Higher Diversity: A Research-Backed Approach for Raising Your Standard for Diversity Recruiting

Diversity recruiting often lives in a separate silo: extra job board contracts, conference sponsorships, standalone sourcing teams. When budgets tighten, that work disappears first, and it rarely fixes what's broken in the underlying hiring process. A different approach treats diversity as a systemic problem to be solved with the data, tools and process fixes already available inside a TA function.

The framework rests on three components: using data that already exists in an ATS or CRM (even if it's messy), applying behavioral science research on decision-making to hiring design, and adjusting choice architecture, the way systems are configured to nudge candidate and recruiter behavior.

Several examples illustrate the approach:

• An application that took over seventeen minutes to complete on a phone caused the vast majority of candidates, diverse and non-diverse alike, to abandon it before finishing. Mobile traffic skewed more diverse than desktop traffic, tying the slow process directly to broadband and device access gaps. Turning on existing ATS features like profile parsing and mobile quick apply, and cutting irrelevant prescreen questions, dropped completion time to under three minutes and cut abandonment from the nineties down to roughly a quarter, for all applicants.
• An organization drowning in unqualified applicants found that free job scrapes on a major platform were burying a small number of high-quality, diverse candidates coming from paid job slots and internal sourcing. Cutting the noisy free channel, dropping a diversity job board that hadn't produced a qualified candidate in years, and reallocating that spend toward channels that were already overperforming produced measurable gains in applications from women and from Black and Hispanic candidates within months, without any increase in total spend.
• A vendor that claimed strong performance among women of color was tracked closely enough to catch a silent algorithm change that had quietly shifted its applicant pool toward white men, giving the recruiting team leverage in renegotiating the contract.
• A long-standing sponsorship of a women's engineering conference turned out to draw an audience that was mostly men, prompting a rethink of what recognition and outreach actually resonate with the audience being targeted.

Across each case, the pattern is the same: pull the data that already exists, question spending that has continued out of habit or fear, and fix the process itself rather than layering a separate diversity initiative on top of it.