How do I calculate the ROI of an ATS?

You cannot show a return without a starting number, and the starting number has to be captured before the tool arrives. Record three things over a normal quarter. First, time: how many hours recruiters and hiring managers spend per hire on scheduling, screening, chasing feedback and status updates. Second, direct spend: job advertising, agency fees, assessment tools and any contractor coverage while roles are open. Third, duration: how long roles stay open, and what the business loses while they do. Most teams find the time figure hardest because nobody logs it, so estimate it with a two-week sample and be transparent that it is an estimate. A rough honest baseline beats a precise number invented after the fact, which is what happens when you skip this step and try to prove value retrospectively.

Treat administrative time reduction as the most reliable line. Automating scheduling, feedback chasing and status emails removes work that genuinely existed, and you can verify it by asking the same people the same questions after rollout. Reduced advertising waste is also verifiable if you track source performance. Be more careful with two claims. The first is headcount savings, because most teams redeploy recruiter hours into sourcing rather than removing roles, and that is a better outcome but not a cash saving. The second is revenue attributed to faster hiring, which is real for quota-carrying and billable roles and speculative elsewhere. Split your model into hard savings you would defend to finance and soft benefits you describe qualitatively. Tracking cost per hire consistently on both sides of the change makes the argument much harder to dispute.

More than the subscription. Include implementation and configuration effort, data migration, integration work, training time for recruiters and managers, and the productivity dip during the first weeks when people are learning. Add any per-integration or per-module charge, and any cost that grows with your hiring volume under the vendor's pricing model. Then extend the whole picture across the realistic life of the contract rather than the first year, because implementation is front-loaded and the benefit curve is not. A system that looks unattractive over twelve months often looks obvious over three years. State your assumptions in the model itself so a sceptical reviewer can change one input and see the effect, rather than arguing with a single number they cannot inspect.

Expect the time savings first, usually within the first full hiring cycle after adoption settles, and the pipeline and quality effects considerably later. Scheduling and coordination gains appear quickly because they replace work that was already happening. Improvements in source quality, drop-off and offer acceptance depend on having enough hires recorded in the system to see a pattern, which for a smaller team can take several quarters. Set that expectation with your budget holder before signing, because the most common way an ATS business case fails is being reviewed at month three against benefits that were never going to appear until month nine. Agree the review date and the specific measures in advance, and use recruiting metrics your team already tracks so the comparison is credible.