` clicks, iOS app webviews (Instagram, TikTok, Facebook, X in-app browsers all strip the referrer by default), enterprise VPN proxies that strip referrers, and ad blockers that strip query strings.
A high share of Direct traffic is one of the strongest signals that your paid campaigns aren't tagged thoroughly enough, see the first FAQ for the common causes.
## Custom channel rules
You can override the defaults with custom channel rules:
- **Promote a source** — `utm_source=newsletter` → custom "Newsletter" channel instead of generic Email
- **Split a channel** — Paid Social → "Meta Ads", "LinkedIn Ads", "TikTok Ads" as separate channels
- **Bucket referrers** — `pinterest.com` referrals → custom "Pinterest" channel separate from Organic Social
- **Match by landing page** — visits to `/partners/...` → "Partner Program" channel
Rules are evaluated before the defaults; first match wins. Set them at **Setup -> Channel rules** in your workspace (rollout in progress; if you don't see it yet, contact hello@sourceloop.ai for early access).
## How channel attribution propagates
Once a visit is classified into a channel, the value persists across:
- **Every conversion** that visitor produces, form, meeting, chat, payment, both first-touch and last-touch
- **Every report** that groups by channel (Traffic dashboard, Content dashboard, Funnels, Contacts Hub filters)
- **Every CRM contact** created from the visitor, written into the channel field of HubSpot / Salesforce / Pipedrive
- **Every outgoing webhook** that fires when the lead is captured or updated
A channel set on the first session sticks as the **first-touch channel**. A channel set on the converting session is the **last-touch channel**. These are independent and most leads end up with different values for each (e.g., first-touch: Paid Search; last-touch: Direct).
## What's next
- **Understand the click IDs that drive paid-channel detection:** [Click IDs explained](/help/click-ids-explained/).
- **Tag your campaigns so they classify correctly:** [UTM best practices](/help/utm-best-practices/).
- **Diagnose why a specific visit shows up as Direct:** [Why is my source showing as Direct?]() (companion troubleshooting article).
- **See the channels in action on the dashboard:** [Traffic dashboard](/help/traffic-dashboard/).
## Frequently Asked Questions
### Why does so much of my paid traffic show as Direct?
Three usual reasons. (1) Your campaign URLs aren't tagged with UTM parameters and the click ID was stripped (iOS Safari with strict tracking prevention, browser extensions, or a redirect chain that drops query strings). (2) The visitor came back later through a bookmark / typing your URL after first seeing the ad. (3) The ad platform's auto-tagging is disabled and you haven't added manual UTMs. The fix is always to tag every paid link with UTM parameters in addition to relying on click IDs.
### What's the difference between Paid Search and Organic Search?
Paid Search means the visitor clicked a paid ad on a search engine (Google Ads, Microsoft Ads, etc.), detected by a click ID (gclid, msclkid) on the landing URL or by `utm_medium=cpc`. Organic Search means the visitor clicked an unpaid result on the same search engines, detected by the referrer domain and the absence of any paid click ID. Both flow from the same engines but answer different marketing questions.
### I added utm_source=facebook and utm_medium=cpc but it still shows as Direct. Why?
The UTMs are read on the landing session only. If the visitor came directly, then visited again later through a tagged link in a new session, the new session captures the UTMs but the first one stays Direct. If you mean a single landing session is still showing Direct, double-check the URL the visitor actually arrived on (some redirect chains drop query strings); open the page with `?sl_debug=1` to see what the tracker recorded.
### Can I create custom channels?
Yes, custom channel rules let you override the defaults. For example, you can map `utm_source=newsletter` to a custom channel called "Email Marketing" instead of the default Email bucket, or split Paid Social into separate "Meta Ads" and "LinkedIn Ads" channels. See custom channel rules for the setup.
### Does SourceLoop's channel definition match Google Analytics 4's?
Mostly yes, the defaults follow GA4's standard channel grouping (Paid Search, Organic Search, Paid Social, Organic Social, Email, Referral, Direct, Display, Affiliate, Audio, SMS, Push). The differences are (1) SourceLoop adds Paid Shopping and Organic Shopping for ecommerce, (2) SourceLoop uses click IDs as a stronger signal than referrers (a referrer-less click on a gclid still resolves to Paid Search), and (3) SourceLoop's Direct bucket is narrower because click IDs can rescue visits that GA4 would flag as Direct.
### How is the channel decision made, in what order?
Click IDs first (highest confidence), then UTM parameters (utm_medium + utm_source), then the referring domain (matched against a curated list), then Direct as the fallback. The first signal that produces a confident classification wins; later signals don't override earlier ones in the same session.
### Why does my Reddit / Twitter / YouTube visit show as Referral instead of Organic Social?
Because the referrer crossed into your site without a UTM tag or platform click ID. Without those signals, SourceLoop classifies by referrer domain, and Reddit / Twitter / YouTube referrals land in the broad Organic Social bucket by default. The fix is to tag any links you control (your own social posts, your channel description, your community bio) with `utm_medium=social` so the visit lands in Organic Social explicitly. Untagged shares from other people will continue to land as Organic Social automatically based on the referring domain.
---
# 9 Click ID and their Ad Platforms [With Example]
Plain-English reference for every paid click ID SourceLoop reads, gclid, gbraid, wbraid, msclkid, fbclid, li_fat_id, ttclid, epik, rdt_cid.
Source: https://sourceloop.ai/help/click-ids-explained/
Updated: 2026-05-29
---
A **click ID** is an opaque token that an ad platform appends to your landing URL when someone clicks one of your ads. Different platforms use different names, all of them serve the same purpose: a high-confidence proof that "this visit came from this ad."
For attribution accuracy, click IDs do three things UTMs alone can't:
1. **Survive privacy stripping.** Some iOS browsers and ad blockers remove the `Referer` header but leave query strings intact. The click ID hangs on.
2. **Encode platform-specific data the ad network can decode.** Pushing an offline conversion back to Google Ads with the original `gclid` lets Google match the conversion to the exact campaign, ad group, keyword, audience, and creative the click came from, depth no UTM can carry.
3. **Are set automatically.** Auto-tagging is on by default in most ad accounts, so click IDs land on every paid click without per-campaign URL-tagging discipline.
This article lists every click ID SourceLoop reads, what platform sets it, and what channel it implies.
## The full table
| Click ID | Platform | Channel | Notes |
|---|---|---|---|
| **`gclid`** | Google Ads | Paid Search | The classic Google Ads click ID. Set on every paid Google Search and Display click when auto-tagging is on. |
| **`gbraid`** | Google Ads (iOS) | Paid Search | iOS app campaigns when the user has restricted Apple's App Tracking Transparency (ATT). Functionally equivalent to `gclid` for SourceLoop's purposes. |
| **`wbraid`** | Google Ads (iOS web) | Paid Search | iOS web clicks under ATT restrictions. Functionally equivalent to `gclid`. |
| **`msclkid`** | Microsoft Ads (Bing) | Paid Search | The Microsoft equivalent of `gclid`. Set on every Microsoft Ads click when Final URL Auto-tagging is on. |
| **`fbclid`** | Meta (Facebook + Instagram) | Paid Social | Set on every click from inside Meta's ecosystem, paid ads, organic posts, profile-bio links, Messenger shares. The presence of `fbclid` doesn't strictly mean paid; it means the click originated from Meta. |
| **`li_fat_id`** | LinkedIn | Paid Social | LinkedIn's First-Party Ads Tracking UUID. Set by the LinkedIn Insight Tag on the first paid-ad landing for a visitor; persists across sessions. |
| **`ttclid`** | TikTok | Paid Social | TikTok Ads click identifier. |
| **`epik`** | Pinterest | Paid Social | Pinterest Promoted Pin click identifier. |
| **`rdt_cid`** | Reddit | Paid Social | Reddit Ads click identifier. |
## How SourceLoop reads them
The tracker reads every click ID present on the landing URL and stores them on the visitor's session and (on conversion) on the lead. From there:
- **For channel classification**, the click ID is the highest-confidence signal. A `gclid` on the URL means Paid Search even if the referrer is missing and no UTMs are set.
- **For offline conversion sync**, when you push a lead back to Google Ads / Microsoft Ads / Meta / LinkedIn, SourceLoop sends the stored click ID alongside the conversion so the platform can match it to the original click record.
