How I Use Clay to Build Hyper-Personalized Lead Lists (And Why Most People Get It Wrong)

I've spent well over 300 hours inside Clay and made hundreds of thousands of dollars as a direct result of that time. It's probably the best lead generation tool available right now. It's also one of the most difficult pieces of software I've ever had to learn. Most people land on the homepage, see a blank screen, and quit. This article is the map they never had.

Here's exactly how I use Clay to generate tens of thousands of leads and book 250+ sales calls every month for us and our clients, kept as simple as I can make it.

Start With Company Sourcing, Not Enrichment

The first thing I do inside Clay is pull in the companies I want to target. For this walkthrough, let's say we're going after software development companies in the US, SMB size, anywhere from 2 to 200 employees.

Clay will surface 50,000+ results. I preview 50 rows to start. Here's where most people make their first mistake: Clay offers enrichment options right out of the gate, and beginners burn credits on all of them immediately. I don't do that. I skip the bulk enrichment at this stage and pull only the specific data points that are actually relevant to my offer. That keeps costs down and keeps the table clean.

The base data Clay pulls, company name, description, size, location, tech stack, is already useful. But the real value starts when you get surgical.

Use Tech Stack Data to Find a Hook

Once I have the company list, I add a field to pull the technologies each company is running on their website. This is where targeting gets interesting.

If I'm selling Framer migration services, I filter for companies built on Webflow. If I'm selling email marketing strategy, I check whether they're running Klaviyo. If I'm selling a support solution, I look at what support stack they're already using and position against it.

I also pull website traffic data directly inside Clay rather than jumping over to a separate tool. You can see real traffic numbers, identify companies underperforming against their competitors, and use that as the opening line in your outreach. "Your traffic is significantly behind two of your direct competitors" is a much stronger opener than anything generic.

Every company has something you can latch onto. The tech stack is usually the fastest place to find it.

Find the Right Person, Then Run Waterfall Enrichment

Once I know a company fits my criteria, I need to find who to actually contact. Inside Clay, I add an enrichment step to find people at the company, filtering by titles like Site Owner, Partner, or VP depending on the offer.

That data pushes cleanly into a new table: first name, last name, job title, location, LinkedIn profile. Clean and organized, all in one place.

Then comes waterfall enrichment for email addresses. Instead of bouncing between Apollo, ZoomInfo, or other databases separately, Clay runs through multiple sources in sequence, I use Lead Magic and Prospeo as my primary API integrations because they're cost-effective and accurate. Clay pulls from the full name, company domain, and LinkedIn profile to find the verified work email.

The waterfall approach means you're not relying on a single database. If one source doesn't have the email, the next one fires automatically. The result is a higher match rate and a cleaner final list.

Personalize at Scale With Clay AI (Claygent)

This is the part that separates a good Clay user from a great one. Clay has a built-in AI research agent called Claygent, and it's genuinely powerful.

Here's a concrete example of how I use it. Let's say I'm selling content marketing services. I configure Claygent to visit each company's domain and find three of their direct competitors. Then I run a second Claygent workflow to generate SEO blog ideas that would help that company take market share from those specific competitors.

Now my outreach email isn't generic. It references their actual competitors by name and offers a specific content angle that maps directly to their market position. That level of personalization at scale used to require a full research team. Claygent does it in a workflow.

One practical note: run Claygent on the Mini model when the task doesn't require deep reasoning. It burns significantly fewer credits and the output quality is more than sufficient for most research tasks.

📥 Best Cold Email Software 2026

The 7 cold email tools worth your money in 2026 — ranked by an agency managing 25,000+ inboxes.

Get it here →

Export, QA, and Push to Your Sending Tool

Once the table is built, verified emails, competitor data, personalization angles, I export it. Clay connects directly to HubSpot, Salesforce, Google Sheets, and Smartlead, among others.

My process: export to Google Sheets first, have a VA do a quick scan to confirm the data looks clean, then push to Smartlead. That one manual QA pass has saved us from sending campaigns with bad data more times than I can count. Automation is great; a final set of human eyes is still worth it.

The Two Mistakes That Kill Most Clay Users

Starting without a workflow plan. Clay's homepage is a blank canvas. If you don't know what you're building before you open it, you'll waste hours and credits going in circles. Map out your targeting logic, enrichment steps, and personalization needs before you touch the tool.

Skipping the API setup. I strongly recommend buying the GPT-4 package and getting to at least Tier 3. Connect your own API keys for your data providers rather than running everything through Clay's native credits. It's significantly cheaper, and you get better data quality because you control which sources you're pulling from.

Clay rewards patience and proper setup. Rushed configurations produce expensive, mediocre lists.

Key Takeaways

• Pull company data first, skip bulk enrichment until you know exactly what you need.

• Use tech stack and traffic data to find a specific, credible hook for your outreach.

• Waterfall enrichment across multiple databases (Lead Magic, Prospeo) beats any single source for email coverage.

• Claygent can research competitors and generate personalized content angles at scale, use the Mini model to save credits.

• Always QA your export in Google Sheets before pushing to your sending platform.

• Set up your own API keys and reach Tier 3 before running serious volume. It cuts costs and improves data quality.

Frequently Asked Questions

What is Clay and what makes it different from tools like Apollo or ZoomInfo? Clay is a lead enrichment and research platform that pulls data from multiple sources in a single workflow. Rather than relying on one database, it runs waterfall enrichment across providers like Lead Magic and Prospeo simultaneously, and it adds AI-powered research capabilities that tools like Apollo don't offer natively.

What is Claygent and how do I use it for personalization? Claygent is Clay's built-in AI research agent. You can configure it to browse the web, find competitor information for each company on your list, generate content ideas, or pull any other research that makes your outreach more specific. For most tasks, running it on the Mini model keeps credit costs low without sacrificing output quality.

Why does Clay have such a steep learning curve? The platform opens to a blank screen with no guided workflow. Without a clear plan for what data you need and in what order, it's easy to burn credits on unnecessary enrichments and end up with a disorganized table. The learning curve is mostly about workflow design, not the tool itself.

How do I keep Clay costs under control? Connect your own API keys for data providers rather than using Clay's native credits for every enrichment. Reach at least Tier 3 and add the GPT-4 package. Only pull the specific data fields you actually need for your offer, don't enrich everything by default just because the option is there.