Kmart Australia has slashed its average time-to-hire for store roles by 73.8%, dropping from 44 days to just 11.8, after deploying an AI-powered chatbot that replaces conventional CVs with a blind, five-question written interview. The retailer now processes roughly 600,000 applications annually, saving an estimated $5 million to $6 million over three years, while candidates rated as stronger fits stay about two and a half times longer in the job. The results, first reported by Stockhead and detailed in a Sapia.ai case study, offer a real-world benchmark for automation in high-volume hiring—and a wake-up call for IT teams managing the infrastructure behind it.

The Chat-Based Interview That Ditched the CV

Since July 2023, Kmart and Target have used a chat interview tool developed by Melbourne-based Sapia.ai, integrated directly with SAP SuccessFactors. Instead of uploading a résumé, applicants for many store roles answer five behavioral questions designed to assess teamwork, helping others, adaptability, problem solving, and communication—traits that map directly to frontline retail demands. The system uses natural-language processing to evaluate responses and return a recommendation, but the final hiring decision stays with human recruiters, who select approximately 85% of the recommended candidates.

Crucially, the first stage is blind: recruiters do not see a candidate’s name, age, gender, school, address, or any visual CV elements. This means a 16-year-old with no work history can demonstrate potential through structured examples rather than being filtered out because their CV looks thin. The approach compresses what was often a six-week silence into a consistent, automated feedback loop—every applicant receives a report explaining their strengths and suggested roles, and candidate satisfaction scores sit at 9.1 out of 10, with 80% saying they’d recommend the experience.

Kmart Group operates over 450 stores and employs about 20,000 people under the age of 21, making entry-level hiring both massive and strategically sensitive. By removing CVs, the system addresses a long-standing fairness problem: conventional résumé screening rewards previous opportunity, keyword-stuffing, and formatting skill more than the behaviors that actually predict success on a shop floor.

What It Means for Job Seekers and Everyday Users

If you’re a teenager or young adult applying for retail positions, the shift is tangible. Gone are the days of formatting a CV with minimal content and hoping a store manager spots it in a pile. Instead, you’ll sit through a brief, text-based chat where you share real examples—how you handled a team conflict, helped a customer, or adapted to a sudden change. The bot is not a general-purpose conversational AI; it’s a structured assessment tuned for the role, so you won’t need to game keywords.

For applicants who dread phone or video interviews, the written format offers time to formulate thoughts and avoids the anxiety of live eye contact or vocal judgment. But accessibility isn’t universal. Candidates with dyslexia, limited literacy, or non-native English skills might find text-based assessments challenging, and Kmart must provide assistive technologies, extra time, or alternative pathways to compete fairly. The chatbot also gives every participant a personalized feedback report—something that manual processes rarely delivered at scale—turning rejection into a modest developmental resource.

From a customer perspective, Kmart’s observation that “applicants are also customers” resonates. A recruitment process that ghosts candidates for weeks can sour a brand, while a quick, respectful interaction builds goodwill even among those not hired.

The Behind-the-Scenes Infrastructure That Makes It Work

For IT administrators and enterprise architects, the headline speed isn’t magic—it’s infrastructure. Kmart’s deployment integrates with SAP SuccessFactors, connecting the Sapia.ai chatbot into an established HR workflow. This means candidate recommendations flow into existing recruitment pipelines, and recruiters can manage the process without toggling between isolated tools. But integration brings a host of security, compliance, and data-governance challenges that any business considering similar automation must address.

Data Privacy and Security: The chatbot collects sensitive information from people who aren’t yet employees, meaning the platform must handle both structured scores and unstructured free-text responses that might inadvertently reveal health conditions, ethnicity, or financial stress. IT teams need to enforce data minimization (ask only what’s necessary), define retention policies (don’t keep responses forever), and ensure encryption in transit and at rest.

Access Controls: Recruiters and store managers require role-based permissions. Central HR might need analytics dashboards, while a store manager should only see candidates for their open vacancies. That demands single sign-on (SSO), multi-factor authentication (MFA), and robust audit logs that track who viewed, exported, or overrode a recommendation. Using Azure AD or a similar identity provider is table stakes; without it, a compromised recruiter account could expose thousands of applicant records.

Model Versioning and Change Management: AI models evolve through vendor updates, retraining, or employer reconfiguration. A score that meant something in 2024 might not in 2026 if the underlying model changes. Kmart must version-control its assessment configurations and monitor for statistical drift, treating the chatbot like any other critical business application subject to change management. IT teams should demand vendor transparency: what’s in the model, when it was last retrained, and how updates are tested.

Incident Response: If a breach occurs, timelines matter. Contracts with AI vendors should specify notification deadlines, forensic support, and evidence preservation. The U.S. Workday litigation—where plaintiffs allege AI screening disadvantaged protected groups—underscores that courts will examine not just the employer but the software supplier’s role, making solid vendor agreements essential.

