Case Study · AI Prospecting

Know who's buying. Before they raise a hand.

How we designed Clay's prospecting intelligence landing page — turning a six-tab research problem into a page that sells one thing: a shorter, better list.

Prospecting for revenue teams at
Customer logos: Paperz, Dorfus, Martino, Squares, Galaxos
Client
Clay
Category
AI Prospecting Intelligence
Audience
Revenue & sales leaders
Focus
Demos & free trials
The Problem

Prospecting today is six tabs
and a guess.

A rep opens LinkedIn, a data vendor, the CRM, a news tab, the sequencer and a spreadsheet. Forty minutes later they have one email and no idea whether the account is even in market. The page had to make that felt — not explained.

Clay's 3D clay-rendered illustration of a tangled prospecting workflow
Research eats the week
0 hrs
Reps spend more time assembling context than talking to buyers.
Data rots quietly
0% / yr
Titles change, people leave, and nobody tells your CRM.
Signals arrive too late
0 days
By the time a funding round hits your list, three competitors already emailed.
Personalization doesn't scale
0%
Good first lines take fifteen minutes. Quotas don't allow fifteen minutes.
The Core Idea

Four moves, decided
before we opened Figma.

Every competitor sells volume — more contacts, more records, more coverage. A rep doesn't want more. They want fewer, better names and a reason to call each one. So the whole page argues for subtraction.

The hero promises knowing who's buying before they raise a hand. The final CTA promises a shorter, better list. Everything between them exists to prove that one claim.

Buyers burned by bad data don't believe accuracy claims. They believe a visible process — so we put the enrichment waterfall itself on the page.

Every generated line traces back to a real source — a job post, a filing, a launch. Citations are the only credible answer to "isn't this just AI slop?"

Sales tooling is dashboards and tables. A soft, tactile 3D clay world in violet and lime buys attention in a feed of identical SaaS pages.

The partner lockup panel from the Clay page
Design Craft

Everything a rep needed
a spreadsheet for.

Three claims a buyer won't take on faith — coverage, scoring and personalization. Each got its own block, its own colour, and a screenshot of the feature actually doing the work.

Watch a blank list become a pipeline.

One provider misses. Two providers disagree. So we showed the waterfall itself — 140+ sources running in sequence, each result checked against the last, with a confidence score attached.

  • 94% deliverability on verified emails, or the credit is refunded
  • Re-verifies every record quarterly so the CRM stops rotting
  • Credits pooled across the team, never per seat
The enrichment waterfall UI running across 140+ providers

A score your reps will actually trust.

Most scores are a black box with a number on the front. We surfaced the four inputs behind every score — so a rep can defend the call to their manager, and a manager can defend the forecast to the board.

  • Signal fit, reachability and timing — weighted, and visible
  • Re-scores hourly as new intent signals land
  • Writes back to Salesforce as a native field, not a note
The intent scoring UI showing the signals behind a score

Outreach that reads like a person wrote it.

The strongest anti-AI-slop argument on the page: every line in the draft is cited back to its source — a job post, a filing, a launch — so a rep can verify the claim before it goes out.

  • Every line is cited back to its source, so reps can verify it
  • Learns your team's voice from the threads that actually landed
  • Pushes to Outreach, Salesloft or Apollo — reps stay where they are
The enriched contact card and drafted outreach with cited sources

Show the product working, not sitting still.

Static screenshots show a tool. We framed the workspace mid-task — an agent running a list, checking off steps, pushing to the CRM — so the buyer sees outcomes arriving rather than an interface waiting for them.

  • Pull accounts matching the ICP, automatically
  • Enrich contacts through the waterfall, then draft outreach
  • Push everything back to the CRM — no copy-paste
The Clay workspace running an agent task list
Proof

Proof from reps,
not from marketing.

We placed testimonials from working sellers directly before the ROI block — the exact moment doubt peaks and price enters the buyer's head.

★★★★★

"My reps used to open the week with a list and a prayer. Now they open it with twelve accounts that did something last Thursday. That's the whole difference."

Dana MirelesVP Sales Development, Northwind
★★★★★

"The part I didn't expect: onboarding. New SDRs used to take a quarter to learn the territory. Sightline hands them the context on day one, so they're calling in week one."

Marcus FeldDirector of Revenue Ops, Arcline
★★★★★

"We replaced two data vendors and a very tired spreadsheet. The credit pool is per team, not per seat, and the scoring gave my managers something to coach against."

Priya RamanHead of Demand Gen, Vaultly
Clay's 3D clay illustration for the ROI section
The Numbers

"Move the sliders. It's your number, not ours."

Rather than publishing an ROI figure nobody believes, we let the buyer build it. Reps get roughly eleven hours a week back, and the accounts they call are pre-qualified — the math runs on their team, their numbers.

  • Payback in under two months for the median team
  • Retires spend on two or three point tools you're already paying for
Questions

The things buyers ask us
on the first call.

We asked Clay's team which five questions come up on every first call, then answered them in the buyer's own words. An FAQ isn't filler — it's the cheapest way to remove the objections that stall a deal.

An AI prospecting intelligence platform. It continuously researches your target accounts, enriches contact data, detects buying signals like hiring, funding and tech-stack changes, and drafts personalised outreach — then writes it all back to your CRM.

It watches public signals across job posts, filings, funding announcements, tech-stack changes and news, then scores each account on fit, reachability and timing — with the inputs behind every score visible to the rep.

Yes. It writes to Salesforce and HubSpot as native fields, and pushes drafted outreach to Outreach, Salesloft or Apollo, so reps keep working where they already work.

From 140+ licensed providers and public sources, run as a waterfall so each result is checked against the last. Every record carries its source, and coverage is GDPR and CCPA compliant.

Credits are pooled across the team rather than charged per seat, so you're not paying for reps who prospect lightly. Most teams see payback in under two months.

What We Now Treat As Defaults

Five rules for AI product pages.

In a category selling "more," promising less is the sharpest position available.

Buyers burned by bad data believe a visible process, not an accuracy claim.

Sourced output is the only credible answer to "isn't this just AI slop?"

A number the buyer built themselves is a number they'll defend internally.

Distinct art direction is a conversion asset, not decoration.

Let's Build Yours

Got a product that's hard to explain? That's the fun part.

Design + build · No retainer required · Reply in one working day
Case Study Library · 9 Projects

Browse all nine case studies.

You're viewing one of nine landing-page projects. Click any card below to open a different case study.