Essays and strategic takes.

I have the $200 Max plan and I kept reaching for the most expensive model for everything. This is the system I built for picking a Claude model and effort level, why the cheap one isn't always cheaper, and the free skill that makes Claude do the picking for me.

Four strange things started happening to my posts this summer. So I went digging through what LinkedIn has actually published about its feed. There was no summer update. There was an eighteen-month rebuild nobody announced.

There is no universal list of best LinkedIn keywords. How to pull the words that matter from the jobs you actually want, and where they belong on your profile.

Search appearances go up and down, but the number alone tells you nothing. What LinkedIn actually measures, and how to check whether you're findable for the right roles.

LinkedIn Recruiter now mixes filters, Boolean, and AI search. What that means for getting found, what LinkedIn documents, and what nobody actually knows.

Six and a half months of running my own job search like an instrumented funnel: 148 applications vs. 70 inbound leads, why findability beats volume, and every tool and write-up I've published, in priority order.

People keep asking why juliedx.com doesn't look AI-designed. Because AI didn't design it. It executed a brand I built first. The full process: interrogation, pressure-testing, voice, visual territory, and a locked system, with every prompt to run it yourself.

It's not you. In 2022, a remote marketing director had roughly 1,200 open roles with ~290 applicants each. Today: 389 roles, ~1,100 applicants each. About 92% of your odds are simply gone. Here's the data, and what I'd do about it.

Two upgrades to my automated job search system after months of daily runs: a relay of scheduled tasks that finishes the long jobs, and a loop that keeps the system's own instructions current. The original system still works. This makes it better.

A free Claude skill that turns AI into a real interview coach. It asks the hardest questions your actual interviewer would ask, in their voice, then catches the delivery habits draining your answers, one pattern at a time. Copy it below.

Six months, 70 recruiters and hiring managers reaching out first, starting from 3,500 followers. What the data says works, what doesn't, and why the KPI matters more than the tactics.

Put your own expertise into a free AI tool that becomes a version of you: a lead magnet, a second opinion, a taste of how you think. The talk I gave, written up, with the four exact prompts to build your own.

Your data lives in a handful of tools that don't talk to each other. Here's how to pull it into one dashboard that answers the questions you actually have, plus the prompts to build your own.

On paper I'm a big-company person. I've been a builder the whole time. Twenty years of the story my resume never told.

I spend maybe an hour a week looking for jobs and still see more of what's actually open than people who scroll their feed every day. The Boolean search strings, the right ATS platforms to target, and the automated company-list sweep I use.

A landing page audit tells you what's actually wrong with the page and what to fix first. Here's what a real audit checks, why ChatGPT isn't a standard you can defend, and the compliance lens regulated pages need.

After seven years marketing in healthcare, here are the questions I learned to raise about a landing page before it went live, and the people to bring in for the answers a checklist can't give you.

Most landing page audit checklists grade formatting. A real audit names the root cause. Here's the six-step process that tells you what's actually killing conversions.

I tracked 172 of my own LinkedIn posts during a job search. The hooks, cadence, and post types that drive profile views from people who hire, and the one metric that lies.