How We Arrived at 600,000 Applications a Year

Kmart’s previous hiring process was a classic retail dysfunction. Stores were inundated with CVs that sat unread for weeks, while managers spent hours skimming documents that often said little about a candidate’s actual capabilities. For a company handling 600,000 applications annually, even a generous five-minute human review per application would consume 50,000 working hours—before any interviews, scheduling, compliance checks, or offers. That’s simply non-viable, so the reality was worse: most CVs never got meaningful attention.

The problem compounds for entry-level roles. A 16-year-old’s CV is an exercise in formatting minimal information. Traditional applicant tracking systems (ATS) that rank by keyword matches further tilt the field toward candidates who understand keyword optimization or can afford professional résumé services. Worse, the slow, inconsistent response times meant Kmart often lost top candidates to competitors who could make an offer in a week or two. The hidden costs of unfilled roles—overtime for existing staff, reduced service quality, missed sales—further eroded operations.

The move to structured, chat-based assessment didn’t just automate screening; it redesigned the first stage to collect relevant behavioral evidence from everyone, not just those who survived a random triage. By July 2023, Kmart and Target rolled out the system across their network in roughly six weeks, integrating with SAP SuccessFactors in a timeline that hints at careful piloting and strong vendor collaboration.

What to Do Now: A Pragmatic Checklist for Businesses

If your organization faces volume hiring for similar frontline roles, Kmart’s results are compelling but not a copy-paste template. The success stems from aligning the tool to a specific problem: entry-level retail where CVs offered poor predictive value. Here’s a staged approach for IT and HR leaders:

  1. Define the actual recruitment bottleneck. Is it volume? Delay? Poor retention? Inconsistent screening? Technology should solve a genuine pain point, not chase a trend.
  2. Validate competencies empirically. Don’t just encode what current high performers look like. Use job analysis to identify traits that predict success, then design questions that gather relevant behavioral examples without encouraging oversharing of personal data.
  3. Pilot before scaling. Run a controlled trial to surface integration glitches, accessibility gaps, and scoring quirks. Involve HR, legal, and IT from day one.
  4. Implement rigorous IT controls.
    - Enforce SSO with MFA for all administrative and recruiter accounts.
    - Set role-based access to limit data exposure.
    - Turn on audit logging for every interaction with candidate data.
    - Establish data retention and deletion policies, and ensure the vendor abides by them.
    - Prepare incident response procedures that include the vendor.
  5. Test for demographic adverse effects. Examine the full funnel: application completion rates, recommendation rates, interview invitations, offers, and retention, broken down by gender, ethnicity, disability, and intersectional groups. A headline diversity percentage can hide subgroup disparities.
  6. Provide accessible alternatives and a human appeals process. Text-based chats aren’t for everyone. Offer accommodations and a clear path to request human review or correction of inaccurate system judgments. Train recruiters to not blindly trust scores—automation bias is real, and small numerical differences aren’t scientifically meaningful.
  7. Monitor continuously. Labor markets shift, candidate language evolves, and AI models drift. Schedule regular audits, both internal and external, and require vendors to notify you of major model updates.

What’s Next: Audits, Adaptation, and the Slow March of Regulation

Kmart’s most urgent next step is independent, longitudinal validation. While the 74% time reduction, $5-6 million savings, and improved diversity figures (8.25% First Nations hires vs. a 3% benchmark) are remarkable, all those numbers currently come from the company or its vendor. External auditors should scrutinize the full recruitment funnel for hidden biases, examine how the system performs across different store types and economic conditions, and verify retention claims.

The landscape is also shifting. The Australian Human Rights Commission has published AI recruitment compliance guidance, and Australian Public Service agencies were expected to adopt AI hiring principles by June 1, 2026. In the U.S., the Workday case continues to set the tone, with courts willing to probe whether a tool is truly just a tool or a de facto decision-maker. Employers can’t outsource accountability through a software purchase; contracts must include audit rights, breach notification obligations, and model documentation.

Meanwhile, candidates are adapting. Generative AI tools and coaching services will soon train applicants to produce the kinds of answers that score well on Sapia.ai’s dimensions. Kmart will need to decide where the line is between authentic responses and polished AI-generated narratives. Attempts to automatically detect AI-written text are unreliable, so the only robust defense may be to use the chat interview as an initial filter and rely on later human interaction for deeper verification.

Kmart’s deployment is a powerful proof point that structured AI automation can transform an operation that overwhelmed manual systems. But the same scale that delivers savings can amplify errors, and the real test will be whether the retailer can maintain accessibility, transparency, and genuine human accountability as the model, the workforce, and the applicant pool evolve. For enterprise IT teams watching, the message is clear: AI in hiring is not a simple plug-in—it’s a whole-system endeavor demanding rigorous governance, security architecture, and a commitment to fair outcomes that goes beyond the dashboard.