- **For lead detail and CSV export**, all stored click IDs appear as columns on the lead row.
## Why both UTM and click ID
UTMs and click IDs are complementary, not redundant. Each survives where the other fails:
| Scenario | UTM survives? | Click ID survives? |
|---|---|---|
| iOS Safari with strict tracking prevention (referrer stripped) | ✅ Yes | ✅ Yes |
| URL shortener that strips query strings | ❌ No | ❌ No |
| Redirect chain that drops your custom params but keeps platform-set ones | ❌ Often no | ✅ Yes |
| Bookmarked URL revisited later (no original click) | ❌ No | ❌ No |
| Email forwarded with link intact | ✅ Yes | ✅ Yes |
| Ad platform auto-tagging is disabled | ✅ Yes (if you set manual UTMs) | ❌ No |
| You forgot to add UTMs to a paid campaign | ❌ No | ✅ Yes (auto-tagging) |
The best practice is to run both: keep auto-tagging on in every ad account AND add manual UTMs (`utm_source`, `utm_medium`, `utm_campaign`) to every paid link.
## Where click IDs power offline conversions
When you connect Google Ads, Microsoft Ads, Meta, or LinkedIn to SourceLoop and turn on offline-conversion sync, the lead's stored click ID is what makes the round trip possible. The flow:
1. Visitor clicks a Google Ad. `gclid=ABC...` lands on the URL. SourceLoop stores it.
2. Visitor fills out a form. SourceLoop creates a lead with the `gclid` attached.
3. Sales qualifies the lead in your CRM.
4. SourceLoop's sync to Google Ads fires an offline conversion event with the `gclid` and the lead value.
5. Google Ads receives the event, looks up the `gclid` to find the original click, and credits the conversion (with full revenue) to the right campaign, ad group, and keyword.
Without the click ID, the offline conversion still uploads (matched by hashed email instead), but Google's match rate drops from ~99% to roughly 60-80%, and you lose campaign-level granularity.
See the per-platform guides:
- [Configure Google Ads offline conversions](/help/configure-google-ads-offline-conversions/)
- [Configure Microsoft Ads offline conversions](/help/configure-microsoft-ads-offline-conversions/)
- [Configure Meta Conversions API](/help/configure-meta-conversions-api/)
- [Configure LinkedIn offline conversions](/help/configure-linkedin-offline-conversions/)
- [Configure TikTok Events API](/help/configure-tiktok-events-api/)
## How to verify click IDs are landing
The fastest way to check that auto-tagging is on and SourceLoop is picking up the click IDs:
1. Click one of your own live ads (preferably from an incognito window so the click isn't deduped against your existing session).
2. After landing, look at the URL bar. You should see the click ID as a query parameter.
3. Open `?sl_debug=1` on the landing page and check the browser console; SourceLoop logs every captured click ID under the `[sourceloop/v3]` prefix.
4. Convert (submit a form or book a meeting) on the same session.
5. Open the lead in the SourceLoop Contacts Hub. The click ID is listed in the lead detail drawer under the technical/attribution panel.
If you can see the click ID on the URL but not on the lead, the most common cause is the tracker not loading on the landing page; double-check the tracking pixel install.
## What's next
- **How click IDs feed into the channel decision:** [How SourceLoop classifies marketing channels](/help/channel-definitions/).
- **Set manual UTMs alongside click IDs:** [UTM best practices](/help/utm-best-practices/).
- **Push the click IDs back to the ad platform for offline conversions:** [Configure Google Ads offline conversions](/help/configure-google-ads-offline-conversions/), [Configure Meta Conversions API](/help/configure-meta-conversions-api/), [Configure LinkedIn offline conversions](/help/configure-linkedin-offline-conversions/).
## Frequently Asked Questions
### Do I need click IDs if I'm already using UTM parameters?
Both, ideally. Click IDs are auto-appended by the ad platform's auto-tagging and survive when the visitor lands on a page that ignores or strips your UTMs (a redirect chain, a cached old page, an embed without query-string passthrough). UTMs are explicit, controlled by you, and survive when the click ID is stripped (some iOS browsers, some ad blockers). Running both is belt-and-suspenders and produces the cleanest attribution.
### How do I check if my ad platform is auto-tagging click IDs?
Google Ads, Account Settings -> Auto-tagging -> ensure "Tag the URL that people click through from my ad" is on. Microsoft Ads, Accounts & Billing -> Auto-tagging -> Final URL Auto-tagging. Meta, Account Settings -> Advanced -> URL parameters -> ensure fbclid is appended (it usually is by default). Click a live ad in private/incognito and inspect the URL you land on; if it has the click ID in the query string, auto-tagging is on.
### What does it mean if a lead has a gclid but no UTMs?
It means Google Ads auto-tagging is on and you haven't added manual UTMs to your campaigns. SourceLoop can still attribute the lead to Paid Search via the gclid, but you lose the granular campaign / ad-group / keyword detail that UTMs would carry. For full attribution depth, set up manual UTMs (utm_source=google, utm_medium=cpc, utm_campaign={campaign_name}) in addition to auto-tagging.
### What's the difference between gclid, gbraid, and wbraid?
All three are Google Ads click IDs. gclid is the classic identifier, set on most desktop and Android web clicks. gbraid replaces gclid for iOS app campaigns when the user has opted out of ATT. wbraid replaces gclid for iOS web clicks where ATT consent is restricted. Functionally for SourceLoop attribution, any of the three signals "Paid Search" with high confidence. For pushing offline conversions back to Google Ads, the same field accepts all three, see the Google Ads offline conversion guide.
### Why does fbclid show on direct-typed URLs sometimes?
Because Meta's tracking embeds fbclid in any link the user clicked from inside Facebook or Instagram, including organic posts, profile bios, and shared messages in Messenger, not only paid ads. SourceLoop treats a URL with fbclid + no utm_medium as Paid Social by default, which usually matches intent (the click came from inside Meta's ecosystem), but you can override this with a custom channel rule if you specifically want organic Facebook clicks bucketed as Organic Social.
### Are click IDs personally identifiable information (PII)?
No. A click ID is an opaque identifier the ad platform generated, it doesn't contain the user's name, email, location, or any directly identifying field. It can be linked back to a click record on the ad platform's side, which is how offline-conversion uploads work, but on its own it's not PII under GDPR / CCPA.
### How long does a click ID stay valid?
Click IDs are valid for the lifetime of the ad platform's attribution window, typically 30 days for Google Ads, 7 days for Meta (default click-attribution window). SourceLoop stores the click ID on the lead indefinitely, but pushing it back to the ad platform for offline-conversion attribution only works within the window. After it expires, the click ID still exists on the lead but the ad platform won't accept it for new offline-conversion uploads.
---
# UTM best practices for clean marketing attribution
Tag campaigns with UTM parameters that produce clean attribution. The five standard parameters, naming conventions, common mistakes, and templates.
Source: https://sourceloop.ai/help/utm-best-practices/
Updated: 2026-05-29
---
**UTM parameters** are the five query-string fields you can add to any URL so SourceLoop (and Google Analytics, HubSpot, and every other attribution tool) knows where a visitor came from. They look like this in a URL:
```
https://yoursite.com/pricing?utm_source=newsletter&utm_medium=email&utm_campaign=jan-launch
```
When the visitor lands on your site with a UTM-tagged URL, the tracker reads the parameters and stores them on the visitor's session and on every conversion that follows. This article is the practical guide: which parameters matter, naming conventions that scale, and the mistakes that cause the most cleanup pain six months in.
## The five UTM parameters
| Parameter | Required? | What it means | Example |
|---|---|---|---|
| **`utm_source`** | Yes | The platform or publication that sent the traffic | `google`, `linkedin`, `newsletter`, `partner-name` |
| **`utm_medium`** | Yes | The marketing channel | `cpc`, `email`, `social`, `affiliate`, `referral` |
| **`utm_campaign`** | Strongly recommended | The specific campaign name | `q1-launch`, `pricing-relaunch-2026`, `mql-nurture` |
| **`utm_content`** | Recommended for paid | Differentiates ads in the same campaign | `headline-a`, `image-2`, `cta-blue` |
| **`utm_term`** | Optional, paid search only | The search keyword (auto-filled by Google Ads when ValueTrack is on) | `{keyword}`, `marketing+attribution+software` |
Two are mandatory for channel attribution to work: `utm_source` and `utm_medium`. Without them, the URL falls back to referrer or click ID classification, which is less reliable.
