How to Save Clay Credits & AI Tokens
If you’ve been using Clay for any length of time, you’ve probably noticed how quickly credits and AI tokens can add up. Whether you’re enriching company data or qualifying leads, inefficient workflows can drain your budget faster than you’d like. But here’s the good news: there’s a simple strategy that can save you a ton of resources, and it all comes down to how you structure your tables.
The Problem: Running Enrichments Multiple Times
Let’s paint a picture of what commonly happens. You’re building a lead list, and you’ve got multiple people from the same company in your database. Maybe you’ve scraped LinkedIn, imported a list, or pulled data from another source. For each person, you run company enrichments to gather firmographic data, qualify the company, or feed information into AI prompts.
Here’s where the problem kicks in: if you have 20 people from the same company in your list, and you’re running company-level enrichments on each row, you’re actually enriching that same company 20 times. That’s 20x the credits, 20x the API calls, and 20x the AI tokens being consumed for the exact same information.
This isn’t just wasteful—it’s expensive. And when you’re scaling your outreach and working with thousands of leads, these costs multiply quickly.
The Solution: Separate Your Company Qualification
The fix is actually quite straightforward, but it requires thinking about your workflow differently. Instead of running company enrichments on your main people table, you want to create a separate “qualified companies” table.
Here’s how it works: you extract all the unique companies from your leads list and write them to a dedicated company qualification table. In this table, you run all your company-level enrichments once per company. This means gathering information like company size, industry, funding status, technology stack, or any other firmographic data you need to determine if they’re a good fit.
Once you’ve enriched and qualified the companies in this separate table, you can then reference that data back in your main people table without having to run those enrichments again. This way, all 20 people from the same company are pulling from the same enriched company record.
The Impact on Your AI Token Usage
This approach is particularly powerful when you’re using AI for qualification. Let’s say you’re using OpenAI through Clay to analyze company websites, news articles, or other data points to determine if they match your ideal customer profile.
If you’re running these AI prompts on every single person row, you’re burning through tokens at an alarming rate. But when you move this logic to a company-level table, you’re only running it once per company. This can reduce your OpenAI token consumption by 10x, 20x, or even more depending on how many contacts per company you’re working with.
The same principle applies to other enrichment providers. Whether you’re using Clearbit, Apollo, or any other data provider through Clay, running enrichments at the company level rather than the contact level means you’re only paying for each unique company once.
How to Implement This Strategy
Here’s a simple workflow to get started:
Step 1: In your main people table, identify the company domain or company name for each contact.
Step 2: Create a new table called “Qualified Companies” or something similar.
Step 3: Use Clay’s “Write to Table” feature to send unique companies from your people table to your qualified companies table. Make sure you’re only writing unique values so you don’t create duplicates.
Step 4: In your qualified companies table, run all your company-level enrichments. This includes firmographic data, technographic data, AI qualification prompts, website analysis—anything that’s relevant at the company level rather than the individual contact level.
Step 5: Back in your people table, use lookups or enrichments to reference the qualified company data. This way, each person inherits the company qualification without having to run those enrichments again.
Additional Benefits Beyond Cost Savings
While saving credits and tokens is the main benefit, this approach has other advantages too. Your workflows become cleaner and more maintainable. When you need to update your qualification criteria, you only need to change it in one place rather than multiple tables or workflows.
It also makes your data more consistent. When all contacts from the same company are pulling from the same enriched record, you eliminate discrepancies that can happen when enrichments are run at different times or return slightly different results.
Plus, your tables run faster. Instead of waiting for the same enrichment to process 20 times, it only runs once, which means your workflow completes more quickly.
Start Optimizing Your Workflows Today
If you’re serious about scaling your lead generation while keeping costs under control, implementing a company-level qualification table is one of the most impactful changes you can make. It’s a simple structural shift that can save you thousands of credits and dramatically reduce your AI token consumption.
The next time you’re building a workflow in Clay, take a moment to think about what data is truly contact-specific versus what’s company-specific. Making this distinction and structuring your tables accordingly will pay dividends in both cost savings and workflow efficiency.
Want help automating your lead generation? Book a call