |
| Marketers who can clearly determine the impact of each channel | 18% | [Funnel and Ravn Research, via PPC Land](https://ppc.land/36-of-marketers-keep-funding-channels-they-think-dont-work-funnel-finds/) |
| In-house marketers with high trust in their attribution model | 10% (60% moderate, 30% low) | [Funnel and Ravn Research, via PPC Land](https://ppc.land/36-of-marketers-keep-funding-channels-they-think-dont-work-funnel-finds/) |
| Marketers dissatisfied with their ability to reconcile results from different tools | 80% | [MMA, State of Attribution 2024](https://www.mmaglobal.com/node/44120) |
| Senior marketing leaders who can prove marketing's value and get credit for it | 52% | [Gartner, via StockTitan](https://stocktitan.net/news/IT/gartner-survey-finds-only-52-of-senior-marketing-leaders-can-prove-0jchbtk87ys6.html) |

The Funnel figures come from two separate studies run with Ravn Research. The 2026 Marketing Intelligence Report surveyed 238 in-house marketers plus agency staff. The later study, published in October 2026, surveyed 233 respondents. The two questions were worded differently, which is why one gives 86% struggling and the other 18% able. They point the same way.

Gartner's 52% comes from 378 senior marketing leaders surveyed in April and May 2024. It measures proving value to the business, which is a wider question than channel attribution, but attribution is usually how that proof gets made.

## Which attribution models companies use

| Statistic | Figure | Source |
|---|---|---|
| In-house marketers relying mainly on platform-reported data and simple attribution | 49% | [Funnel and Ravn Research, via PPC Land](https://ppc.land/36-of-marketers-keep-funding-channels-they-think-dont-work-funnel-finds/) |
| Expanding into multi-touch attribution or early independent measurement | 42% | [Funnel and Ravn Research, via PPC Land](https://ppc.land/36-of-marketers-keep-funding-channels-they-think-dont-work-funnel-finds/) |
| Using triangulated measurement (incrementality tests, MMM and attribution) | 9% | [Funnel and Ravn Research, via PPC Land](https://ppc.land/36-of-marketers-keep-funding-channels-they-think-dont-work-funnel-finds/) |
| Marketers still relying on a single-touch model | 25.5% | [Ruler Analytics](https://www.ruleranalytics.com/blog/insight/marketing-attribution-stats/) |
| Marketers who prefer a multi-touch model | 75.5% | [Ruler Analytics](https://www.ruleranalytics.com/blog/insight/marketing-attribution-stats/) |

These two sources do not line up neatly. Ruler's 25.5% on single touch is much lower than Funnel's 49% on platform data and simple attribution. Ruler does not publish its sample size or survey year on the page, and "platform-reported data" in Funnel's question includes anyone who reads results straight out of Google Ads or Meta, whatever model those platforms use. Treat the Funnel figure as the better described of the two.

If you are choosing between models, [first touch and last touch answer different questions](https://sourcetag.io/blog/first-touch-vs-last-touch). Keeping both on the lead record is cheaper than any multi-touch tool.

## Touchpoints before a lead converts

| Statistic | Figure | Source |
|---|---|---|
| Average B2B customer journey length | 272 days (up from 211) | [Dreamdata, LinkedIn Ads Benchmarks Report 2026](https://www.prnewswire.com/il/news-releases/dreamdata-linkedin-ads-benchmarks-report-2026-linkedin-outperforms-all-platforms-delivering-positive-return-on-ad-spend-302709177.html) |
| Touchpoints in a typical B2B journey | 88 (up from 76) | [Dreamdata](https://www.prnewswire.com/il/news-releases/dreamdata-linkedin-ads-benchmarks-report-2026-linkedin-outperforms-all-platforms-delivering-positive-return-on-ad-spend-302709177.html) |
| Channels in a typical B2B journey | 4 (up from 3.7) | [Dreamdata](https://www.prnewswire.com/il/news-releases/dreamdata-linkedin-ads-benchmarks-report-2026-linkedin-outperforms-all-platforms-delivering-positive-return-on-ad-spend-302709177.html) |
| Stakeholders in a typical B2B journey | 10 (up from 6.8) | [Dreamdata](https://www.prnewswire.com/il/news-releases/dreamdata-linkedin-ads-benchmarks-report-2026-linkedin-outperforms-all-platforms-delivering-positive-return-on-ad-spend-302709177.html) |