## Naming conventions that scale
The hardest part of UTMs isn't adding them; it's keeping the values consistent across a team and over years. Three rules cover 90% of the pain:
### Rule 1: Lowercase only
UTM values are case-sensitive in reports. `utm_source=Facebook` and `utm_source=facebook` will show as two separate sources. Establish a lowercase-only convention on day one.
### Rule 2: Use dashes, not spaces or underscores
Spaces get encoded as `%20` and break the eye-readability of URLs in dashboards. Underscores are fine technically but visually crowded. Dashes (kebab-case) sort alphabetically, read cleanly in the URL bar, and copy/paste reliably:
```
✅ utm_campaign=q1-launch-2026
❌ utm_campaign=Q1 Launch 2026
❌ utm_campaign=Q1_Launch_2026
```
### Rule 3: Establish a `utm_campaign` template
Without a template, every marketer types `utm_campaign=` differently and your dashboards fragment. A simple template that scales:
```
{objective}-{audience}-{date}
```
Examples:
- `demo-smb-2026q2`
- `signup-enterprise-jan2026`
- `pricing-test-2026q1`
The objective is what you want the visitor to do (demo, signup, trial, content). The audience is who you targeted (smb, enterprise, freelancers). The date lets you compare campaigns across periods without name collisions.
For paid ads, add a fourth segment for variant identification (so `utm_content` is reserved for ad-level variation):
```
{objective}-{audience}-{date}-{variant}
e.g., demo-smb-2026q2-googlecpc
```
## Recommended UTMs per channel
### Google Ads / Microsoft Ads
```
utm_source=google (or `bing` for Microsoft)
utm_medium=cpc
utm_campaign={campaign_name} ← use ValueTrack to auto-fill
utm_content={creative}
utm_term={keyword} ← ValueTrack auto-fill
```
ValueTrack parameters: `{campaign}`, `{adgroup}`, `{keyword}`, `{matchtype}`, `{creative}` are Google's auto-fill placeholders that Google replaces with the actual values when an ad is served. Set them once at the campaign level (Google Ads → Campaign settings → Tracking template) and every ad gets the right UTMs automatically. See Google's [ValueTrack reference](https://support.google.com/google-ads/answer/2375447).
Even with auto-tagging click IDs on, manual UTMs are still recommended for the campaign-level reporting they unlock.
### Meta Ads (Facebook + Instagram)
```
utm_source=facebook (or `instagram`)
utm_medium=paidsocial
utm_campaign={campaign-name}
utm_content={ad-creative-name}
```
Meta has its own dynamic parameters (`{{campaign.name}}`, `{{ad.id}}`, `{{adset.name}}`) you can set at the ad-set level. See Meta's [URL parameters reference](https://www.facebook.com/business/help/2360940870872492).
### LinkedIn Ads
```
utm_source=linkedin
utm_medium=paidsocial
utm_campaign={campaign-name}
utm_content={creative-id}
```
LinkedIn doesn't have ValueTrack-style auto-fill at the same depth, so this is mostly manual setup per campaign.
### Email (newsletters, broadcasts, transactional)
```
utm_source=newsletter (or your-app's name)
utm_medium=email
utm_campaign={send-name} e.g., 2026-04-newsletter
```
Most email tools (Mailchimp, ConvertKit, Klaviyo, Loops, Postmark) support per-link UTM templates or per-send UTM macros. Set the template once at the brand level so every email send produces consistent UTMs without remembering to tag.
### Organic social (your own posts)
```
utm_source=twitter (or linkedin, facebook, instagram, threads)
utm_medium=social
utm_campaign=organic-{topic} e.g., organic-launch-week
```
Worth tagging your own posts because referral classification by domain lumps all social platform clicks into Organic Social. Manual `utm_campaign` on your own posts lets you see exactly which posts drove conversions.
### Partner / affiliate / sponsored placements
```
utm_source=partner-name (use a slug, e.g., g2, captera, productscope)
utm_medium=partner (or affiliate, sponsorship)
utm_campaign={placement} e.g., g2-roundup-2026
```
### Internal cross-links (do NOT use UTMs)
UTMs on internal links restart the visitor's session and overwrite the original attribution. Bad:
```
❌ [See pricing](/pricing?utm_source=blog&utm_medium=internal)
```
Use plain links for internal navigation. Only tag links that take the visitor away from your site and back.
## Common mistakes
| Mistake | What goes wrong |
|---|---|
| **Mixed case** | `Google` and `google` show as two sources, splitting the same campaign in reports |
| **Spaces in values** | URL encodes to `%20`, looks ugly in dashboards and breaks some sorting |
| **UTMs on internal links** | Restarts the visitor's session, overwriting the original first-touch attribution |
| **No `utm_medium`** | Visit falls back to referrer / click ID classification, often lands as "Direct" |
| **Free-form `utm_campaign`** | "Spring promo" vs "spring_promo" vs "Spring Campaign 2026" all show as different campaigns |
| **Reusing the same `utm_campaign` across years** | Year-over-year reports can't separate the 2025 launch from the 2026 launch |
| **Putting the email subject in `utm_campaign`** | Subjects get long; URL becomes ugly; comparison across emails is hard |
| **Tagging organic search URLs** | Pointless; search engines manage their own click parameters. Tagging cannibalises the Organic Search bucket |
## Tooling that helps
You don't need fancy tools, but a few make UTM discipline easier:
| Tool | What it does |
|---|---|
| **A UTM builder** | Anything that enforces your naming convention. Google's [Campaign URL Builder](https://ga-dev-tools.google/campaign-url-builder/), [utm.io](https://web.utm.io), or a Google Sheet with formulas |
| **A shared UTM tracker (spreadsheet)** | Single source of truth for "what UTMs did we use for the Q1 demo campaign?". One row per campaign with the exact UTM string |
| **Ad platform campaign-level templates** | Google Ads tracking templates, Meta dynamic parameters, LinkedIn campaign URL macros. Each lets you set UTMs once per campaign and the platform auto-fills every ad |
| **Email tool default UTM template** | Mailchimp, Loops, Postmark all support per-link UTM templates. Set them once at the brand level |
## What's next
- **How SourceLoop turns UTMs into channels:** [How SourceLoop classifies marketing channels](/help/channel-definitions/).
- **The other half of the attribution signal:** [Click IDs explained](/help/click-ids-explained/).
- **See the UTM breakdown on every dashboard:** [Traffic dashboard](/help/traffic-dashboard/).
- **Pick the right attribution model now that your data is clean:** [7 Types of Attribution Models](/help/types-of-attribution-models/).
## Frequently Asked Questions
### Do I really need UTMs if my ad platforms auto-tag click IDs?
Yes, for two reasons. (1) Click IDs flag the channel (Paid Search, Paid Social) but don't carry campaign / ad-group / keyword / creative detail in a form SourceLoop can decode without calling back to the ad platform. UTMs do, instantly and offline. (2) Click IDs only exist for paid channels, your newsletter, blog cross-links, partner placements, and Slack-shared links all need UTMs because no platform sets a click ID for them.
### Should I use uppercase or lowercase in UTM values?
Lowercase, always. UTMs are case-sensitive in reports, so utm_source=Google and utm_source=google show as two separate sources. Establishing a "everything lowercase" rule from day one prevents weeks of cleanup later.
### What's a clean naming convention for utm_campaign?
Something machine-parseable. A common pattern is {objective}-{audience}-{date}, e.g., demo-smb-2026-q2 or signup-enterprise-jan2026. The dashes (not spaces) let you sort campaigns alphabetically and filter cleanly. Avoid free-form descriptions like "spring campaign for SMB" which make analytics filtering painful.
### How many UTM parameters do I need?
Two are required for channel attribution to work (utm_source, utm_medium). Three is the minimum for useful reporting (add utm_campaign). Four is best for paid (add utm_content to differentiate creatives). Five if you care about keyword-level reporting (add utm_term for paid search).
