Good morning, my chronically employed friend. This week in the make work suck less group chat...

👏 One good thing. Google DeepMind workers just voted to unionize in the UK over militarized AI tools potentially creating a model for workers to object to other morally bankrupt business objectives. Speaking of AI...

📮State of the business. Over the last four years, the most common question I get from clients is "How do I beat the ATS?" There's a lot behind the question: conflicting resume guidance, auto-rejections, and ghosting, all underlaid by the fear of emerging AI capabilities. Back in 2022-2023, the answer was relatively simple: there was no AI overlord judging your resume and preventing a human from giving you a chance. That isn't the case anymore; over the last two years, the AI fears have come to fruition. However, the key to "beating the ATS" remains counterintuitively the same. Let's break it down.

🤖 Today, every major ATS uses active AI ranking in their resume intake process, including 79% of hiring platforms used by the Fortune 500 and 72% of platforms used by the general market. Greenhouse was the last human-first holdout.

🗄️ What is an ATS? Right: ATS stands for Applicant Tracking System. The big ones are Workday, Greenhouse, Lever, iCIMS, and SAP. Back in the day, these tools worked as a digital filing cabinets for recruiting teams to manage open roles and applicants. They could scan resumes, filter the different parts into a database, and recruiters could keyword search for the right candidates. The quickest way to get your resume to the top of the pile was to have an internal referral.

🌪️ You're fighting volume, not the robot. Since 2022 (when ChatGPT launched) applications per role have increased 158%. The market is flooded. Recruiters are wading through 300+ resumes... for some roles, thousands. And even though the good recruiters will put eyeballs on every single resume, 82% get less than 1 minute's consideration. They needed a way to figure out where to prioritize their time. We were using AI to help us write resumes. They started using AI to evaluate them.

✅ How hiring platforms work today:

1. You submit your resume through a company's ATS vendor
2. The platform parses (reads) your resume and sorts the data into a standard format
3. Some platforms strip out identifying factors
4. Then AI evaluates your parsed resume for a match to the role
5. AI assigns you a score and ranks you against all other applicants
6. The highest ranked applicants are surfaced to the top of the human review pile

📍So how do you get through all of that? Like I said, the answer I've given for the past four years hasn't changed. You need to optimize your resume so that it's easy for the tool and the human to read and understand.

1. Format: Ditch the visually fancy resumes. No tables, columns, or text boxes. Upload in .docx or PDF so the ATS can accurately read your resume. If the parsing fails, the AI is evaluating garbage.
2. Clear career history: Demonstrate steady progression (time in role, increasing responsibilities, titles, impact) using industry-standard terms. Note for my former fed friends: If you're pivoting into private sector, you need to do that translation up front. If not, you'll never get the chance to explain it to a human.
3. Story: Show, do not tell, what your superpowers are and what you're passionate about. Don't use vague terminology. Use real data with context. Don't say you're an innovative visionary, describe the revolutionary thing you created and the impact it had on the world.

☎️ But how do you increase your chances of getting called? Again, the answer is the same. Be one of the first 50-100 to apply and if possible, use your network to get a referral or introduction.

🤖 It's not all beep-beep boop-boop gravy. Biased humans created biased AI. Biased humans read AI-assisted resumes. ATS vendors have been sued (including a class action lawsuit against Workday) for age, gender, and race discrimination. In response, many of these vendors are publishing third-party audits (like Greenhouse partnership with Warden AI) to demonstrate good faith efforts to combat systematic bias... or to meet statutory requirements for automated employment decision tools like NYC Local Law 144. Some issues: audits may make us feel good, but findings do not automatically fix models, and who's to say the auditors themselves aren't using biased data? Nothing to see here folks.

💬 Coffee break gossip. Some news to share over an iced Dunkin coffee this week:

1. Consulting firms are now partnering with AI companies to... automate themselves out of business? OpenAI announced a 14 billion dollar AI consulting venture with the likes of Bain, McKinsey, Capgemini, and Goldman Sachs.
2. (Unverified rumor but still sharing) LinkedIn may be testing a feature that alerts your employer every time you apply for a role. Yikes.
3. If you haven't listened to the Love Trapped podcast yet, highly recommend you binge it this weekend. It involves a former Bachelor, faked pregnancy, and weaponized justice system. It's delicious.

Do me a favor - if you enjoyed this Saturday morning drop in your inbox, forward it to a few friends.

Until next Saturday, you got this.
Cassandra

Did someone share this group chat with you?