Dreamdata's dataset is over 66 million sessions across 3.5 million customer journeys. It covers B2B companies using its platform, so it skews toward software firms with long sales cycles. A plumber or a law firm will see far shorter journeys. The point that carries over is the gap between journey length and cookie length. A 272 day journey cannot be tracked by a cookie that lasts 7 days.

## Self-reported source vs tracked source

Refine Labs added a required free-text "How did you hear about us?" field to its booking form and compared the answers with HubSpot's tracked source over 620 conversions.

| Statistic | Figure | Source |
|---|---|---|
| Conversions software credited to web search (direct and organic) | 78% | [Refine Labs, The Attribution Mirage](https://www.refinelabs.com/article/attribution-mirage) |
| Conversions customers credited to web search | 12% | [Refine Labs](https://www.refinelabs.com/article/attribution-mirage) |
| Self-reported mentions crediting social, podcasts, word of mouth or community | 85% | [Refine Labs](https://www.refinelabs.com/article/attribution-mirage) |

This is one agency's data, and an agency that sells a dark social approach. It is still the clearest published comparison. The pattern is common: software sees the last step, usually a Google search or a typed URL, and the customer remembers what made them search. Neither answer is wrong. They describe different moments, which is the same split as first touch and last touch.

## Data lost to ad blockers, Safari and consent banners

| Statistic | Figure | Source |
|---|---|---|
| Internet users worldwide who use an ad blocker (Q2 2025) | 29.5% | [GWI, via Backlinko](https://backlinko.com/ad-blockers-users) |
| US internet users who use an ad blocker (Q2 2025) | 32.5% | [GWI, via Backlinko](https://backlinko.com/ad-blockers-users) |
| US ad blocking on desktop / other devices | 27% / 22% | [Blockthrough, via Backlinko](https://backlinko.com/ad-blockers-users) |
| Lifetime of cookies set by JavaScript in Safari | 7 days | [WebKit Tracking Prevention](https://webkit.org/tracking-prevention/) |
| Lifetime of JavaScript cookies on a page reached through a decorated link from a site classified as a tracker | 24 hours | [WebKit Tracking Prevention](https://webkit.org/tracking-prevention/) |
| Visit data lost with a compliant consent banner, where consent is required | 60% average | [etracker Cookie Consent Benchmark](https://www.etracker.com/en/cookie-consent-benchmarks/) |
| Lost ad-click-to-conversion journeys recovered by Google's consent mode modelling | More than half, on average | [Google Ads Help](https://support.google.com/google-ads/answer/12445060?hl=en) |

The ad blocker figures differ because GWI asks people whether they use one, while Blockthrough measures blocking on real page views. Both land between a fifth and a third of users.

The WebKit rule matters most for lead tracking. Outside the EU every iOS browser runs on WebKit, so the 7 day limit covers Chrome on an iPhone too. Anyone who clicks an ad and comes back to fill in your form a week later arrives as Direct. Our [Safari cookie guide](https://sourcetag.io/docs/safari-cookies) covers the fix, which is setting the cookie from a server.

etracker does not publish its sample size, and its 60% applies to European sites where consent is legally required. Google's consent mode recovery is modelled. It fills in totals inside Google Ads but does not put a source on an individual lead in your CRM.

## What we could not verify

Some numbers get quoted often and we could not trace them to a primary source. We left these out:

- The share of CRM records with a missing or "unknown" lead source. We found anecdotes, such as "a third of leads" in one vendor blog, but no study with a sample.
- The share of Google Ads spend or conversions that is unattributed.
- Touchpoint counts for B2C purchases.
- "72% of marketers lack confidence in their attribution data", attributed to Gartner in several roundups. We could not find it in a Gartner release.

