Clay tested the demo forms of more than 6,000 companies. Agents found each form, submitted a realistic inquiry, watched for an email or phone response, timed the reply, and evaluated how personal it was.
The research could have ended as one benchmark report. Clay connected the dataset to a company lookup, tailored recommendations, and a follow-up workflow so each visitor could use the findings instead of only reading them.
Speed to lead: what 6,346 form fills revealed
Most teams treat inbound follow-up as a routing problem. The benchmark shows a larger gap between speed, coverage, and relevance.
The businesses we tested each had varying results. Let's see how they did.
Customer pain
Method
Findings
What to do next
Write the report
Shape the argument before the page
Before Alex Lindahl became Clay’s Creator in Residence, he spent years as a GTM engineer working directly with customers. One complaint kept surfacing: inbound leads were taking too long to reach the right person. When Alex got room to build his first lead-magnet experiment, that was the problem he chose.
Start with a problem your product can solve
Clay’s product is built for optimizing speed to lead. So, starting from a pain point that their prospects were facing that their product also solves for, was a clear fit. Clay’s internal process looks like this:
- An inbound system enriches and qualifies form submissions
- It routes a good-fit prospect to the right booking link
- Meetings appear directly on an account executive’s calendar without an SDR manually researching every lead
Alex heard the opposite process on customer calls. A form submission would wait for someone to notice it, research the company, decide who should respond, and write an email. Strong prospects waited. Poor-fit leads sometimes received attention first.
That led to a testable question: how quickly do B2B companies respond to a realistic demo request, how personal is the response, and did the company appear to recognize whether the prospect matched its ICP?
“One lead magnet could work really well for one industry. It could not work well for another industry.”
Alex Lindahl, Creator in Residence at Clay
To find what can work for your company, follow the methodology Alex shared:
- List the problems customers repeat in calls
- Pick one your product already helps solve
- Turn it into a question you can test with evidence
- Decide what result would help a buyer take the next step
An interactive experience can encourage prospects to explore the result, but it cannot make the wrong strategy useful. Start with a pain point customers actually raise, then build an exceptional experience around it.
How Clay tested thousands of demo forms
Clay started with more than 10,000 tier-one and tier-two accounts. The team used several specialized agents to test companies with suitable forms.
- Claygent found the form. It visited each domain and returned the correct demo or contact URL.
- AgentMail created a monitored inbox. When a company replied, the event returned to Clay with a timestamp.
- Saperly provided a phone number. A callback created the same kind of timed event.
- A Subconscious agent submitted the inquiry. It used a consistent small-software-company persona so the tests could be compared.
- Clay analyzed the response. The workflow measured elapsed time, classified the reply’s personalization, and checked whether the test company fit the target account’s stated ICP.
The ICP check gave Clay’s GTM team more than a response-time score. Each company record could show how quickly the team replied, how personal the response was, and whether the test persona appeared to match its ICP.
That creates three different starting points for a conversation:
- If the company replied, a rep can discuss its speed, personalization, and qualification process
- If Clay tested the company but received no reply, a rep can show where follow-up stopped
- If Clay did not test the company, the submission can trigger a new test and produce a report later
Clay planned to pass this context to its SDR team so reps would have one talk track, one use case, and company-specific data to discuss. The same findings could power a personalized report or landing page for each company. A rep can begin with what happened in that company’s inbound process, then show how Clay orchestrates agents to improve it.
How Clay turned one dataset into hundreds of personalized experiences
Clay drafted the report narrative in Google Docs, outlined the web experience in Ploy, and loaded the research into PloyDB. Alex could inspect the records, choose the useful breakdowns, and connect the same source to charts and company-level reports.
A visitor’s work email became the switch for the entire experience. It matched the visitor to Clay’s research, selected the right report path, and carried the same company context into the next sales conversation.
Visitor input
Work email matches the visitor to a company record
Path 01
Tested and replied
Speed, personalization, and ICP fit
Personalized benchmark and recommendations
Path 02
Tested, no reply
A recorded gap in the follow-up process
A report showing where follow-up stopped
Path 03
Not tested yet
No existing company record
A new test runs and a report follows later
Shared destination
Clay’s GTM team receives the same company context for its next conversation
One talk track, one use case, and evidence from that company’s own inbound process.
Live demo: Alex turned a signup page into a data system
Clay hosts weekly GTM Builder AMAs for its Reddit community. Guests include RevOps leaders, agency founders, internal operators, and GTM engineers. Alex still faced a basic programming question: who should Clay invite next, and what did the audience want them to teach?
He built the answer into the signup flow. After entering an email, a visitor chose the guest profile and workflows they wanted to learn from. Those answers flowed into PloyDB, where Clay could compare interests across audience segments and see which topics were gaining momentum.
Building the right aesthetic: use references
The data model was working, but the page still had to feel like Clay. Feedback such as “make it more dynamic” or “make the charts feel better” would leave too much open to interpretation. Alex brought screenshots of palettes, animations, and chart treatments that showed the direction he wanted.
When he could not choose between treatments, he asked Ploy to place several options on one workbench page. He compared them inside Clay’s design system, chose one, and refined it in context. The reference gave Ploy a concrete direction to build from.
The finished component could then join the design system alongside the PloyDB schema, agent workflows, calculations, and lead-routing logic. Each experiment left useful building blocks for the next one.
How to apply Alex’s strategy to your business
Choose a customer problem for which your team has evidence or can collect it responsibly. Then design the experience around one decision the visitor needs to make.
A useful first brief to start with:
- Customer problem: What repeated issue are you helping someone evaluate?
- Activation path: How does solving it connect to your product?
- Evidence: What original data can you gather or already own?
- Visitor result: What comparison, recommendation, or output will they receive?
- Required inputs: What must they provide for that result to work?
- Follow-up: What context should reach sales or customer success?
- Measurement: What behavior will tell you whether to keep iterating?
Keep the first test narrow. Ship it to a relevant audience, watch which parts people use, and ask sales what context improved the next conversation. Save the pieces that worked before building the next version.