The way buyers research and evaluate vendors has changed. They're not running discrete searches; they're having extended advisory conversations with AI tools — ChatGPT, Claude, Gemini, Perplexity, coding assistants, search-enhanced AI — across many sessions, often weeks, frequently with multiple people on the buying committee using AI independently. The AI is acting as a guide and an advisor across the full journey: research → category framing → vendor shortlist → comparison → objection-handling → procurement.
By the time a prospect reaches your sales team, the AI has often already shaped which vendors made their shortlist, what they believe about your category, and which specific objections they're carrying in.
The single most impactful step you can take is to tag every incoming lead as AI-influenced or not. From there you can compare every downstream outcome — conversion to opportunity, opportunity-to-close, deal size, cycle time — between the AI-influenced cohort and the rest. Without that baseline, you can't measure how AI is shifting your funnel. The funnel is shifting quickly as AI plays an increasingly important role in the buying journey.
There are three steps that build on each other. Step 1 is the most impactful and the easiest to land. Steps 2 and 3 deepen the picture.
One question added to your existing lead form gets you the cohort. Every incoming lead lands tagged as AI-influenced or not, and from there every downstream metric is comparable across cohorts.
The key move: add the question to the form itself. Every submitter sees it, so you get a complete signal on every inbound lead. If form-completion drop-off is a serious concern for your team, the question can instead live on the confirmation or thank-you step that appears right after submission — the lead is captured first, and the AI-influence answer is layered on top.
Did AI play a role in how you found or evaluated us?
This is the same wording we use on the Unusual website. Three options is better than two — privacy-conscious buyers get a graceful out without skewing the data.
Optional progressive disclosure, only shown if "Yes":
Which tool(s)? (multi-select, skippable)
Two complementary moves:
Add AI as a value in your existing Lead Source field. A meaningful share of inbound now comes directly from AI tools — buyers who clicked a citation in ChatGPT, Claude, Perplexity, or Gemini and landed on your site through that AI tool. That's a channel.
Add AI Influenced as a second attribution dimension. AI's influence often shows up on top of the channel a buyer used to reach you — someone who came in via SEO can still be heavily AI-influenced, because they asked Claude for vendor recommendations before searching your name on Google.
| Field | Values |
|---|---|
Lead Source | SEO / Email / Paid Social / Events / Direct / AI / etc. |
AI Influenced (new) | Yes / No / I'd rather not say |
Capturing both means every Lead Source (including the new "AI" channel) can also be sliced by AI-influence status. The strongest reading of the data is usually "AI is a multiplier on the channels you already trust, and increasingly a channel of its own."
AI influence happens at the individual stakeholder level, not just the deal level. One person on the buying committee may have used AI heavily; another may not have used it at all. The post-submission question captures attribution at the contact level on inbound. Downstream, the contact-level tags roll up into deal-level cohort flags — a deal is AI-influenced if any stakeholder is tagged AI-influenced.
Our default recommendation is on the form itself. Every submitter sees the question, which gives you a complete cohort signal on every inbound lead. A required (or near-required) field on the form produces higher-fidelity data than a partial signal collected after submission, where many buyers click away before answering.
The tradeoff is real: every additional field on a marketing form reduces completion. For most teams, the value of complete AI-attribution data outweighs the small dip in form completion — this is the question that tells you how the funnel is shifting, and the signal compounds over months.
For teams where conversion drop-off is a serious concern — for example, teams that have already minimized the form to the bare essentials and removed classic questions like "How did you find us?" for that reason — the AI-influence question can live on a second screen that appears right after submission. The lead is captured first; the AI-influence answer is layered on the confirmation or thank-you step the buyer is already looking at. Enough buyers answer to build the cohort, and form completion stays untouched. This is the safer placement when every percentage point of form-completion matters.
We're at the very early innings of AI attribution. AI's influence on the buying funnel is rising — and it will keep rising. AI is moving from research, to action on behalf of the buyer (booking demos, summarizing options), to eventually purchase and product use. Companies that don't establish a baseline now lose their ability to understand and adapt to a buying motion that's getting more AI-mediated by the quarter. Capturing the signal now is the lowest-friction way to establish that baseline.
Once the post-submission question is live, the second step deepens the cohort: have your sales transcript data analyzed by your own AI tools (Claude, ChatGPT, an internal AI agent) to tag the contacts and opportunities where AI use comes up during the sales conversation. Many deals will come in without an AI flag from the confirmation step — the buyer skipped the question — but a stakeholder later mentions during a discovery call: "Yeah, I asked Claude about this." You want that signal too.
If your GTM team has access to the raw sales transcript data — via Granola, Otter, Avoma, Fireflies, Chorus, or an internal notetaker — your AI tools can read the transcripts directly and produce a per-call tag. This is the highest-fidelity option. Your Unusual team can share their best practices for using AI tools (Claude, ChatGPT) to surface AI-influenced mentions in sales transcripts — the prompts and the workflow that travel directly back to this option.
If you're on Gong, AI-research mentions can be added as a custom signal directly inside the Gong UI. Similar add-on signals are available in other sales-intelligence platforms with comparable extensibility — useful for teams that don't have direct transcript access but want a tagged signal layered into the workflow they already use.
A note on data residency: all analysis happens within your own system. Unusual provides platform-side data and the analysis patterns; your GTM team's AI tools read your transcript data inside your environment. Nothing crosses the boundary.
The third step is to brief your sales team to start asking buyers about AI use as part of their evaluation, around close or post-close when trust has been built.
The briefing artifact for your sales team is on the hub: When your buyers ask AI first. It's designed to be shared with your AEs and CSMs as a standalone page. Forward it directly — it gives them the context they need on what's happening in their buyers' heads, what your company is doing about it, and the one question to ask.
This layer supports and reinforces the transcript analysis. The buyer's own words about which AI tool shaped their thinking — captured intentionally, after trust has been built — are the highest-fidelity attribution signal you can pair with the transcript layer's organic mentions.
Once Steps 1, 2, and 3 are in motion, you'll start to see the cohort take shape. One to three months after the post-submission question goes live is when there's typically enough cohort data to build meaningful comparison reports — AI-influenced vs not — on conversion rates, deal sizes, and cycle times. At that point, your Unusual team will lay the cohort data next to the model-side analysis we're producing on your account, and the attribution loop closes.