Measurement
How to Track AI Referral Traffic in GA4 (2026): The AI Assistant Channel, a Tested Regex and a Dashboard Block
To track AI referral traffic in GA4, start with the AI Assistant channel that Google added to the Default Channel Group in May 2026, then build one custom channel group that combines it with a tested source regex for the assistants Google’s list does not name. That gives you a single AI row in every acquisition report, including your history, in about 15 minutes.
This guide covers what the native channel does and does not catch, a referrer table and regex current to October 2026 (with the test output), the step-by-step channel group setup, a reporting block for Looker Studio (now called Data Studio), and why consent mode makes European AI numbers look smaller than they are.
Key takeaways
- Since 13 May 2026, GA4 assigns the medium
ai-assistant, the campaign(ai-assistant)and the AI Assistant channel when the referrer matches Google’s list of AI assistants. It needs no setup, and it excludes AI Overviews and AI Mode, which stay in Organic Search. - Google’s documentation names ChatGPT, Gemini, Claude, Deepseek, Copilot and Grok as examples, but not Perplexity. Check where your
perplexity.aisessions land before you trust the native row. - A custom channel group whose AI channel matches Default channel group = AI Assistant OR Source matches your regex, placed above Referral, catches both and applies retroactively.
- Google’s own example regex for AI assistants also matches every
.aidomain,platform.openai.comand the Gemini crypto exchange. The anchored pattern in this guide does not. - Visits with no referrer (native apps, copied links) still land in Direct, and visitors who decline consent in the EEA send a referrer but no session. Both are undercounts you report as ranges, not fixes you configure.
What is AI referral traffic?
AI referral traffic is website traffic that originates from AI assistants such as ChatGPT, Perplexity, Gemini, Claude and Microsoft Copilot, when a user clicks a cited link inside an AI-generated answer. Unlike standard organic search, it arrives with the visitor’s intent partly formed inside the AI conversation.
Until May 2026 none of it had its own row in GA4’s default reports: it was spread across Referral, Direct and Unassigned. Google’s native channel fixes part of that. The rest, which assistants it recognises, what happens to links without a referrer and how consent changes the count, is what the sections below are for. The scale of the gap on the Google side, where AI answers are counted as Organic Search, is measured in our six-site study of AI Mode impressions.
GA4’s AI Assistant channel (May 2026): what it does and does not do
Google announced the change in the 13 May 2026 entry of What’s new in Google Analytics. It describes a new AI Assistant channel for “chatbots like ChatGPT, Gemini, and Claude” and three changes to traffic-source dimensions, all applied automatically.
How GA4 classifies an AI visit now
The Default channel group definitions spell out the mechanics. When the referrer matches Google’s list of AI assistants, GA4 sets the medium to ai-assistant and the campaign to (ai-assistant), and the AI Assistant channel is simply “the medium exactly matches ai-assistant”. The same page describes the channel as users arriving “from sources like ChatGPT, Gemini, Deepseek, Copilot, or Grok” and says it “excludes Google’s AI Overviews and AI Mode”. Those two are named in the Organic Search definition instead.
| What you see | Before May 2026 | After the AI Assistant channel |
|---|---|---|
| Session medium for a recognised assistant | referral | ai-assistant |
| Session campaign | Whatever the source carried | (ai-assistant) |
| Session default channel group | Referral | AI Assistant |
| AI Overviews and AI Mode clicks | Organic Search | Organic Search (unchanged) |
| Assistant clicks with no referrer | Direct | Direct (unchanged) |
Two practical consequences follow. Any report or dashboard filter you built on medium = referral to find AI sources will quietly lose those sessions after the switch, so rebuild it on the channel or on Session source. And any year-on-year comparison that crosses 13 May compares two classification rules, so compare on Session source, which did not change, rather than on channel.
What the AI Assistant channel still misses
Perplexity is not in Google’s named examples. Google’s default channel page names ChatGPT, Gemini, Deepseek, Copilot and Grok; the announcement names ChatGPT, Gemini and Claude. Perplexity appears in neither list, although Google’s own custom channel group guidance lists it among “popular assistants” to put in a regex. Google links a downloadable list of sources from the default channel page; until you have checked your own property, treat Perplexity as possibly still sitting in Referral. The quickest check is in the Perplexity section below.
