Skip forward into hiring notes
RosterAcorn

Vendor hiring-tool claims: 5 checks before you buy

Hiring software vendor claims are only useful when you can trace each figure to a source, a date and a measure.

Key takeaways before you buy

  • check whether a customer result describes the same work your team does.
  • check any Hackett rating against its scope of 12 providers and 28 criteria before you cite it.
  • treat Workable's 1000+ and 50+ counts as library sizes, not as time saved.
  • ask Indeed for the amount behind the Sponsored Job budget before you commit any sponsored job spend.

Lever reports customer results and faster hiring on its pages about customer results and time to fill. Those figures come from Lever itself, so each one needs a check before it reaches your notes.

Workable says its figures count reusable HR text and recruiting emails, so read them as library sizes rather than hiring results. Keep them out of any time-saving estimate you make for your team.

Why vendor figures need a source check

Before a vendor figure shapes a purchase, find out who measured it, when it was measured, and what it measured. Copy the full sentence from the vendor page into your notes, with the page address beside it. Keeping the exact wording lets you compare offers against the same text rather than a memory of a sales call. Mark any figure that lacks a date or measure, so you can ask about it later.

A figure can match its source and still say little about your own hiring. A customer report describes what customers reported, and a rating describes the scope its assessors used. Read each claim as a statement about that source, and write down the subject it covers. Treat any mismatch in subject as a question for the vendor.

Put a date beside every figure. The Hackett rating quoted in this guide belongs to the 2026 Talent Acquisition Vendor Assessment, according to Lever. Ask the vendor which date a customer figure reflects. If you cannot find a date, write that gap down rather than filling it in with the day you read the page. Save the day you read each page as well.

Ask each vendor to name the unit behind every figure. A unit tells you what a number is counting, so a figure about reusable text, customers or ratings should carry that exact label in your notes. If the vendor cannot name the unit, leave the figure out of your comparison and record why. A missing unit is a question to keep open rather than a reason to choose.

Keep a note for each vendor, with a line for each figure that holds its source, date and measure. Figures with all three can go into your shortlist discussion. Figures missing any of them stay open until the vendor answers. Pull your own hiring records beforehand, so you have a baseline for comparison. Use an interview scorecard example to keep interview records consistent with that comparison.

Use this record format to keep a question beside its evidence.

RecordSourceDateMeasureFollow-up
ExampleSaved page addressDate question remains open Customer-reported resultAsk which work was measured

Read the 73% customer result

73% improvement in recruiting automation

Lever customers reported a 73% improvement in recruiting automation and a 60% improvement in recruiter efficiency, according to Lever. These are separate figures with separate subjects, so each needs its own line in your notes. Read the word customers closely, because the figure describes what customers reported rather than a result your team will record.

Ask Lever which customers and which time period the 73% figure covers, and write the reply beside the number. A short written answer is enough, since you need the group of customers and the period behind it. If the reply is vague, mark the figure as unconfirmed until you hear back. Store the reply with the source link, so the figure never travels without its context.

The 60% figure in the same sentence refers to recruiter efficiency, a different subject from recruiting automation. Keep the figures apart in your comparison, because each one answers a separate question about the software. Do not add the percentages together or average them into one score. Each figure stays on its own line with its own source.

A small staffing team may work differently from the customers behind that report, so test the tool on your own job orders before you rely on the number. Use the same wording in your test notes as the vendor figure, so the two can be compared line by line.

Write a line in your notes with the source, the date, the measure, and a yes or no on whether you tested the figure yourself. Mark the test result as untested until you have run the tool on real job orders. That label is a reminder that the claim is still open.

Check the 2.5x time-to-fill claim

2.5x faster time to fill

Lever's wording is that "teams are seeing 2.5x faster time to fill and cut manual screening time by 70–90%", according to Lever. That sentence joins a speed claim and a screening claim in a single line, so split them apart in your notes. Read the speed part as a description of what teams are seeing, not as a benchmark for your team.

Write down how your team starts and stops the clock on time to fill, then ask the vendor to use the same start and stop points. If the vendor measures a different span, note the gap next to the figure rather than comparing the two numbers directly. That note shows where the two measures differ.

Ask Lever which teams and which time period the 2.5x figure covers. Write that limit next to the figure before you share your comparison with a colleague.

The screening result needs its own question about the period measured. Ask the vendor what period the screening figure covers, and record the reply next to it.