## How to measure yours

You do not need a survey to know your own numbers. Open your CRM, filter leads from the last 90 days, and count how many have an empty or "unknown" source field. That is your attribution gap. Then compare the source field against "how did you hear about us" answers on the same leads, if you collect them.

To close the gap, tag paid links with [UTM parameters](https://sourcetag.io/blog/what-are-utm-parameters), store the first and last touch on arrival, and write them into hidden fields on the form so the source lands on the CRM record. [SourceTag](https://sourcetag.io/) does this on any form, falls back to the referrer for untagged traffic, and can set the cookie from your server, through the WordPress plugin or a Cloudflare Worker, to get past Safari's 7 day limit. Its script is also not on the main filter lists, so most [ad blockers](https://sourcetag.io/docs/ad-blockers) leave it alone.

## FAQ

### What percentage of marketers can attribute leads to a channel?

Very few. Funnel's 2026 Marketing Intelligence Report found 86% of in-house marketers and 79% of agency marketers struggle to determine each channel's impact. A later Funnel study put the share who can clearly determine it at 18%.

### How much tracking data do ad blockers remove?

GWI found 29.5% of internet users worldwide and 32.5% in the US used an ad blocker in Q2 2025. Blockthrough measured US blocking at 27% on desktop and 22% on other devices. Not every blocker blocks every script, so the loss depends on which tools you rely on.

### How long does Safari keep attribution cookies?

WebKit deletes cookies set by JavaScript after 7 days without user interaction with the site, and after 24 hours on pages reached through a decorated link from a site classified as a tracker. This applies to Safari and, outside the EU, to every browser on iOS. Cookies set by your server are not covered by the 7 day cap.

### How many touchpoints happen before a B2B lead converts?

Dreamdata's 2026 report, built on 3.5 million B2B customer journeys, found a typical journey has 88 touchpoints across 4 channels and 10 stakeholders and lasts 272 days. Shorter-cycle businesses will see far fewer.

## Sources

- Funnel and Ravn Research, [The 2026 Marketing Intelligence Report](https://page.funnel.io/hubfs/Gated%20content/2026-Marketing-Intelligence-Report_Digital.pdf) (PDF, published late 2025)
- PPC Land, [36% of marketers keep funding channels they think don't work, Funnel finds](https://ppc.land/36-of-marketers-keep-funding-channels-they-think-dont-work-funnel-finds/) (2026)
- MMA Global, [State of Attribution 2024](https://www.mmaglobal.com/node/44120) (2024)
- Gartner press release via StockTitan, [Only 52% of senior marketing leaders can prove marketing's value](https://stocktitan.net/news/IT/gartner-survey-finds-only-52-of-senior-marketing-leaders-can-prove-0jchbtk87ys6.html) (2024)
- Ruler Analytics, [Marketing attribution statistics](https://www.ruleranalytics.com/blog/insight/marketing-attribution-stats/) (2026, sample not stated)
- Dreamdata via PR Newswire, [LinkedIn Ads Benchmarks Report 2026](https://www.prnewswire.com/il/news-releases/dreamdata-linkedin-ads-benchmarks-report-2026-linkedin-outperforms-all-platforms-delivering-positive-return-on-ad-spend-302709177.html) (2026)
- Refine Labs, [The Attribution Mirage](https://www.refinelabs.com/article/attribution-mirage) (2024)
- Backlinko, [Ad blocker usage and demographic statistics](https://backlinko.com/ad-blockers-users) (2026, GWI data from Q2 2025)
- WebKit, [Tracking Prevention in WebKit](https://webkit.org/tracking-prevention/) (current documentation)
- etracker, [Cookie consent benchmarks](https://www.etracker.com/en/cookie-consent-benchmarks/) (2025, sample not stated)
- Google, [About consent mode modelling](https://support.google.com/google-ads/answer/12445060?hl=en) (current documentation)
