Creating a Data Comparison Table for LinkedIn Job Changes in Clay

Bob X. · 2025-08-03T19:45:45.178Z

Need an expert's input judgment on this: I am looking to create a table for data comparison to see if someone changed their work experience and I want to achieve this with scheduled runs in clay. This is the logic I was thinking of creating: Having 2 tables with identical list of person LinkedIn Urls in one workbook. Both tables will have identical claygent prompts pulling the same data from those LinkedIn profiles. HOWEVER, one table will run everyday while the other runs every 2 days. So if there is a difference in the job experiences between the output of the table that runs every day compared to the one that runs every other day, then that means someone changed their job experience. If there is a job change, there will be a notification that is sent out. Would this logic work with the scheduled run function in clay?

Creating a Data Comparison Table for LinkedIn Job Changes in Clay | Clay

11 comments

· Sorted by Oldest

Ben S.
·
·
Why not use the Job Change signal for this?
Bob X.
·
·
I need it to instantly send a notification when the job experience changes on LinkedIn, and the job change signal is quite delayed and inaccurate
Bharat D.
·
·
Bob X. - Do not rely on Claygent for Linkedin signals. Try to use Apify Linkedin actor to do the same thing, it should work.
Bob X.
·
·
Hmm yeah makes sense thanks!
✅1
Vinit
·
·
Bob X. while Bharat's suggestion is good, there is one problem I.e. you need to give your LI cookie to those Apify scraper and that might lead to profile suspended or banned. The way I do it is I have a LinkedIn URL of lead, I enrich it and extract the latest experience. Now after 15 days or a month, I enrichment it again and extract latest experience. If there is a difference between latest experience, that means the lead has changed the job. You can enrich it using Lead Magic(or any other tool). This way you do not need to give your LI cookie and your work will be done as well wo running two tables. I hope this helps.
LuisArturo
·
·
Hey there Bob thanks for reaching out, and appreciate everyone else jumping in here. Just to confirm if the goal here is to track job changes, then you can use our signal to track a job change for a person's profile here instead. This will help make it so you wouldn't have to have two tables running two columns concurrently and also you can set the cadence to when do you want it to refresh.
👍1
Bob X.
·
·
Job changes signal might be hard to achieve with the intention of this table. So the goal of this table is to have a list of employees that work at a company (Company A) and look for employees that have concurrent work experiences, ie working 2 jobs at the same time. So they would have an active role at Company A but also an active role at another company. Then there would be slack alerts that are sent out under 3 conditions. (Alert 1): An employee went from no concurrent employment to having a concurrent employment (Alert 2): An employee went from concurrent employment to no concurrent employment (Alert 3): An employee switched from one concurrent employment to another concurrent employment. MY CURRENT TABLE: uses claygent GPT 4o to extract the employees concurrent employment status through their LinkedIn url. Outputting the concurrent employment’s company name, company LinkedIn URL and job title at the concurrent employment. (This is duplicated across 2 tables with one table running the same claygent every 2 days and the other running everyday. Both tables then port data into a 3rd table where the concurrent employment gets compared and a formula is used to decide if one of the 3 alerts gets sent out. THE ISSUE: The claygent run is inaccurate sometimes and can pull other experiences from the persons LinkedIn profile (not the concurrent one) There also might be delays MY QUESTION: curious if there is a way to accurately output people’s job experience data from their LinkedIn profile in an compare the concurrent employment results between day 1 and 2 to see if there has been a job experience change. Need the data to be super real time (at most, 1 day delay) and accurate
Abdul S.
·
·
Hey Bob X. it's smart but yeah claygent isn’t 100% reliable for LinkedIn job history, especially with concurrent roles linkedIn data isn’t real time and GPT can pull outdated or irrelevant experiences best bet: compare timestamped outputs across runs, add logic to ignore older roles, and maybe add a confidence filter (like titles with “current” in them) real time accuracy is tough without access to deeper APIs, but we can reduce false alerts this way
Hari K.
·
·
Hey Bob! Just wanted to check and see if you were able to figure this out with the community or if you still need guidance here?
Bob X.
·
·
Hey Abdul S. when you say compared timestamped outputs acrosss multiple runs do you mean enriching concurrent employment using other methods at the same time as claygent, such as Claude, Apollo, lead magic etc to compare and see what’s the most common output? Then use formulas to extract that one that is the most “common”? Also for timestamped output, is there a formula or way to get timestamps per column or would I need to create a separate table for every type of data source I use to enrich concurrent employment? Thanks!
Channeled
APP
·
·

