Soko Healthcare: Website, Intake & Fractional Operations

The problem

Soko Healthcare places nurses directly with South Florida facilities, and every candidate is screened by a working RN. The firm had a live site but no system behind it: no way to catch a half-finished inquiry, no alert when a facility reached out, no CRM, and copy that talked about temporary staffing when the model is direct hire only.

The goal was the smallest system that takes a facility from “found the site” to “signed agreement,” and a nurse from “saw a job” to “placed,” and then an operating rhythm to keep it running as the firm grows.

What we built

• A new home on Cloudflare Pages. The site moved off its original builder and now deploys from Git, with every non-production branch published as a no-index preview. The firm’s email was never touched.
• Direct-hire copy. Every page leads with the facility’s problem, and the temporary-staffing language came out. No claims about fees, guarantees, or outcomes go on the site until the firm confirms them.
• Two intake paths. A facility inquiry and a nurse application, each asking the questions first and the contact details second, with résumé upload on the nurse side and a thank-you page with a way to book a call.
• One source for the forms. The questions, picklists, and a 15-point lead score live in one file. The site renders it, the server function validates and scores against it, and a test checks that the score only names real options.
• A CRM the firm already knows. Every submission lands, scored, in a Google workbook the founder works from daily. It is the CRM for now; a dedicated one comes when volume justifies it.
• Measurement. GA4 events for both forms plus phone and booking clicks, running only on the production hostname, and Search Console.

Fractional operations

The engagement didn’t stop at launch. OpsPal is Soko’s fractional operations team: we own the intake and follow-up rules, the recruiting pipeline stages, reporting, and the operating cadence, so the founder’s time goes to nurses and facilities instead of systems. Changes go in one at a time, through a shared log, against a written plan.

How it works

The site is built for search engines and AI assistants from the start: Organization, Person, Service, FAQ, and breadcrumb structured data, an llms.txt file, and robots rules that welcome AI crawlers. Every month we check whether Soko shows up when someone asks ChatGPT, Perplexity, Claude, or Google how to hire a nurse in Broward County.

The rollout is phased: Broward first, then Miami-Dade, then Palm Beach. City and specialty pages get added after the first placements prove the model.

Stack

Cloudflare Pages and Pages Functions, Google Apps Script and Sheets, GA4, Google Search Console, Schema.org structured data, llms.txt.

Where it stands

Live at sokohealthcare.com. The site, both intake paths, and the workbook CRM are running, and the fractional operations retainer continues month to month.