Recruiting Automation Pricing: Paradox, Workday and Greenhouse on Three Meters

Workday's Recruiting Agent charges 6 Flex Credits to screen one resume, while the Workday Recruiting module underneath it has historically been priced on your total employee count. That is two meters inside one vendor before you add a conversational agent or a second system. Recruiting automation pricing is now the clearest case in HR software of agents and the ATS billing on units that do not move together.

Recruiting automation pricing runs on three meters at once

A meter is the unit a vendor counts to produce your invoice. Seats, employees, credits and hiring volume are all meters. The problem in recruiting is that a single hire passes through tools that each count something different.

Before agents arrived, this mattered less. Your applicant tracking system, the database that holds requisitions, candidates and interview feedback, was priced on one unit and stayed there for the contract term. Agents changed that by attaching usage charges to individual steps inside the same workflow.

The ATS meter counts employees or hiring volume

Workday has said plainly that its historical model is employee based. A post on Workday's own blog states that the company has historically operated on an employee-based model that ties pricing to the total number of employees in an organisation. Greenhouse prices differently, saying cost depends on the plan, your hiring volume, your organisational complexity and the features you need.

Both of those meters are fixed at signing. Neither changes when a requisition receives 40 applicants instead of 400. That stability is the entire value of a seat or headcount meter, and it is exactly what agent pricing removes.

The agent meter counts actions

Workday's Flex Credits are an annual block of credits drawn down as agents complete work. Workday says each metered skill is assigned a credit value based on its estimated value in terms of time saved and efficiency gained. The same blog post says agents are metered when they complete an action, not every time a query runs or a token is consumed.

That is a defensible design. It also means the unit you pay for is the resume or the requisition, which are inputs to hiring rather than the hire itself.

The conversational meter is an unpublished quote

Paradox, whose assistant Olivia handles chat screening and interview scheduling for high-volume roles, became part of Workday when the acquisition closed on 1 October 2025. Workday's release says the Workday Paradox Candidate Experience Agent is available for purchase through Workday or Paradox. Neither that release nor the Flex Credits page says the product draws on Flex Credits.

Paradox's own pricing page lists no prices, no tiers and no billing unit. So the third meter in this stack is not just unpublished in dollars. Its unit is unpublished too.

What each recruiting software vendor actually publishes

The table below uses only what the vendors state on their own pages, plus one rate card figure reported by a trade publication. Where a column is empty in reality, it says so rather than filling the gap with a reseller estimate.

Workday rate card values come from CIO's July 2026 reporting of the card as of 21 May 2026. The Greenhouse Notetaker metering terms are reported by a pricing tracker and could not be opened on Greenhouse's legal index, which is discussed in the limitations section.

Read across the last column and the problem is visible. Four different quantities drive five line items: headcount, applicants, requisitions and hiring volume. Not one of them is the number of people you actually hired.

My view is that this is not an accident of packaging. Each vendor has chosen the unit that grows fastest in its own part of the funnel. Applicants outnumber hires, requisitions outlast them, and headcount compounds every year whether you are hiring or not.

How Workday's Recruiting Agent burns Flex Credits

Lynn Greiner's July 2026 analysis in CIO is the most specific public account of the rate card. It reports two Recruiting Agent skills. Screening and grading a resume against a job opening costs 6 credits. Identifying leads in existing talent pools and rediscovering past candidates costs 750 credits per requisition.

For scale, Workday's own Flex Credits page gives the Self-Service Agent's information retrieval skill at 1 credit per action and its autonomous task completion at 5 credits. A resume screen therefore sits in the same band as an ordinary agent task. The requisition charge is in a different class entirely.

Credits expire and the allowance scales with company size

CIO reports that Flex Credits expire after one year and unused credits do not roll over. Workday says its complimentary allowance is set by company size. Workday also says the Platform Consumption Console alerts administrators at 80%, 90% and 100% of consumption.

Notice what that combination does in recruiting. Your free allowance is sized on headcount, the ATS is priced on headcount, and the agent consumes credits on applicants. A 2,000-person company running a seasonal hiring surge gets the allowance of a 2,000-person company and the consumption of a much larger one.

Workday says usage is now where the value sits

Workday's chief technology officer Gabe Monroy put the argument directly in the same CIO piece. He said value is no longer derived from a fixed factor like how many employees you have, but from how much use you get out of the system. That is a clear statement of intent, and it comes from a company whose AI business is growing quickly.

In its fiscal 2027 second quarter results, Workday reported more than 5,500 customers using at least one organic agent, up more than 35% in a quarter. It also said AI drove more than 25% of new annual contract value. Usage pricing is not a side experiment for Workday. It is where new revenue is coming from.

Where recruiting automation meters conflict inside one workflow

Meters conflict when the same hiring decision pushes one bill up and another down. Three examples show it, using only units the vendors have stated.

First, applicants are not hires. A job advert that attracts a flood of weak applications costs more on a per-resume agent meter even if it produces the same single hire. Under a headcount ATS meter, that same flood costs nothing extra. Your recruiters will feel the second meter and your finance team will see the first.

Second, requisitions are not hires. A retailer opening one requisition for 20 store associates and a retailer opening 20 separate requisitions for the same roles make the same hires. At 750 credits per requisition for talent rediscovery, the second structure costs 20 times as much on that skill. Requisition hygiene, until now an operations preference, becomes a cost lever.

Third, the ATS bill does not move at all inside the term. If agents genuinely cut recruiter workload, the headcount meter gives you none of that saving back until renewal, and Workday's meter is headcount across the whole company, not recruiters. Savings show up in your team's capacity. Costs show up on the credit statement.