### My team keeps forgetting to add UTMs. What's the easiest fix?
Three habits. (1) Use a UTM builder, every link you create goes through it; the builder enforces lowercase and a naming convention. (2) Bookmark a Google Sheet of approved UTMs per campaign so the team copies from it. (3) For ad platforms, set the UTM template at the campaign level (Google Ads -> Campaign settings -> Tracking template) so every ad's URL is appended automatically. With a campaign-level template, you stop relying on humans to remember.
### Will UTMs survive iOS Safari's privacy features?
Yes. iOS Safari (and ITP-style privacy modes generally) strip the Referer header but leave query strings intact. UTM parameters survive intact. The same goes for most ad blockers and privacy extensions. UTMs are the most resilient attribution signal in modern browsers.
---
# How SourceLoop stitches identities and deduplicates contacts
How SourceLoop merges sessions, devices, and form submissions into one contact. The signals used, when merges happen, and handling duplicates.
Source: https://sourceloop.ai/help/identity-stitching-and-dedup/
Updated: 2026-05-29
---
When a visitor lands on your site multiple times, from multiple devices, through multiple campaigns, and converts via multiple channels, SourceLoop has to decide: is this one person or several? The way it decides is **identity stitching**, and the by-product is contact **deduplication**, one row in the Contacts Hub per real human, not one row per session or form submission.
This article covers how stitching works, when merges happen, and what to do about the duplicates that occasionally slip through.
## The model
SourceLoop maintains an **identity graph** per website. Each contact is one node. Every node can have:
- **One or more anonymous identifiers** — the cookie-based IDs the tracker assigns when a new browser lands on the site. One per device/browser usually.
- **At most one canonical email** — the primary contact email used for deduplication.
- **At most one canonical phone** — secondary identifier for phone-only conversions.
- **Optional external IDs** — your internal user ID, your CRM contact ID, your Stripe customer ID, etc., used to stitch across systems.
Two browser sessions that share **any** of those identifiers point to the same contact. Two sessions that share **none** are treated as two contacts.
## How merging happens
A merge happens whenever a session emits an identifier (email, phone, or external ID) that's already attached to an existing contact in your workspace.
### Example: same person, two devices
1. Tuesday: Jane browses your pricing page from her laptop (Chrome). The tracker creates **anonymous_id_A** for her browser. Jane is anonymous, one row in your Visitors view.
2. Wednesday: Jane reads your blog from her phone (Safari). The tracker creates **anonymous_id_B** for that browser. Two anonymous visitors in your Visitors view now (laptop and phone, not yet linked).
3. Thursday: Jane submits a form from her phone with `jane@example.com`. SourceLoop creates a contact with `anonymous_id_B` linked. One row in Leads, one row left in Visitors (the laptop, still anonymous).
4. Friday: Jane visits your pricing page from her laptop and logs into her account. Your app calls `window.sourceloop.identify({email: "jane@example.com"})`. SourceLoop sees `jane@example.com` already exists as a contact, and merges `anonymous_id_A` into that contact's anonymous ID list. The laptop visitor disappears from Visitors. The contact in Leads now shows BOTH the laptop journey and the phone journey, every page view from both, every session from both, in chronological order.
### Example: two different emails, one person
1. Jane submits a contact form with her work email `jane@acme.com`. Contact created.
2. Two months later, Jane signs up for the newsletter with her personal email `jane.smith@gmail.com`. Different email, different contact created.
These stay separate by default, because from SourceLoop's view, they're two distinct identifiers. If you want to merge them, mark one as a duplicate of the other in the lead detail drawer (see below) or have your CRM sync the merge so the next inbound sync from your CRM aligns the records.
## How dedup happens within a single conversion path
Beyond cross-session merging, SourceLoop dedups single-conversion paths so a fast double-submit, a quick reload-and-resubmit, or a webhook retry doesn't create two rows.
| Path | Dedup signal | Window |
|---|---|---|
| **Form submission** (auto-captured by tracker) | Email address | A few minutes |
| **Incoming webhook delivery** | The `chat_id` field if you pass one, otherwise email | Workspace-wide, no time limit |
| **Stripe / Lemon Squeezy / Paddle / Polar / Dodo Payments** | Provider's event id | Workspace-wide |
| **CRM inbound sync** | The CRM's contact id | Workspace-wide |
| **Manual contact creation** | None, you can intentionally create two contacts with the same email; the system warns but allows it |
If the same email submits the same form three times in 10 seconds (the visitor clicked submit too many times), SourceLoop produces one contact with one conversion event, not three.
## The Mark as duplicate flag
When two real contacts represent the same human (different email addresses, accidental double-create by two team members, a CRM merge that didn't propagate), the cleanest fix is the **Is duplicate** flag in the lead detail drawer:
1. Open the duplicate contact (the one you want to mark, not the canonical one).
2. In the **Additional Information** section, toggle **Mark as duplicate**.
3. Save.
What changes:
- The contact stays in the database with all its attribution intact (no data loss).
- The default Contacts Hub view hides duplicates; the **Show duplicates** filter shows them back.
- Outgoing webhooks for that contact include `is_duplicate: true` so downstream systems can filter accordingly.
- CRM sync respects the flag, the canonical contact gets the new updates and the duplicate is skipped.
If you'd rather merge two contacts entirely (combine their journeys, conversions, revenue), that's a more involved operation. For now, contact hello@sourceloop.ai with the two contact IDs and we'll do the merge manually.
## Anonymous IDs in practice
The anonymous ID is a long random string the tracker assigns the first time a visitor lands on your site, stored in the visitor's browser cookies / local storage. It persists across:
- Multiple page views on the same site
- Multiple sessions (the visitor returning days or weeks later)
- Browser restarts
It does NOT persist across:
- Different browsers on the same device (Chrome and Safari are two visitors)
- Different devices (laptop and phone are two visitors until they identify)
- Cleared cookies / incognito → normal browsing (a fresh anonymous ID gets assigned)
Where you see the anonymous ID:
- In the tracker's debug output (`?sl_debug=1` on any page)
- As the `sourceloop_id` field on every payment integration's `checkoutMetadata()` call
- In the `anonymous_id` field on outgoing webhook payloads
- In the visitor's URL query string when crossing into a cross-domain destination (the linker appends it as `_sl=...`)
## Cross-domain stitching
If your visitor moves from `yoursite.com` to `checkout.yoursite.com` (subdomain stays within the same eTLD+1), the tracker handles stitching automatically; the cookies are shared.
If they move to a different domain (`yoursite-checkout.com`), you need to list that domain in `window.SourceLoopConfig.crossDomains` and have a tracker on the destination site too. The tracker then appends the anonymous ID to outbound URLs as a `_sl=...` parameter, and the destination tracker reads it on landing and continues the same visitor's session.
See [Cross-domain and cross-subdomain tracking](/help/cross-domain-and-subdomain-tracking/) for the config example.
## Common pitfalls
**Calling `identify()` with the wrong email.** If a user has email A and your code accidentally calls `identify({email: "...wrong-email..."})`, SourceLoop creates a second contact for the wrong email or attaches the wrong identifier to the right contact. Always identify with the email the user actually owns.
**Clearing cookies during testing.** When QA testing, repeatedly clearing cookies creates a new anonymous ID each time, which makes the Contacts Hub look like it's accumulating dozens of duplicates. Use incognito + a fresh email per test to isolate test data, then delete the test contacts when done.
**Hoping the tracker fingerprints across devices without an identifier.** It doesn't, and it can't without invasive techniques no one should be running in 2026. The only way to bridge devices is for the visitor to identify (log in, submit a form, complete a purchase) on each device.
**Treating duplicate emails in the CRM as "SourceLoop's bug".** If your CRM has two contacts for `jane@example.com` and SourceLoop syncs both, you'll see two contacts on the SourceLoop side too. Fix it in the CRM first, then let the next sync align.
## What's next
- **See the contact view that this article describes:** [How to use the SourceLoop Contacts (Leads) table](/help/contacts-leads-table/).
- **Understand how sessions work alongside identities:** [How SourceLoop defines a session](/help/session-definition/).
- **Learn how the JavaScript SDK exposes identify and identifiers:** [Install the SourceLoop SDK](/help/install-the-sourceloop-sdk/).
