Clay Icon

Improving Find Contacts Efficiency: Strategies to Reduce Lead Loss

·
·

Hey Clay S. I'm trying to tighten up our Find Contacts as right now, 33% of qualified jobs in our workflow are not returning contacts. We have a 2 stage find contacts search function;

  • First stage is searching for exact job title match for the hiring company - this doesn't have a high hit rate. Currently 41/315 qualified jobs return contacts on this search.

  • When stage 1 fails, stage 2 then has a broader contact search criteria. Ie. Seniority and a whole host of "generic" job titles. This has result in 156 of the 284 jobs that didn't hit in search 1, to hit in search 2, leaving roughly 33% of the qualified jobs not finding contacts / people, at all.

At scale, this has massive impact with most positive hits returning ~2 people. Do you have any ideas as to how to tighten this so we can decrease that lead flow loss? There are cases where i've identified personnel on linkedin, that the search SHOULD be finding - ie. include keyword "Owner" which matches the person at the hiring company - yet it doesn't hit. It doesn't make much sense. Example in the snips.

  • Avatar of Channeled
    Channeled
    APP
    ·
    ·

    Here are several ways to improve your Find Contacts search results: 1. Account for Job Title Variations: • Find People does not support semantic search, so "Business Owner," "Owner," and "Company Owner" are treated as different titles • Include all possible variations of the job titles you're targeting 2. Use the New Auto-Query Feature: • Our improved Find People search now automatically includes similar job titles • You can toggle "Exact Keyword Match" if you need precise matching • The system can search by job function to find people in specific departments 3. Consider Multiple Data Providers: • Leverage Clearbit integration for additional contact discovery • Clearbit's Prospector API can help find contacts based on specific criteria like employment role, seniority, and job title 4. Optimize Your Search Strategy: • Start with exclusion filters for more precise results • For example, if searching for "Business Owner," you might exclude terms that could dilute your results • Consider linking your Find People search to your Companies table for a richer view of your data Sources: Find People [Clay 101] - Clay University Find Contacts with Clearbit Find Contacts with Clearbit | Clay.com Find AI, Improved People Search, JSON Inputs, Weighted Round Robin, New Integrations, and more!

  • Avatar of Channeled
    Channeled
    APP
    ·
    ·

    Was that helpful?

  • Avatar of Channeled
    Channeled
    APP
    ·
    ·

    Looks like you're still looking for help. If you’re dealing with a specific table, drop the URL below so we can help you quicker. Otherwise, someone from our team will be in touch soon!

  • Avatar of Alex J.
    Alex J.
    ·
    ·

    Like how is it not finding this Owner? Unless the find contacts doesn't utilise linkedin?

  • Avatar of Bo (.
    Bo (.
    ·
    ·

    Hey Alex - Absolutely. It can be a wide range of reasons (privacy, restrictions, low connections, data being missed) but luckily the reason we created Clay is that if there's data online, we can find it. I've recorded a video showing exactly how to find more people at these companies: https://cln.sh/pHgLKqxZ Step by step: 1. Run a lookup in your people table to spot missing values 2. Click "add enrichment" (top right) 3. Pick Claygent > Select GPT 4o 4. Create a metaprompter: "Find me the {roles} at {company} who are currently working there - I need their linkedin.com/in URL, full name and job title. Get as many as possible. Leave empty if none found" 5. Choose JSON output and generate to get results as a list 6. Add a run condition so it only runs when the lookup is empty 7. Use "write to other table" to send new finds to your people table (send: full name, job title, linkedin.com/in URL, and company domain) 8. Then you can enrich them Tip: Feel free to flag missing data via row left-click > leave feedback for our team to fix.

  • Avatar of Alex J.
    Alex J.
    ·
    ·

    Would you suggest this in conjunction with the current find contacts operations Bo (. or instead of?

  • Avatar of Bo (.
    Bo (.
    ·
    ·

    Since you're using the paid Find People search, I'd recommend replacing it rather than running both. If you were using the free version, then running both would make sense - free Find People first, then this method as a second pass.

  • Avatar of Alex J.
    Alex J.
    ·
    ·

    Sorry Bo (. just clarfiyng. You're recommending that we use the method in the video above first, and then have the native Find People search in Clay as back up? Or the other way around?

  • Avatar of Alex J.
    Alex J.
    ·
    ·

    I'm passing this through to a guy that's built out the environment, that's all.

  • Avatar of Bo (.
    Bo (.
    ·
    ·

    Hey! Don't worry happy to clarify. Your current setup (most expensive): You're using "Find contacts at Company" enrichment in-table, which costs 1 credit per row, followed by multiple find contacts steps writing to another table. Less changes: Start with Claygent, then use your current paid enrichments if needed (with a formula to not run if Claygent found results) But the best approach: Use the table-level Find People feature 1. Go to Actions > Find people at these companies 2. Add your titles 3. Follow with the Claygent method explained earlier here Let me know

  • Avatar of Alex J.
    Alex J.
    ·
    ·

    Could we use 4o mini for that initial scrape as opposed to 4o?

  • Avatar of Bo (.
    Bo (.
    ·
    ·

    Hey! Absolutely - you can use GPT-4-mini instead of full GPT-4o. While GPT-4o gives the highest quality results, mini can also works well. Test both and see what quality level works for your needs. Let me know if you notice any difference

  • Avatar of Alex J.
    Alex J.
    ·
    ·

    Is the Claygent integration (when triggered) also 1 credit?

  • Avatar of Bo (.
    Bo (.
    ·
    ·

    When using Claygent, the credits work in two ways: 1. If you use Clay credits, you'll see the exact cost displayed when you select your model 2. If you connect your own OpenAI API key, it uses your OpenAI credits instead and won't touch your Clay credits at all https://downloads.intercomcdn.com/i/o/w28k1kwz/1369066736/45323cf30a278f8d1aef9ef7402a/CleanShot+2025-02-06+at+_45GZbQcoql%402x.png?expires=1738885500&signature=06f0a9f6a721b118a940c1684e511ab77c3f7881507f46b7eebbd9fc57a67f77&req=dSMhH8l4m4ZcX%2FMW1HO4zUcDKJCzwBdvAPYGcIdM%2BJaWAm%2Bf%2BYDbBFvy9KUO%0Ah8HZ%0A

  • Avatar of Alex J.
    Alex J.
    ·
    ·

    Do you recommend the latter (from a $ POV)? We're on 10K credit plan.

  • Avatar of Bo (.
    Bo (.
    ·
    ·

    Hey! From a cost perspective, using an OpenAI API key could save you more credits in the long run. If you pre-pay around $100, you can reach Tier 2, which is needed for web research. Here's what I'd recommend: 1. Create an account with their API Key, 2. Upgrade to Tier 2 by pre-paying 3. Connect your API key inside Clay to reduce credit usage. [App: Article Inserter] Many users have done this and saved credits, but since this is on OpenAI's side, you might want to check with their support if needed. Let me know if you need help setting it up! 🚀

  • Avatar of Alex J.
    Alex J.
    ·
    ·

    We have Tier 5 already

  • Avatar of Bo (.
    Bo (.
    ·
    ·

    Ah sweet - Even better then! Let me know how it goes.

  • Avatar of Channeled
    Channeled
    APP
    ·
    ·

    Hi Alex J.! This thread was recently closed by our Support team. If you have a moment, please share your feedback: