Twenty Years in IT Staffing: What's Changed, and What Stubbornly Hasn't

OKAYA is twenty years old this year. That's old enough to remember faxing resumes, young enough to have an opinion about AI screening tools, and exactly the right age to notice which things in this industry actually change — and which things people keep predicting will change and never do.

Some honest notes from two decades of matching people with work.

What genuinely changed
The speed. In 2006, a search had a rhythm of weeks — post, wait, sift, schedule around everyone's office calendar. Today a hot candidate is in play across multiple companies within days, and a client who takes two weeks to schedule a first interview has usually already lost. The entire game compressed, and every part of our process had to compress with it.

The geography. Remote work didn't just widen candidate pools; it repriced them. A client in Manhattan now competes with a company in Austin for a developer in Ohio — for some roles the local market simply stopped being the market. Hybrid pulled some of that back, but the old assumption that talent lives within commuting distance never returned, and we don't think it will.

The transparency. Candidates once needed a recruiter to know what roles existed and what they paid. Now rates are a search away and pay ranges sit in the postings, which is mostly a wonderful thing — it burned off an entire generation of staffing firms whose business model was information hoarding. What survived is the part that was never about information: access, judgment, and advocacy.

And yes, the tools. We went from a filing cabinet's worth of resumes to a database of 2.5 million profiles with twenty years of placement history in it. AI now drafts job descriptions and screens resumes across the industry. The tooling is unrecognizable.

What stubbornly hasn't
Hiring is still a trust decision made by nervous humans. Every manager, before every offer, still asks some version of: will this person make me regret this? No platform has ever answered that question. A recruiter who has placed someone before, watched them deliver, and stakes their name on the match — that still moves a hire across the line in a way no scoring model does.

The resume still isn't the person. The best placement in this firm's history looked mediocre on paper and got sold on a phone call — someone here had seen her work. Twenty years of tooling has not changed the fact that the map is not the territory, and the fashionable screening technology of every era has mostly been a new way to misread maps faster.

And relationships still compound better than any asset we own. Consultants we placed in 2008 are hiring managers who call us today. That arithmetic — do right by people, wait a decade, be pleasantly shocked — has survived four recessions, one pandemic, and every technology wave that was supposed to replace us.

What the next twenty asks of us
AI will keep absorbing the mechanical parts of recruiting, and honestly, it can have them. Parsing, matching, scheduling — that was never the job. The job was always the conversation where a candidate admits what they actually want, and the one where a client admits what the role actually is. Whoever keeps doing those conversations well stays valuable in any decade. That's the plan for this one.

To the clients and consultants who've been part of these twenty years — thank you.