Anything without a referrer stays in Direct. The channel keys off the referrer, so a click from a native app, a link copied out of a chat, or a browser that strips the header arrives with nothing to match. Search Engine Journal’s report on the launch makes the same point: the new channel only captures what GA4 can identify through the referrer. Recovering those visits is a different method, covered in AI traffic attribution when the referrer is gone.
ChatGPT’s campaign tag can take a different path. OpenAI’s publisher FAQ states that ChatGPT “automatically includes the UTM parameter utm_source=chatgpt.com in referral URLs”. A UTM-tagged click carries a campaign source, not just a referrer, so check how those sessions resolve in your own property: open Traffic acquisition with Session source / medium as the dimension and look for chatgpt.com rows with a medium other than ai-assistant. If you find chatgpt.com / (not set) sitting in Unassigned, that is ChatGPT traffic the native channel did not claim, and the custom group below picks it up by source.
Google AI surfaces are deliberately out. AI Overviews and AI Mode clicks are Organic Search by definition. Measuring those is a Search Console job, covered in how to track AI Overviews.
AI referrer and regex table (October 2026)
The native channel is the floor. For a complete AI row you still want a source regex, and it needs to be precise, because a loose pattern pulls in unrelated referrals and inflates the channel you are trying to defend.
| Assistant | Session source value(s) to match | Named in GA4’s AI Assistant docs? | Notes |
|---|---|---|---|
| ChatGPT | chatgpt.com, legacy chat.openai.com | Yes | Links carry utm_source=chatgpt.com (OpenAI) |
| Perplexity | perplexity.ai, www.perplexity.ai | No (only in Google’s custom-group example) | Check which channel it lands in on your property |
| Claude | claude.ai | Yes (announcement) | |
| Gemini | gemini.google.com, gemini | Yes | gemini covers tagged links that use it as the source |
| Microsoft Copilot | copilot.microsoft.com, copilot.com | Yes | Check for both hostnames |
| DeepSeek | chat.deepseek.com | Yes | |
| Grok | grok.com | Yes | Do not add x.com, which is all of X |
| Meta AI | meta.ai, www.meta.ai | No | |
| Mistral Le Chat | chat.mistral.ai | No | |
| You.com | you.com | No |
The hostnames are the assistants’ product domains. Confirm each one against your own Session source list before you rely on it, because what a property actually receives depends on the surface and the month. You do not need the Bard patterns in older examples; Gemini’s web app is at gemini.google.com.
The pattern that matches every row in the table, with or without a www. or other subdomain prefix:
^(.*\.)?(chatgpt\.com|chat\.openai\.com|perplexity\.ai|claude\.ai|gemini\.google\.com|gemini|copilot\.microsoft\.com|copilot\.com|chat\.deepseek\.com|grok\.com|meta\.ai|chat\.mistral\.ai|you\.com)$
It is anchored at both ends on purpose. GA4’s regex documentation says that “by default, Google Analytics Regex is full regex”, so an unanchored chatgpt\.com never matches www.chatgpt.com, and a pattern written for partial matching behaves differently in GA4 than in a regex tester. Writing the anchors yourself makes the pattern behave the same under both rules.
Why not use Google’s example regex?
Google’s custom channel group article gives an example pattern that begins ^.*ai|.*\.openai.*|.*chatgpt.*|.*gemini.*|.*gpt.*|.... The first alternative, ^.*ai, matches any source ending in “ai”. I ran both patterns in Python against a list of real and near-miss source values, using full-match semantics to mirror GA4’s default (the syntax used is valid in both Python and Google’s RE2):
| Session source | Google example | This guide |
|---|---|---|
chatgpt.com | match | match |
www.perplexity.ai | match | match |
chat.deepseek.com | no match | match |
grok.com | no match | match |
you.com | no match | match |
jasper.ai | match | no match |
character.ai | match | no match |
platform.openai.com | match | no match |
business.gemini.com | match | no match |
gptzero.me | match | no match |
The false positives are the problem. jasper.ai is a writing tool, platform.openai.com is developer documentation, and gemini.com is a crypto exchange; none of them is an assistant citing your page. On a B2B site with a few hundred AI sessions a month, a handful of those referrals is enough to move the conversion rate you report. Review the pattern each quarter and add hostnames as they appear in your Referral list, rather than widening it with wildcards.