## Frequently Asked Questions
### I see duplicate leads in my Contacts Hub. What's happening?
Three usual causes. (1) The visitor submitted the form twice quickly enough that both submissions made it through before dedup could match them. (2) The visitor used two different emails (a personal and a work email) and SourceLoop treated them as two different people because that's accurate. (3) Two team members manually created a contact at the same time. For (1), SourceLoop's dedup window catches most rapid duplicates automatically. For (2) and (3), use the Mark as duplicate toggle on the lead detail drawer to keep both records linked but mark one as the canonical version.
### When does SourceLoop merge two browser sessions into one contact?
As soon as the same email or phone is captured on both sessions. If session A on a laptop submits the form with jane@example.com, and session B on a phone the next day also identifies as jane@example.com, SourceLoop merges both sessions' anonymous identifiers into one contact named Jane. From then on, every new session under either anonymous identifier (laptop OR phone) updates the same contact.
### Can SourceLoop merge across devices without an email?
No. Without a shared identifier (email or phone), two browser sessions on two different devices are treated as two different anonymous visitors. They merge the first time the same identity signal appears on both. This is the same behaviour as every modern attribution tool, cross-device identity without shared signals is impossible without invasive fingerprinting.
### What's an anonymous ID and where do I see it?
An anonymous ID is the cookie-based identifier SourceLoop sets on the visitor's browser the first time they land on your site. It's stored client-side, persists across sessions on the same browser, and powers the pre-conversion journey timeline. You can see it in the SourceLoop tracker's debug output (load any page with `?sl_debug=1`) or as the sourceloop_id field passed to payment integrations.
### How does this affect attribution?
When two sessions merge into one contact, their journeys merge too. The contact's first-touch attribution is the earliest session in the merged set; the last-touch attribution is whichever session contained the conversion event. The merged journey is what you see on the lead detail drawer's timeline, every session, every page view, every event, in chronological order.
### How does dedup work for incoming webhook deliveries?
Two ways. (1) If you pass a chat_id (or similar unique identifier) in the JSON payload, SourceLoop dedups by that field within the workspace. (2) Otherwise, SourceLoop dedups by email within a short window (so a fast double-fire of the same form doesn't create two contacts). The first delivery wins; later deliveries with the same identifier update the existing contact instead of creating a new one.
### Does the Mark as duplicate flag delete the duplicate?
No. It leaves the record in place but hides it from your default Contacts Hub view (you can re-show it with the filter). The original lead remains intact with all its attribution. The is_duplicate flag is also passed through to outgoing webhooks and CRM syncs so downstream systems can filter accordingly.
---
# How SourceLoop defines a session
When SourceLoop starts a session, when it ends one, how that shapes per-visit metrics on your dashboards, and where it differs from Google Analytics 4.
Source: https://sourceloop.ai/help/session-definition/
Updated: 2026-05-29
---
A **session** in SourceLoop is one continuous visit, a single bounded period during which a visitor interacts with your site. Every metric on the Traffic dashboard expressed "per session" (session count, pages per session, session duration, bounce rate) depends on this definition. Per-channel and per-campaign attribution at the session level is what powers the marketing-source story.
This article covers exactly when SourceLoop starts a new session, when it ends one, and how those decisions feed into your dashboards.
## What starts a new session
Four triggers can start a new session for an existing visitor:
| Trigger | Example |
|---|---|
| **First page view ever** | A new visitor lands on your site for the first time. |
| **30+ minutes of inactivity** | The visitor was browsing, walked away from their computer, came back 45 minutes later. New session. |
| **Attribution change** | The visitor arrives with a new `utm_source` / `utm_medium` / `utm_campaign`, a new paid click ID, or a referring domain that maps to a different channel. New session. |
| **Midnight (workspace timezone)** | A visitor browsing at 11:50 PM keeps browsing until 12:10 AM. Two sessions, one ending at midnight, one starting. |
These triggers are checked on every page view and every custom event.
## What does NOT start a new session
| Non-trigger | Why it doesn't fire a new session |
|---|---|
| **Tab switch** | The 30-minute timer tracks tracker activity (page views, custom events, scroll pings), not tab visibility. A visitor with your tab open in the background isn't "active" but isn't "inactive" either; the timer doesn't tick unless they switch back and resume. |
| **Browser refresh** | Same anonymous ID, same session, same attribution. The refresh just emits another page-view event in the same session. |
| **Internal navigation** | Clicking from `/pricing` to `/docs` to `/contact` is one session, regardless of how many page views. |
| **Cross-subdomain navigation** | `yoursite.com` → `app.yoursite.com` is the same session (cookies are shared on the same eTLD+1). |
| **Cross-domain navigation with the linker** | `yoursite.com` → `yoursite-checkout.com` with the cross-domain linker configured is the same session (the anonymous ID is passed via URL). |
## How attribution attaches to a session
Each session has exactly one set of attribution fields:
| Field | Set from |
|---|---|
| Channel | The classification rules (see [Channel definitions](/help/channel-definitions/)) |
| Source | `utm_source` parameter, or platform inferred from click ID, or referring domain |
| Medium | `utm_medium` parameter, or `cpc/cpm/organic` inferred from click ID, or `organic`/`referral` inferred from referrer |
| Campaign | `utm_campaign` parameter |
| Content | `utm_content` parameter |
| Term | `utm_term` parameter |
| Landing page | The URL of the first page view in this session |
| Referrer | `document.referrer` at session start |
| Click IDs | `gclid`, `fbclid`, `msclkid`, etc., from the URL on session start |
Once a session starts, these values stay frozen for its duration. The visitor navigating internally on your site doesn't change them. This is why the attribution-change trigger above is so important, it ensures each session has exactly one truthful set of attribution.
## Sessions roll up into the visitor and the contact
```
1 contact (1 real person)
→ N anonymous IDs (1 per device/browser they've used)
→ N sessions per anonymous ID (1 per visit)
→ N page views per session (1 per page they viewed)
→ N events per session (custom events, scroll pings, conversion events)
```
The Traffic dashboard's "Sessions" metric counts the sessions inside the date range. The "Visitors" metric counts unique anonymous IDs. The "Contacts" metric counts identified contacts (with an email or phone).
A single high-engagement contact who browsed your site five separate times across two weeks contributes:
- 1 contact
- 1 anonymous ID (assuming same browser each time)
- 5 sessions
- 25-or-so page views (assuming 5 pages per session)
## Session-level vs visitor-level metrics
Most dashboards offer both:
| Metric | Session-level | Visitor-level |
|---|---|---|
| **Count** | Number of sessions | Number of unique anonymous IDs |
| **Conversion rate** | Conversions ÷ sessions | Conversions ÷ visitors |
| **Bounce rate** | Sessions with 1 page view ÷ all sessions | Visitors who only bounced ÷ all visitors |
| **Pages per session** | Total page views ÷ session count | Total page views ÷ visitor count |
| **Duration** | Time from first to last event in a session | Sum of all sessions per visitor |
For attribution work, session-level metrics are usually more meaningful. Visitor-level metrics make sense for "how engaged is my audience over time?" but session-level metrics answer "did this specific campaign produce engaged visits?".
## How session resets affect attribution
The attribution-change trigger is the one with the biggest implication. Consider:
1. **9:00 AM** — Jane arrives via a Google Ads click. Session 1 begins with first-touch channel = Paid Search.
2. **10:00 AM** — Jane leaves your site without converting.
3. **10:25 AM** — Jane sees a LinkedIn ad in her feed and clicks it. She returns to your site. **30 minutes haven't elapsed, but the click ID and referrer have changed, so SourceLoop starts a new session.** Session 2 begins with channel = Paid Social.
4. **10:45 AM** — Jane submits the demo form during Session 2.
The conversion is attributed:
- **First-touch** = Paid Search (Session 1's channel — the earliest session in Jane's identity graph)
- **Last-touch** = Paid Social (Session 2's channel — the converting session)
Both touches are visible on the lead, and you can group revenue / leads by either model on the [Attribution dashboard](/help/types-of-attribution-models/).
If SourceLoop hadn't reset the session on the attribution change, Session 1 would still be active when the form fired, and the conversion would attribute entirely to Paid Search, hiding the role LinkedIn played in closing the deal.