This is the same pattern that has appeared in support software, where Intercom bills Fin per resolution on top of seats. Recruiting is worse in one specific respect. The support agent's unit, a resolved conversation, is at least close to the outcome a buyer wants. The recruiting agent's unit is two steps removed from it.

One 100-hire year billed three ways

The worked example below is illustrative. The applicant and requisition counts are assumptions chosen to show how each meter responds, not data from any customer. The credit rates are the reported Workday figures. No dollar conversion is shown, because Workday does not publish a price per credit.

Assume a company makes 100 hires in a year. In the base case it opens one requisition per hire and runs talent rediscovery once on each. The only variable is how many applicants each hire attracts.

In the base case, the Workday agent meter consumes 105,000 credits. The requisition skill is the larger share, at 75,000 of the total. Most buyers would assume screening drives the bill, because it is the per-candidate action. At 50 applicants per hire, it does not.

Now vary the inputs. At 25 applicants per hire, total consumption falls to 90,000 credits. At 250 applicants per hire, a realistic figure for some frontline roles, it rises to 225,000. Keep 100 applicants per hire but consolidate the 100 hires into 20 requisitions of 5 hires each, and consumption drops to 75,000 credits.

The first way to bill this year is the full Workday stack. That means a headcount subscription that ignores all of the above, an agent meter that swings threefold, and a Paradox licence whose unit nobody outside a sales call can state. The second way is Greenhouse, where 100 hires feed into a quote and the AI is mostly inside the plan. The third way is Paradox bought on its own, where you negotiate a number with no public reference point at all.

I would rather budget the Greenhouse shape for a company hiring 100 people a year. The cost is opaque at the quote, but it is stable for the term. A swing of 150,000 credits on a fixed hiring plan is a forecasting problem that most HR budgets are not built to absorb. That point is made more broadly in the breakdown of why credits and tokens confuse AI buyers.

Greenhouse AI features: bundled, with one meter creeping in

Greenhouse's AI recruiting page says its AI is available across all plans, naming Core, Plus and Pro. The same page also says advanced features such as AI-powered report filters and resume anonymisation sit on higher tiers. Both statements can be true, and together they describe a bundle with tier gates rather than a meter.

The pricing page confirms the split. Core lists the AI Notetaker and Talent Matching. Plus adds AI-powered report filters, Greenhouse Sourcing Automation, and fraud and spam detection. Pro adds enterprise-level Real Talent features. There is no price anywhere on the page, and every plan leads to a demo request.

The one exception to the bundle is the Notetaker. A pricing tracker, UsagePricing, reports that Greenhouse Notetaker terms dated 31 July 2026 meter the product in usage credits and recording-minute limits, and allow Greenhouse to require additional credits before further use. Greenhouse's own product page says the Notetaker is available on all three new tiers and lets admins set retention up to 60 days.

If those terms apply to your order form, Greenhouse has a small interview meter sitting on top of a volume-priced plan. That is the beginning of the same drift, just slower. The tracker also reports that Greenhouse's subscription agreement allows fee increases at renewal with no stated cap. Read that alongside the concessions buyers can win on usage-based contracts before you sign.

Where this comparison is weakest

This post compares meters, not prices, because prices are not public. That is a real limitation, and it cuts in more than one direction.

No dollar value per credit

Without a published price per credit, 105,000 credits could be trivial or material. A large Workday customer's complimentary allowance might cover the whole worked example. A figure for the dollar value of a credit circulates on a document-sharing site, but it cannot be traced to Workday, so this post does not use it. Until you see your own order form, the threefold spread is a spread in exposure, not in spend.

The rate card is reported, not read

The 6 and 750 credit figures come from CIO's reporting of the rate card as it stood on 21 May 2026. Workday links the rate card from its Flex Credits page, but the copy retrieved for this post could not be parsed. Credit values may have changed since May. Consultancies that negotiate Workday contracts also report that rate card values for the same skill differ between customers, which, if true, makes any public figure a starting point rather than a price.

The case for usage meters is stronger than it looks

Monroy's argument has merit. A headcount meter charges a 10,000-person company that hires 50 people a year far more than a smaller company that hires 2,000, regardless of how much recruiting work either does. A usage meter fixes that mismatch. For low-volume hirers with large headcounts, Workday's credit model may well be cheaper than any seat or employee price for the same capability.

Terra Higginson of Info-Tech Research Group told CIO that seat pricing and seat counts are not going away. That supports the reading here, that buyers will carry both meters for years, rather than one replacing the other. The weakness of my argument is that it treats volatility as the main cost. For some buyers, paying for unused headcount capacity is the bigger waste.

What to put in the recruiting software contract

If the meters cannot be aligned, the contract can at least make them visible. These are the clauses I would ask for, in order of how much exposure they remove.

The strongest single ask is the first one. A fixed credit rate turns the Workday agent from an open exposure into a forecastable unit cost per resume, which your talent team can then manage through requisition design. Without it, every one of the scenarios in the worked example is a guess.

This is part of a wider shift that is compressing seat-based SaaS pricing across categories. Recruiting is simply the first HR module where you can watch it happen inside one hiring workflow. How Workday is using credits to hedge its own seat exposure is covered in the analysis of Flex Credits as a seat hedge.

Frequently asked questions

How to price your next 100 hires before renewal

Pull last year's recruiting data this week: number of requisitions, total applicants and total hires. Those three numbers are all you need to translate any vendor's meter into your own exposure, and almost no buyer has them on one page when the quote arrives.

Then send the same file to every vendor on your shortlist and ask for a priced quote against it, with the billing unit named in writing. A vendor that will not price your real history is telling you which meter it expects to grow. If skills testing is also on your roadmap, the assessment side of the funnel has its own economics, covered in the review of skills-based hiring assessments.

Related reading