How to track AI referral traffic in GA4: step by step
This builds one custom channel group that keeps everything from GA4’s default group, including the new AI Assistant channel, and adds a single AI channel that catches both the native classification and your regex. Standard properties can have two custom channel groups in addition to the default, so this uses one of your two slots.
Step 1: Check the native channel first
Go to Reports → Acquisition → Traffic acquisition and look for an AI Assistant row in Session default channel group. Then switch the primary dimension to Session source / medium and search for perplexity, chatgpt and copilot. Note which rows show ai-assistant as the medium and which still show referral or (not set). That list is what the custom group needs to fix.
Step 2: Create a new channel group
Go to Admin → Data display → Channel groups and click Create new channel group. Google’s documentation explains that a new group starts as a copy of the default group, which is what you want: every standard channel, including AI Assistant, is already in it. Name it something clear such as “Channels with AI (2026)”. You need Editor access or above.
Step 3: Add one AI channel with two conditions
Click Add new channel, name it AI Search, and add two condition groups joined with OR:
- Default channel group exactly matches
AI Assistant - Source matches regex, with the pattern from the table section above
Default channel group is one of the fields custom channel groups support, which is what makes this combined rule possible. The first condition inherits every assistant Google adds to its list in future; the second covers the ones it has not added yet and the tagged ChatGPT sessions that did not resolve to ai-assistant.
Step 4: Reorder above Referral
Click Reorder and drag AI Search above Referral, and above Unassigned and Organic Search as well. Traffic “is included in the first channel whose definition it matches”, so an AI channel sitting below Referral catches nothing that Referral already claimed. This is the most common reason a custom AI channel shows zero.
Step 5: Save and apply
Click Save group. Google states that custom channel groups can be applied to your reports retroactively, so the AI Search row covers historical sessions as well as new ones. That is the practical advantage over the native channel for trend reporting: the custom group applies one rule to your whole history, while the default channel changed its rule on 13 May 2026.
/assets/screenshots/ga4-ai-referral-channel-group.png Step 6: Read it in Traffic acquisition
Go to Reports → Acquisition → Traffic acquisition and switch the channel group selector at the top of the table from the default group to “Channels with AI (2026)”. You should now see an AI Search row with sessions, engagement and key events.
/assets/screenshots/ga4-referral-exploration.png If the row is empty, that does not mean you have no AI traffic. It means nothing in the date range arrived with a referrer or source that matches. Check the default Referral channel manually for chatgpt.com and perplexity.ai entries, and check the channel order from Step 4.
Step 7: Break it down in Explore
For ad-hoc analysis, go to Explore → Free-form exploration and set:
- Dimensions: Session source / medium, Landing page + query string
- Filter: Session source matches regex, with the same pattern
- Metrics: Sessions, Engaged sessions, Key events, Engagement rate
This isolates AI sources without touching the channel group and lets you slice by device, country or date range freely.
Step 8: Add landing page as a secondary dimension
In Traffic acquisition (or your exploration), add Landing page as a secondary dimension. This tells you which specific pages AI assistants are citing, which is the most actionable finding in the whole exercise. If AI is sending traffic to your blog posts but not to your services or pricing pages, that is a content gap, not a tracking gap.
Build an AI traffic block in Looker Studio (now Data Studio)
Most teams do not read AI referrals in GA4. They read them in a dashboard. Google renamed the product: in April 2026 “the Looker Studio product name returned to its original name: Data Studio”, at datastudio.google.com. Existing reports carry over, and everything below works the same under either name.
If you already run an SEO dashboard, add the AI block to it rather than building a separate report. The base setup, connecting Search Console and GA4 and choosing which charts earn their place, is in the Looker Studio SEO dashboard guide. The AI block is four charts on the same GA4 data source:
- Scorecard: Sessions, filtered to your AI channel, with comparison to the previous period.
- Time series: Sessions by week, broken down by Session source, filtered to AI. This is the growth chart stakeholders ask for.
- Table: Landing page × Session source, with Sessions, Engaged sessions and Key events. This is the citation map.
- Bar chart: Key event rate for the AI channel against Organic Search, same date range.
There are two ways to get the AI filter into the dashboard.
Use the custom channel group. The Data Studio GA4 connector exposes the fields available in the GA4 Data API, “including any custom fields that are defined for the property”. If the group does not appear after you create it, Two Octobers’ walkthrough notes that you need to edit the data source connection and click Refresh Fields before the new channel group shows up.