## What's next
- **Understand how channels are picked on each session:** [How SourceLoop classifies marketing channels](/help/channel-definitions/).
- **See sessions per visitor on the Contacts Hub:** [How to use the SourceLoop Contacts (Leads) table](/help/contacts-leads-table/).
- **See session-level metrics on the dashboards:** [Traffic dashboard](/help/traffic-dashboard/).
- **Pick the attribution model that scopes how sessions roll up:** [7 Types of Attribution Models](/help/types-of-attribution-models/).
## Frequently Asked Questions
### When does SourceLoop start a new session?
Four triggers. (1) The visitor's first page view, ever. (2) The visitor returns after more than 30 minutes of inactivity. (3) The visitor arrives with a different attribution signal than their last session, a fresh utm_source / utm_medium / click ID, or a new referring domain that resolves to a different channel. (4) Midnight in the workspace's timezone (sessions don't span calendar days for cleaner daily reporting).
### Why don't sessions span midnight?
Daily reports get muddled when one session straddles two days. By starting a new session at midnight, the visitor's sessions-per-day count stays accurate and the Traffic dashboard's day-over-day comparisons are clean. The trade-off is a visitor browsing from 11:50 PM to 12:10 AM appears as two sessions, but the impact is small (most visitors don't browse over midnight).
### How is this different from Google Analytics 4?
GA4 also uses a 30-minute idle timeout but doesn't reset on attribution change or midnight by default. So a GA4 session can span a UTM change and span midnight, which makes campaign-level attribution and day-over-day reports harder to read. SourceLoop's stricter session boundaries mean per-session metrics are more meaningful for attribution work.
### Does a session reset if the visitor clicks a different ad?
Yes. If a visitor arrives via Paid Search, leaves, comes back via Paid Social later that hour, SourceLoop starts a new session for the Paid Social arrival. This is so each session has exactly one channel / source / campaign attribution, which makes the Traffic dashboard's per-session counts directly meaningful.
### How long is the typical session?
Most B2B sites see median session durations of 1 to 3 minutes and median pageviews per session of 2 to 4. Highly engaged visitors (research-mode prospects browsing pricing + docs + case studies) routinely produce sessions of 10 to 20 minutes. The Traffic dashboard's session-duration metric tracks the median across all sessions in the date range.
### Does a tab switch end a session?
No. The 30-minute idle timer is based on tracker activity (pageviews, custom events, scroll-depth pings), not tab visibility. A visitor with your site open in a tab who switches to email and comes back 20 minutes later is still in the same session. The timer only resets to zero when the tracker sees activity.
---
# 7 Types of Attribution Models That Marketers Should Know
A visual guide to the 7 attribution models every marketer should know, last touch, first touch, last/first non-direct, linear, U-shaped, and time decay.
Source: https://sourceloop.ai/help/types-of-attribution-models/
Updated: 2026-05-28
---
If you've ever tried to explain marketing performance in a meeting that includes a paid-acquisition lead AND a content marketer, you've felt the attribution-model problem. Paid wants credit for any conversion their ads touched. Content wants credit for the blog post that started the journey. Both are partially right. **The attribution model is the rule that decides how partial.**
This guide walks through the seven models SourceLoop supports, applies the same journey to all seven so you can see the difference at a glance, and tells you which to pick for which decision.
## The example journey we'll use
To make the seven models concrete, we'll use the same customer journey throughout this article:
| Day | Action | Channel | Touch # |
|---|---|---|---|
| Day 1 | Saw Meta ad and clicked through | **Meta** (paid social) | 1 |
| Day 4 | Searched Google for your brand | **Google** (organic search) | 2 |
| Day 7 | Clicked link in newsletter | **Email** | 3 |
| Day 14 | Converted to paid plan | — | — |
Three marketing touches over a week, one conversion on Day 14. Every model below will assign credit to those three touches differently.
## Quick reference: all 7 models on the example journey
| Model | Family | Meta (T1) | Google (T2) | Email (T3) | Best for |
|---|---|---|---|---|---|
| **1. Last Touch** | Single-touch | 0% | 0% | **100%** | Paid bidding, short cycles |
| **2. First Touch** | Single-touch | **100%** | 0% | 0% | Brand and top-of-funnel reporting |
| **3. Last Non-Direct** | Single-touch | 0% | 0% | **100%** | Daily reporting, GA-compatible |
| **4. First Non-Direct** | Single-touch | **100%** | 0% | 0% | Top-of-funnel for established brands |
| **5. Linear** | Multi-touch | 33% | 33% | 33% | Cross-channel assist credit |
| **6. Position-Based (U-Shaped)** | Multi-touch | **40%** | 20% | **40%** | Balanced reporting, B2B funnels |
| **7. Time Decay** | Multi-touch | ~13% | ~30% | **~57%** | Long sales cycles, considered purchases |
The seven models split into two families: **single-touch** (one channel gets 100%) and **multi-touch** (credit is split across the journey). Each is explained in detail below.
## 1. Last Touch Attribution
**The rule:** 100% of the credit goes to the **last marketing touch** before the conversion.
**Applied to our example:**
```
Meta ad (Day 1, T1) ░░░░░░░░░░░░░░░░░░░░ 0%
Google (Day 4, T2) ░░░░░░░░░░░░░░░░░░░░ 0%
Newsletter (Day 7, T3) ████████████████████ 100%
↑ all credit here
```
### When to use it
- **Paid acquisition bidding** — Google Ads and Meta's bidding algorithms expect last-click signals. Sending Last Touch conversions back is what makes their auctions train correctly.
- **High-intent, short-cycle conversions** — for impulse purchases or single-session conversions, the last touch usually IS the touch that mattered.
- **Quick efficiency comparisons** — "which channel produced the most last-touch conversions this week?" is a fast, defensible answer.
### When NOT to use it
- **B2B or SaaS with long sales cycles** — Last Touch credits the final email or Google search, ignoring all the brand-building that made the prospect ready to convert.
- **When debating whether to keep a top-of-funnel channel** — Last Touch will always make discovery channels look weak, even when they're driving the eventual conversions.
## 2. First Touch Attribution
**The rule:** 100% of the credit goes to the **first marketing touch** in the journey.
**Applied to our example:**
```
Meta ad (Day 1, T1) ████████████████████ 100%
↑ all credit here
Google (Day 4, T2) ░░░░░░░░░░░░░░░░░░░░ 0%
Newsletter (Day 7, T3) ░░░░░░░░░░░░░░░░░░░░ 0%
```
### When to use it
- **Top-of-funnel investment decisions** — First Touch is the only honest way to defend brand campaigns, paid social discovery, SEO content, and PR. These channels rarely convert on the last click, but the conversion wouldn't exist without them.
- **Audience-discovery analysis** — "which channel originally found our best customers?" is a First Touch question.
- **Content marketing reporting** — blog posts and pillar content are first-touch assets by design.
### When NOT to use it
- **Optimising paid-search bidding** — First Touch will tell you to invest in awareness when you may need closing efficiency now.
- **For short single-session journeys** — when the customer converts on their first visit, First Touch and Last Touch are the same thing.
## 3. Last Non-Direct Attribution
**The rule:** 100% of the credit goes to the **last non-Direct touch**. If the final session before conversion was a Direct visit (the visitor typed your URL or clicked a bookmark), skip it and credit the marketing touch before it.
**This is Google Analytics' default model** and SourceLoop's recommended default, because Direct visits usually aren't a marketing channel.
**Applied to our example** (no Direct touches in the journey, so identical to Last Touch):
```
Meta ad (Day 1, T1) ░░░░░░░░░░░░░░░░░░░░ 0%
Google (Day 4, T2) ░░░░░░░░░░░░░░░░░░░░ 0%
Newsletter (Day 7, T3) ████████████████████ 100%
```
**The difference shows up when there IS a Direct touch:**
```
Meta ad (Day 1, T1) ░░░░░░░░░░░░░░░░░░░░ 0%
Google (Day 4, T2) ░░░░░░░░░░░░░░░░░░░░ 0%
Newsletter (Day 7, T3) ████████████████████ 100% ← Last Non-Direct credits here
Direct (Day 10, T4) ░░░░░░░░░░░░░░░░░░░░ 0% ← Last Touch would credit this instead
```
### When to use it
- **GA-compatible reporting** — numbers line up with what your team sees in Google Analytics
- **Default everyday view** — for most marketing-mix decisions, more honest than Last Touch because brand-recall direct visits don't deserve credit
- **Brands with lots of repeat / Direct traffic** — particularly useful when your Direct percentage is high
### When NOT to use it
- **When you specifically want to measure brand recall** — turning a paying customer into a direct-return visitor IS a marketing outcome; this model hides it.
