Improving call back rate of ATS job applications first part: company name and tenure normalization

 TLDR; With the strategy in this blog we are checking whether we can improve the call back rate for (Applicant Tracking Software) ATS job applications where the job seeker objectively matches the job description for at least 80%. We will try as a first step company name and tenure normalization. 


A step based approach to increasing ATS submission success

There is a general sentiment between job seekers and hiring managers that it is hard to get job interviews through ATS [2,3]. Strategies that were applied in the past to bypass this problem was working through referrals, job fairs, reaching out from companies to potential job seekers and reaching out from job seekers to employees of companies. As a reference, the number of employees in the high-tech industry has decreased since April 2023 by 10% from 474,000 to 427,000 in December, this is still without the recent wave of layoffs [1]. The main challenge that we are experiencing now is that a much higher amount of skilled workers is trying to pass through ATS systems [2. 3]. The main challenge is that it is hard to be successful in trying to pass the first application step. 

Therefore we are going to try a step by step approach to understand how we can better match the needs of job seekers and hiring managers. This week we start with company name and tenure normalization. It is important to notice that the job seeker needs to match at least 80% of the skills that are asked for in the jobs that you are applying for.


Uniformization of the resume document

Definition of perfect match for the job

We define a perfect match for the job as having at least the correct years of experience, matching experience level with the tools or similar tools asked for, experience with the industry and matching soft skill set. We know it is not possible to match all of these with 100%, therefore we suggest to match at least 80%. 

If you don’t know whether you can trust  your own judgment on this, ask a trusted friend or (former) colleague to  help you out. You can also ask your (former) colleagues whether they write a referral on Linkedin that is highlighting your strengths. Be sure to return the favor to them by also writing out a referral for them!


Format alignment for the resume document

We want to be sure that we are only evaluating for now for the change in optimization of the company name. Therefore we want to reduce the amount of parsing errors as much as possible. Therefore we are going to 


Company name normalization: two parts for rule ATS company norm 1

We split up the company normalization in two parts. In part 1 we ensure that the company name is correct. In the second part we ensure that the tenure from the job seeker and the company name and location can be parsed correctly. When you apply these two steps and want to help measure the impact, you need to include ATS company norm 1 in the skill set as a job seeker. As a company you need to ensure that you can track the skillset  ATS company norm 1.


Part 1: Company name validation: change management

Companies can undergo a lot of change, they can change their name, they can get acquired, they can merge, they can stop to exist. Sometimes they also might have a slight overlap with the name of another company. To account for all these name variations we suggest to apply the following steps


Part 2: Correct parsing of role, tenure and location of the company

Based on research from a friend [2] and a lot of trial and error with some successes with submitting their resume to various ATS systems, I compiled the following rule for describing the company, tenure and location. I want to check the validity of this rule in the skill ATS company norm 1 as well. 

I have applied this rule to one of the previous companies that I worked at and I am explaining this further below.   

Head of Data Machine Learning Architect                                                 (Jun 2018 - Jan 2020)
Mappedin, Waterloo, Canada



How do we measure success?

The population that I want to test with this approach is the general job seekers population. We can’t judge based on the successes of a few people alone. We also acknowledge that the data from job seekers is confidential. Therefore we propose to evaluate success in the following way keeping confidentiality in mind for both the job seekers and the companies data.


For companies


For job seekers


Next steps

I will work over the following weeks to start implementing more rules that can help to normalize the resumes and track the improvements. I will also work further on figuring out a proper data and privacy model to keep track of the measurements, I will be sure to get this properly evaluated by industry experts. But having a normalized resume is not going to hurt anybody, so therefore I can keep working on building out extra rules at the same time. So till now be sure to keep your metrics at the place where you prefer to keep them safe! Good luck with the job search, finding the correct candidate and be sure to enjoy the weekend!


Resources