Use a calculated field. When you report on several properties, or you have used both custom-group slots, define the AI rule once in the data source instead:
CASE
WHEN Session default channel group = "AI Assistant" THEN "AI Search"
WHEN REGEXP_MATCH(Session source, "^(.*\\.)?(chatgpt\\.com|chat\\.openai\\.com|perplexity\\.ai|claude\\.ai|gemini\\.google\\.com|gemini|copilot\\.microsoft\\.com|copilot\\.com|chat\\.deepseek\\.com|grok\\.com|meta\\.ai|chat\\.mistral\\.ai|you\\.com)$") THEN "AI Search"
ELSE Session default channel group
END
Data Studio’s REGEXP_MATCH “returns true if the target value exactly matches the regular expression pattern”, the same full-match behaviour as GA4, so the anchored pattern carries over unchanged. The same page notes that patterns containing \ “may require additional escaping”, which is why every backslash is doubled inside the quoted string above. Two connector limits to know: GA4 segments and comparisons are not available through it, and when a table is filtered by a calculated field the grand total shows null. Filter the AI table on the channel group where you can, and keep the calculated field for charts without a summary row.
Consent mode: why European AI numbers look too small
If your site serves visitors in the EEA or UK, consent is the second leak after missing referrers, and it hits the AI channel harder than any other because the channel is small.
Google’s consent mode documentation states that when visitors deny consent, tags send cookieless pings instead of storing cookies, and that those pings can include the timestamp, the user agent and the referrer. So an AI-referred landing still reaches GA4 with its source attached. What it does not carry is an identifier, which means the session, the next pageview and the conversion cannot be joined to it. The same page is explicit that a basic implementation, where tags are blocked until consent, gets no modelled data at all.
In practice that gives you three rules for AI reporting in Europe:
- Implement consent mode in the advanced pattern, so denied visitors still produce landing pings with a referrer.
- Report the AI conversion rate from consented AI sessions only, and show the consent rate for the channel next to it.
- Treat modelled AI sessions as an estimate whose distribution by source Google does not document.
The detail, including who qualifies for behavioural modelling and what it does to AI rows, is in Consent Mode v2 and AI referral traffic.
Why it’s worth tracking even when the volume looks small
The objection most teams raise: “It’s less than 1% of our sessions. Why does it matter?” Three reasons.
Concentration on commercial pages. AI traffic is not distributed evenly across a site; it concentrates on the pages that matter most for revenue. The Previsible 2025 State of AI Discovery Report, which analysed over 1.96 million LLM-driven sessions across sites in SaaS, e-commerce, finance, legal, health, and publishing, found that AI traffic concentrates on industry pages (1.14% AI penetration), tools pages (0.95%), and pricing pages (0.46%), all 4–9× higher than the 0.13% site-wide average. These are not awareness visitors. They arrive with a defined problem and, in many cases, a vendor shortlist already formed inside the AI conversation.
Conversion rate vs organic. The conversion data is the strongest argument for the channel, and the thinnest. Seer Interactive measured one client between October 2024 and April 2025: ChatGPT referral traffic converted at 15.9% and Perplexity at 10.5%, against 1.76% for Google Organic on the same site. Read the arms before the ratio: roughly 11,000 AI sessions against roughly 14 million from Google. Seer’s own title is “6 Learnings, 1 site”. It is a case study, not a benchmark. The second measurement is weaker than it looks. Rankability’s analysis of Microsoft Clarity data across 1,200 websites found AI platform visitors converting to sign-ups at 1.66% versus 0.15% from traditional search, an 11× difference. That figure is a third party’s reading of Clarity’s data, not Clarity’s own published finding: Microsoft Clarity itself puts AI-referred conversion at three times other channels, not eleven. Treat 11× as the outer edge of a range whose publisher reports 3×. The direction is the part both agree on: AI-referred traffic converts above organic.
Early data is the actionable data. When volume is small, the channel mix and landing-page distribution are read cleanly. The teams who build the strongest GEO presence are not the ones who wait until AI traffic is large; they are the ones measuring now, while the data is thin enough to act on.
What to look for once you have the data
Once your AI channel is live and populated, four things are worth tracking.