## 4. First Non-Direct Attribution
**The rule:** 100% of the credit goes to the **first non-Direct touch**. Same logic as Last Non-Direct, but at the start of the journey.
**Applied to our example** (no Direct touches, so identical to First Touch):
```
Meta ad (Day 1, T1) ████████████████████ 100%
Google (Day 4, T2) ░░░░░░░░░░░░░░░░░░░░ 0%
Newsletter (Day 7, T3) ░░░░░░░░░░░░░░░░░░░░ 0%
```
**The difference shows up when the journey starts with a Direct visit:**
```
Direct (Day 0, T0) ░░░░░░░░░░░░░░░░░░░░ 0% ← First Touch would credit this
Meta ad (Day 1, T1) ████████████████████ 100% ← First Non-Direct credits here
Google (Day 4, T2) ░░░░░░░░░░░░░░░░░░░░ 0%
Newsletter (Day 7, T3) ░░░░░░░░░░░░░░░░░░░░ 0%
```
### When to use it
- **Top-of-funnel reporting in established brands** — many returning customers' "first touch" is technically Direct (they've been to your site before). First Non-Direct credits the actual marketing channel that brought them back.
- **First-touch attribution with cleaner data**, like First Touch, but without Direct dominating the report.
### When NOT to use it
- **For brand-new audiences** — if your customer base is mostly first-time visitors, First Touch and First Non-Direct produce nearly identical results.
## 5. Linear Attribution
**The rule:** Credit is split **equally across every touch** in the journey.
**Applied to our example** (3 touches, so each gets 33.3%):
```
Meta ad (Day 1, T1) ███████░░░░░░░░░░░░░ 33%
Google (Day 4, T2) ███████░░░░░░░░░░░░░ 33%
Newsletter (Day 7, T3) ███████░░░░░░░░░░░░░ 33%
```
If the journey had 5 touches, each would get 20%. If 10 touches, each would get 10%.
### When to use it
- **Cross-channel assist credit** — when you want to show that every channel contributed something, without arguing about which contributed more
- **As a sanity check against single-touch models** — if Last Touch says a channel produced 50% of conversions and Linear says 18%, the difference is the credit it was over-claiming
- **Executive summaries** — Linear is the easiest multi-touch model to explain in a meeting
### When NOT to use it
- **For ad-platform bidding signals** — Google Ads and Meta won't optimise well off Linear conversions
- **When recency genuinely matters more** — Linear weights a social impression three weeks ago the same as an email click 24 hours before conversion
## 6. Position-Based (U-Shaped) Attribution
**The rule:** 40% to the **first touch**, 40% to the **last touch**, 20% split across the middle touches.
**Applied to our example** (3 touches, so middle gets all 20%):
```
Meta ad (Day 1, T1) ████████░░░░░░░░░░░░ 40% ← first touch
Google (Day 4, T2) ████░░░░░░░░░░░░░░░░ 20% ← middle
Newsletter (Day 7, T3) ████████░░░░░░░░░░░░ 40% ← last touch
```
The name comes from the credit shape: big at both ends, smaller in the middle. The thinking is that the first touch (discovery) and the last touch (close) are the two structurally important moments; middle touches mostly serve to keep the lead warm.
### When to use it
- **Balanced top-of-funnel + closing reporting** — a fair compromise for teams that want to credit both discovery channels and closing channels meaningfully
- **B2B sales reporting** — the U-shape matches how most B2B funnels feel: discovery moment, long middle of nurture, closing moment
- **Marketing-mix decisions involving cuts** — harder to defend cutting purely-middle channels, harder to over-credit pure-closing channels
### When NOT to use it
- **For single-touch and two-touch journeys** — if there are no middle touches, Position-Based degenerates into 50/50 or 100% single-touch
- **For very long sales cycles** — Time Decay is usually a better fit
## 7. Time Decay Attribution
**The rule:** Credit is weighted by recency. The closer a touch is to the conversion, the more credit it gets. SourceLoop uses **exponential decay with a 7-day half-life**: a touch 7 days before conversion gets half the credit of a touch on the same day.
**Applied to our example** (conversion on Day 14):
```
Meta ad (Day 1, 13 days out) ███░░░░░░░░░░░░░░░░░ ~13%
Google (Day 4, 10 days out) ██████░░░░░░░░░░░░░░ ~30%
Newsletter (Day 7, 7 days out) ███████████░░░░░░░░░ ~57%
↑ closest to conversion
```
The 7-day half-life means:
| Days before conversion | Credit weight (relative) |
|---|---|
| 0 days (same day) | 100% |
| 7 days | 50% |
| 14 days | 25% |
| 21 days | 12.5% |
| 28 days | 6.25% |
So even a touch 4 weeks out still gets some credit, just much less than recent touches.
### When to use it
- **Long B2B / SaaS sales cycles** — when journeys span weeks or months, Time Decay correctly reflects that the touch from yesterday is more relevant than the touch from a month ago
- **High-consideration ecommerce** — for purchases involving multiple research sessions, Time Decay credits late-funnel research moments over early-funnel discovery
- **As a smarter Last Touch** — Time Decay is essentially Last Touch with grace for very recent assists. Fairer when the last week of touches all matter, not just the final click.
### When NOT to use it
- **For short single-session conversions** — if most journeys are same-day, Time Decay collapses to Last Touch
- **When measuring top-of-funnel investment** — Time Decay will always under-credit the discovery touch from weeks ago
## Side-by-side: same journey, all 7 models
The headline comparison from earlier, repeated here as a take-home reference:
| Model | Meta (T1, Day 1) | Google (T2, Day 4) | Email (T3, Day 7) |
|---|---|---|---|
| Last Touch | 0% | 0% | **100%** |
| First Touch | **100%** | 0% | 0% |
| Last Non-Direct | 0% | 0% | **100%** |
| First Non-Direct | **100%** | 0% | 0% |
| Linear | 33% | 33% | 33% |
| Position-Based (U-Shaped) | **40%** | 20% | **40%** |
| Time Decay | ~13% | ~30% | **~57%** |
Notice how the same conversion produces wildly different stories:
- Under **Last Touch**, the campaign is "Newsletter wins, kill Meta and Google".
- Under **First Touch**, the campaign is "Meta is the hero, Newsletter and Google just happened to be in the way".
- Under **Position-Based**, "discovery and closing both matter, the middle is supporting".
- Under **Time Decay**, "the last week of touches mostly drove this, the Meta impression two weeks back contributed something but barely".
All four are technically correct. The right one depends on what you're deciding.
## What this looks like in SourceLoop
Every dashboard in SourceLoop has an attribution model selector in the top-right of the filter bar. It defaults to **Last Touch**, but you can change it (or stack multiple models) on every page.
The selector is the same control across the Traffic, Content, Locations, Devices, Paths, and Ads Performance dashboards. Picking a different model rebuilds every number on the page (the metric tiles, the chart, the breakdown table) under the rules of the model you picked.
To compare two models in the same view, click **+ Add model** in the selector. The breakdown table will split each metric column per model so you can see Last Touch and First Touch credit side by side for the same campaigns. The numerical gap between the two columns is your assist value.
## Which model should you actually pick?
For most teams the answer isn't a single model. It's **which model for which question**.
| Decision you're making | Model to use |
|---|---|
| Bidding signal for Google Ads, Meta, TikTok | **Last Touch** (or Last Non-Direct) |
| Quarterly executive update | **Position-Based** or **Linear** |
| Content / SEO reporting | **First Touch** or **First Non-Direct** |
| "Should we cut this channel?" debate | Compare **Last Touch + Linear** side by side |
| Long B2B sales cycle reporting | **Time Decay** |
| Default everyday view | **Last Non-Direct** (matches GA) |
The Traffic, Content, Locations, Devices, Paths, and Ads Performance dashboards in SourceLoop all support multiple models simultaneously via the model selector. The cleanest workflow is to **stack two models** in the same view, the gap between their columns is the actual answer to "how much of this channel is direct conversion vs assist".