Ben S.
·
·
Why not use the Job Change signal for this?
Bob X.
·
·
I need it to instantly send a notification when the job experience changes on LinkedIn, and the job change signal is quite delayed and inaccurate
Bharat D.
·
·
Bob X. - Do not rely on Claygent for Linkedin signals. Try to use Apify Linkedin actor to do the same thing, it should work.
Bob X.
·
·
Hmm yeah makes sense thanks!
✅1
Vinit
·
·
Bob X. while Bharat's suggestion is good, there is one problem I.e. you need to give your LI cookie to those Apify scraper and that might lead to profile suspended or banned. The way I do it is I have a LinkedIn URL of lead, I enrich it and extract the latest experience. Now after 15 days or a month, I enrichment it again and extract latest experience. If there is a difference between latest experience, that means the lead has changed the job. You can enrich it using Lead Magic(or any other tool). This way you do not need to give your LI cookie and your work will be done as well wo running two tables. I hope this helps.
LuisArturo
·
·
Hey there Bob thanks for reaching out, and appreciate everyone else jumping in here. Just to confirm if the goal here is to track job changes, then you can use our signal to track a job change for a person's profile here instead. This will help make it so you wouldn't have to have two tables running two columns concurrently and also you can set the cadence to when do you want it to refresh.
👍1
Bob X.
·
·
Job changes signal might be hard to achieve with the intention of this table. So the goal of this table is to have a list of employees that work at a company (Company A) and look for employees that have concurrent work experiences, ie working 2 jobs at the same time. So they would have an active role at Company A but also an active role at another company. Then there would be slack alerts that are sent out under 3 conditions. (Alert 1): An employee went from no concurrent employment to having a concurrent employment (Alert 2): An employee went from concurrent employment to no concurrent employment (Alert 3): An employee switched from one concurrent employment to another concurrent employment. MY CURRENT TABLE: uses claygent GPT 4o to extract the employees concurrent employment status through their LinkedIn url. Outputting the concurrent employment’s company name, company LinkedIn URL and job title at the concurrent employment. (This is duplicated across 2 tables with one table running the same claygent every 2 days and the other running everyday. Both tables then port data into a 3rd table where the concurrent employment gets compared and a formula is used to decide if one of the 3 alerts gets sent out. THE ISSUE: The claygent run is inaccurate sometimes and can pull other experiences from the persons LinkedIn profile (not the concurrent one) There also might be delays MY QUESTION: curious if there is a way to accurately output people’s job experience data from their LinkedIn profile in an compare the concurrent employment results between day 1 and 2 to see if there has been a job experience change. Need the data to be super real time (at most, 1 day delay) and accurate
Abdul S.
·
·
Hey Bob X. it's smart but yeah claygent isn’t 100% reliable for LinkedIn job history, especially with concurrent roles linkedIn data isn’t real time and GPT can pull outdated or irrelevant experiences best bet: compare timestamped outputs across runs, add logic to ignore older roles, and maybe add a confidence filter (like titles with “current” in them) real time accuracy is tough without access to deeper APIs, but we can reduce false alerts this way
Hari K.
·
·
Hey Bob! Just wanted to check and see if you were able to figure this out with the community or if you still need guidance here?
Bob X.
·
·
Hey Abdul S. when you say compared timestamped outputs acrosss multiple runs do you mean enriching concurrent employment using other methods at the same time as claygent, such as Claude, Apollo, lead magic etc to compare and see what’s the most common output? Then use formulas to extract that one that is the most “common”? Also for timestamped output, is there a formula or way to get timestamps per column or would I need to create a separate table for every type of data source I use to enrich concurrent employment? Thanks!
Channeled
APP
·
·