1. Volume trend, not absolute volume. The absolute number will likely be small. What matters is whether it grows month on month. Flat or declining AI referrals on a growing site is a GEO signal: your content is not being cited.
2. Which pages are getting cited. The landing page dimension shows which URLs appear in AI answers often enough to earn clicks. Cross-reference them against your key commercial pages. If AI sends traffic to blog posts but not to services or pricing pages, your commercial pages are not structured in a way that makes AI engines confident enough to recommend them.
3. Conversion rate versus organic. Compare the AI channel’s key event rate against Organic Search in the same date range. If it is already higher, even at tiny volume, the citation strategy is working. If it is lower, the usual cause is that the AI is citing informational content and visitors arriving on blog posts have no clear next step.
4. Which platforms are sending traffic. The citation pool is largely separate across platforms: Ahrefs analysis of 15,000 prompts found that only 12% of AI-cited URLs rank in Google’s top 10 on average, and the platforms cite mostly different content. Perplexity is the outlier with 28.6% Google top-10 overlap; the rest sit around 8%. Those figures are long-tail prompts on standalone assistants rather than AI Overviews, where the same dataset put overlap at 76%. If you only get ChatGPT referrals and nothing from Perplexity, your content may not be structured for Perplexity’s citation patterns, which favour definition-first, citation-dense content.
Perplexity and Copilot referral traffic in GA4
Perplexity is the assistant most worth checking by hand, because it is the one Google’s AI Assistant documentation does not name.
The fastest check (no setup). Go to Reports → Acquisition → Traffic acquisition, switch the primary dimension to Session source / medium, and type perplexity into the table’s search box. If Perplexity has sent traffic in your date range, you will see perplexity.ai with a medium of either ai-assistant (the native channel claimed it) or referral (it did not). Repeat with copilot for copilot.microsoft.com and copilot.com. That one search tells you whether you need the custom group at all.
Isolate both in an Explore report. For a reusable view, go to Explore → Free-form and set:
- Dimension: Session source / medium, with Landing page as a second row to see which pages each platform cites
- Filter: Session source matches regex
^(.*\.)?(perplexity\.ai|copilot\.microsoft\.com|copilot\.com)$ - Metrics: Sessions, Engaged sessions, Key events, Engagement rate
Why Perplexity clicks are worth isolating. A click from an assistant’s citation means someone read a summary that named your site and chose to visit anyway. In Seer Interactive’s single-client study, Perplexity converted at 10.5% against 1.76% for Google Organic on that single client, whose AI arm was roughly 11,000 sessions against 14 million for Google.
If you get zero Perplexity referrals, check that PerplexityBot can crawl your site (security plugins and WAF rules can block AI crawlers without you noticing), then check whether your content structure matches what Perplexity’s retrieval rewards. The full citation strategy is in How to Get Cited by Perplexity (And Track Your Citations).
Referral tracking cannot tell you how often an assistant names your brand without sending a click. To see mentions and citations inside Perplexity and ChatGPT answers directly, pair GA4 with a citation check; how to see whether your business appears in ChatGPT and Perplexity walks through monitoring that presence.
What low AI referral numbers actually mean
If you set up the tracking and find near-zero AI referral traffic, check the measurement before the content. Two causes sit outside GA4’s configuration: assistants that strip the referrer, handled in attributing conversions from AI assistants, and European visitors who decline consent, handled in the consent section above. A third is inside it: an AI channel placed below Referral in the channel order.
If the measurement is sound, there are three likely content causes, each with a different fix.
Cause 1: Your content is not structured for citation. AI engines consistently cite content with clear definitions, direct answers to specific questions, and structured formatting. A page of long flowing prose with no subheadings is hard to extract a quotable answer from. The fix is restructuring, not more content. For the schema layer, see Structured Data for AI Search; if you publish video, how video content earns AI citations covers the transcript and metadata signals that turn YouTube into an AI referral source.
Cause 2: Your entity is not established. AI engines are more likely to cite sources they recognise as authorities on a topic. If your domain has limited external mentions and no consistent entity signals across the web, your content may be accurate but unverifiable from the model’s perspective. See Why Entities Matter More Than Keywords for AI Search for the framework.