## How to switch models in SourceLoop
On any dashboard:
1. Click the model selector top-right (showing "Last touch" by default).
2. Click **+ Add model** to stack a second (or third) model alongside the current one.
3. Metric columns split per model, so a single table can show Last Touch conversions, First Touch conversions, and Linear conversions side by side.
4. Use the **X** next to a model to remove it.
The model applies to Conversions counts, Revenue totals, and derived metrics (CPA, ROAS). Switching models changes every number on the dashboard, the chart, the tiles, the table, and the totals.
## What's next
- **See the dashboards where these models apply:**
- [Traffic dashboard](/help/traffic-dashboard/)
- [Content dashboard](/help/content-dashboard/)
- [Locations dashboard](/help/locations-dashboard/)
- [Devices dashboard](/help/devices-dashboard/)
- [Paths dashboard](/help/paths-dashboard/)
- [Ads Performance dashboard](/help/ads-performance-dashboard/)
- **See the actual touch sequences** multi-touch models work on: [Paths dashboard](/help/paths-dashboard/)
## Frequently Asked Questions
### What's an attribution model?
A rule for splitting credit when one conversion is preceded by multiple marketing touches. A customer might see your Meta ad, then search Google, then click a newsletter link before buying. An attribution model decides which of those three touches deserves credit. Different models give different answers; the choice depends on the decision you're making.
### What's the difference between single-touch and multi-touch attribution?
Single-touch gives 100% credit to one touchpoint, either the first or the last. Multi-touch splits credit across every touch in the journey using a weighting rule. Single-touch is simpler and matches what ad platforms expect; multi-touch is more honest about how marketing actually works, but harder to defend cleanly to a single channel.
### Which attribution model should I use?
Depends on the decision. For paid bidding signals, Last Touch. For brand awareness justification, First Touch. For everyday reporting that matches Google Analytics, Last Non-Direct. For honest cross-channel credit, Linear. For balancing top and bottom of funnel, Position-Based (U-Shaped). For long sales cycles where recency matters, Time Decay.
### Does my attribution model affect the numbers in Google Ads or Meta dashboards?
No. Each ad platform applies its own model to the conversions you send back to it. Your SourceLoop model affects SourceLoop dashboards and the conversion data pushed back, but the platform's UI still reports under its own rules.
### Why does Google Analytics use Last Non-Direct as the default?
Because Direct visits usually aren't a marketing channel, they're someone typing your URL or clicking a bookmark. Crediting the marketing touch before the Direct visit is more useful than crediting Direct itself. SourceLoop defaults to Last Non-Direct for the same reason.
### Can I look at one campaign under multiple attribution models at the same time?
Yes. SourceLoop's model selector supports stacking multiple models in the same view, so you'll see Last Touch credit alongside First Touch credit (or however many others you add) for the same campaign in the same table. The gap between columns is the assist value.
---
# How to track lead source in Web Form submission
Capture the marketing source and pre-conversion journey behind every form submission. Automated for embedded forms, semi-automated for hosted form tools.
Source: https://sourceloop.ai/help/track-web-form-submissions/
Updated: 2026-05-29
---
Web forms are the most common conversion type SourceLoop captures. Once the tracker is on your site, **every form submission is enriched with marketing source, UTMs, landing page, and the visitor's full pre-submit journey**, so when a lead lands in your dashboard you immediately see what brought them in.
This article is the overview. For setup specifics on individual form tools, the left sidebar has a per-tool article for every supported builder.
## Step 1: Install the SourceLoop tracking script
The tracker is the prerequisite for both capture paths below. It records every visitor's session, UTMs, landing page, and journey, so when a form submission arrives (directly from the browser or later via a webhook), SourceLoop can stitch it to the right visitor.
1. Sign in to [SourceLoop](https://app.sourceloop.ai/).
2. Open **Setup -> Tracking code** in the left sidebar.
3. Copy the snippet shown.
4. Paste it into the `` of every page on your site, especially the pages where forms (or the buttons that open hosted forms) appear.
Once the tracker is live, head to the matching section below depending on whether your form is embedded on your own site (automated) or hosted on the form tool's domain (semi-automated).
## How form capture works
SourceLoop supports two capture paths, depending on where your form actually lives:
| Path | What it means | Setup |
|---|---|---|
| **Automated** | The form is embedded on a page that loads the SourceLoop tracker, the browser fires the submit, the tracker reads the email field. | Install the tracker. That's it. |
| **Semi-automated** | The form is hosted off-site (e.g., `typeform.com/...`, `jotform.co/...`) so the tracker never sees the submit. The form tool forwards each submission to SourceLoop via a webhook. | Install the tracker, add SourceLoop's webhook URL inside the form tool, pass the visitor's identifier through the form URL. |
Both paths produce the same end result: a lead in the **Contacts Hub** with full first-touch and last-touch attribution, the converting landing page, and the entire pre-submit session timeline.
### Why two paths
Forms hosted on your own domain (HTML, Webflow, Framer, every WordPress plugin) submit through the browser. The SourceLoop tracker is already running on that page, so it can read the email field directly as the submission fires. No extra wiring needed.
Forms hosted on the form tool's own domain (Typeform's hosted page, JotForm's hosted page, Fillout's hosted page) submit on a different origin. The visitor leaves your site before the form fires its submit event, so the tracker never sees it.
The semi-automated webhook fills that gap. Before the visitor leaves your site to open the hosted form, the tracker quietly passes the visitor's anonymous identifier along on the form URL. When the visitor submits, the form tool sends a webhook to SourceLoop with the submission data plus that identifier. SourceLoop stitches the submission back to the visitor's pre-submit journey server-side, so the lead arrives with the same full attribution as an automated capture would have produced.
## Automated capture: supported tools
Tracker on the page is the only setup. Pick the matching article for the small confirmation steps.
[
HTML forms](/help/track-lead-source-in-html-forms/)
[
HubSpot Forms](/help/track-lead-source-in-hubspot-forms/)
[
Webflow](/help/track-lead-source-in-webflow/)
[
Framer](/help/track-lead-source-in-framer/)
[
Tally](/help/track-lead-source-in-tally/)
[
Elementor](/help/track-lead-source-in-elementor/)
[
Gravity Forms](/help/track-lead-source-in-gravity-forms/)
[
WPForms](/help/track-lead-source-in-wpforms/)
[
Contact Form 7](/help/track-lead-source-in-contact-form-7/)
[
Fluent Forms](/help/track-lead-source-in-fluent-forms/)
[
Ninja Forms](/help/track-lead-source-in-ninja-forms/)
[
Forminator](/help/track-lead-source-in-forminator-forms/)
[
Everest Forms](/help/track-lead-source-in-everest-forms/)
[
Formidable Forms](/help/track-lead-source-in-formidable-forms/)
[
JetFormBuilder](/help/track-lead-source-in-jetformbuilder/)
[
Kali Forms](/help/track-lead-source-in-kali-forms/)
[
MetForm](/help/track-lead-source-in-metform/)
[
SureForms](/help/track-lead-source-in-sureforms/)
[
HappyForms](/help/track-lead-source-in-happyforms/)
[
WS Form](/help/track-lead-source-in-ws-form/)
[
weForms](/help/track-lead-source-in-weforms/)
[
Cognito Forms](/help/track-lead-source-in-cognito-form/)
[
ConvertKit](/help/track-lead-source-in-convertkit/)
[
Klaviyo](/help/track-lead-source-in-klaviyo/)
[
Omnisend](/help/track-lead-source-in-omnisend/)
[
Mailchimp for WP](/help/track-lead-source-in-mailchimp-for-wp/)
[
AidaForm](/help/track-lead-source-in-aidaform/)
[
ARForms](/help/track-lead-source-in-arforms/)
[
GoHighLevel](/help/track-lead-source-in-gohighlevel-form/)
[
involve.me](/help/track-lead-source-in-involve-me/)