Cause 3: You are not targeting the right queries. AI traffic concentrates on decision and comparison queries, not awareness queries. The Previsible analysis found pricing and tools pages receive 4–9× higher AI presence than site averages, because AI users arrive with defined problems and use AI to compare solutions. A content plan that is mostly top-of-funnel awareness pieces is optimising for a channel AI does not drive heavily.
Three things to do this week, depending on what you find
If you find AI traffic and it is converting well: document which pages are being cited, analyse their structure, and replicate that structure across your highest-value commercial pages. The pages already getting cited are your GEO templates.
If you find AI traffic and it is not converting: the issue is CTA misalignment. The cited pages are likely informational: they answer a question but offer no clear next step. Add a contextual CTA to each cited page that reflects where an AI-referred visitor already is. They understand the problem, so skip the awareness pitch.
If you find very little or no AI traffic: run your priority pages through a GEO readiness check. Does each page have a definition block in the opening paragraph? Does it directly answer the question implied by its primary keyword? Are there FAQ questions in natural language? The AEO Article Analyzer scores any article against 10 citation-readiness criteria, with a 0–100 score, pass or fail per criterion and the top three fixes.
The measurement gap is the strategy gap
Google’s native channel means every GA4 property now shows some AI traffic without configuration. It does not mean the number is complete. The custom group in this guide takes 15 minutes and closes the gaps you can close; the referrer and consent gaps have to be reported as ranges.
What the setup surfaces, which pages AI engines already cite, at what conversion rate and from which platforms, is the most actionable signal available in search right now. The zero-click search data explains why so much AI exposure never produces a click, and the B2B AI search buying-journey breakdown covers how those un-clicked touchpoints still move deals. Wiring all of it into one attribution model is what full-funnel tracking is built to do.
Frequently asked questions
Does GA4 automatically track AI traffic?
Partly, since 13 May 2026. GA4 now assigns the medium ai-assistant and the AI Assistant channel when the referrer matches Google’s list of assistants, which Google’s documentation illustrates with ChatGPT, Gemini, Claude, Deepseek, Copilot and Grok. It does not catch visits with no referrer, it excludes AI Overviews and AI Mode, and it may not cover every assistant you receive traffic from. A custom channel group combining the native channel with a source regex closes most of that gap.
Does the GA4 AI Assistant channel include Perplexity?
Google’s documentation does not name Perplexity among the example assistants for the channel, although its custom channel group guidance lists Perplexity as an assistant to include in a regex. Check your own property: in Traffic acquisition, set the dimension to Session source / medium and search for perplexity. If the medium is referral, add Perplexity to a custom channel group as described in this guide.
Why does some ChatGPT traffic show up as Direct in GA4?
Referrer data is set by the browser, so clicks from ChatGPT’s native apps, links copied out of a chat and browsers that strip the header arrive with nothing for GA4 to classify. OpenAI says ChatGPT adds utm_source=chatgpt.com to referral URLs, which helps where the parameter survives, but any redirect that drops the query string loses it. Measured ChatGPT traffic is therefore an undercount of real ChatGPT visits.
How do I check Perplexity referral traffic in GA4?
Open Reports → Acquisition → Traffic acquisition, set the primary dimension to Session source / medium, and type perplexity into the table search box. perplexity.ai rows show sessions and key events, and the medium tells you whether the native AI Assistant channel claimed them. For a reusable view, build a free-form exploration filtered to Session source matching ^(.*\.)?perplexity\.ai$ and add Landing page to see which pages Perplexity cites.
Can I track AI Overview clicks in GA4?
Not separately. Google’s default channel definitions place clicks from AI Overviews and AI Mode in Organic Search and explicitly exclude them from the AI Assistant channel. AI referral traffic from standalone assistants arrives through their own domains; AI Overview traffic arrives through Google Search and is indistinguishable from other organic clicks in GA4. Use Search Console for the Google side, as set out in how to track AI Overviews.
What is a good AI referral traffic conversion rate?
There is no published benchmark worth planning against, only two narrow measurements. Seer Interactive’s single client recorded ChatGPT at 15.9% and Perplexity at 10.5% against 1.76% for Google Organic, on an AI arm of roughly 11,000 sessions against 14 million from Google. Rankability’s analysis of Microsoft Clarity data across 1,200 sites recorded AI sign-up conversion at 1.66% against 0.15%, a different action in a different industry. What survives both is a direction, not a number: AI-referred traffic converts above organic, by a multiple that varies with ICP, so measure your own.