Keep the vendor's reply beside the figure, and pause if the reply names no date or measure. A reply that gives the source and date but leaves out the measure still leaves the speed claim open. Your own time-to-fill records are the check the figure can be compared against, so pull them before the next vendor call.

Illustrative example: compare matching clocks

Suppose your records show an illustrative 50 days to fill a role, while a tool test takes an illustrative 10 days. Before comparing those results, check that both clocks start and stop at the same points. If they do not, record the difference and keep the comparison open. This test describes your records; it does not confirm a vendor's customer result.

Test the 2026 Hackett rating scope

For its 2026 Talent Acquisition Vendor Assessment, analysts assessed 12 talent acquisition software providers across 28 criteria, Lever reports. That describes a review of software providers across criteria, so treat it as a scope to read, not a measure of hiring outcomes. Read the scope before you look at any rating that came out of it.

Keep each rating with its scale, so a 4 out of 5 is never read without its category.

Lever also earned a Leading 4 out of 5 rating for AI & Emerging Technologies and 10 Top Performer badges, including recognition for Analytics and Platform Flexibility, as Lever reports. Keep the badges on a separate line from the rating in your notes. Ask which of them matter for the job you are filling, and set the rest aside.

Ask the vendor which criteria produced the rating, and whether the 28 criteria cover the product you plan to buy. The reply tells you how closely the rating fits your use. Record the reply beside the rating, along with the date of your conversation. A reply in writing gives you a record to compare later.

Treat the rating as a comparison point among the 12 providers the assessment covered, not as a result for your team's hiring.

Count what Workable's library numbers measure

1000+ reusable HR texts to choose from

Workable offers 1000+ reusable HR texts. Read the 1000+ as the size of a set of choices, not as hires made or days saved.

Ask Workable how the texts in the 1000+ count are grouped, and whether you can search across them before you start a role. Write the answer beside the count, so the number keeps its context. Keep the question on your list until the vendor answers it in writing. Your list should note the date of each question.

Plan time for your own edits to any text you adopt, since the count shows a size and not a measure of fit. In your notes, count the texts you actually use, so the 1000+ figure is never treated as the number you will use. Keep your own edit time on a separate line, so it stays visible when you compare vendors.

50+ recruiting email drafts

Workable also offers 50+ recruiting email drafts and describes creating an email in seconds. Do not treat the words "in seconds" as a measured time saving, because the figure counts email drafts.

Write the 50+ count next to its exact wording, so the number is never read as a speed result. Then ask the vendor what the count includes, such as which types of recruiting email it covers. Compare the email count with the other library counts only after you have the answers in writing.

Ask what sponsored job charges cover

Indeed states that upon expiration of the credit, users are charged based on Sponsored Job budget. That sentence ties the charge to a budget and to the moment a credit expires. Write these conditions into your notes, so the charge can be traced to its trigger. Keep the wording exact, since a paraphrase can drop a condition.

Ask Indeed to explain how the Sponsored Job budget is set for your account, and request the reply in writing before a credit expires. A written reply gives you a record to check against later charges. File the reply with the quoted sentence so the two stay together. Note the date of the reply beside it.

The source is an Indeed guide on recruitment yield ratios, so its sentence about sponsored job charges belongs in your notes as a stated rule rather than a quoted amount. Treat that sentence as a description of how charges work, and nothing more. Compare it with any written terms you receive, and note any mismatch.

Keep the credit terms and the sponsored job terms in separate lines of your notes, so each one can be checked on its own. Ask for any amount in writing rather than assuming one, and note the date of that reply. If a figure has no written source, leave it out of the cost comparison. A short, dated note is enough for this check.

Compare that reply with the credit wording before you commit any budget, so the two sentences agree. Keep this comparison with your other vendor figures so the set stays in one place. Use hiring decision debrief notes when recording how the comparison informed your decision.

Today, choose a figure you plan to cite and email the vendor to ask for its source, date and measure. Hold off on citing any figure that comes back without an answer.

Hiring tool claims: questions and answers

What did Lever customers report?

Customers reported improvements in recruiting automation and recruiter efficiency, as Lever says. Check them against your own hiring records before you rely on them. Keep the source and date beside each figure in your notes.

How many providers did the Hackett assessment cover?

Lever says the assessment covered 12 providers. Ask the vendor which providers and criteria apply to the tool you are considering. Record that reply beside the rating in your notes.

What do Workable's 1000+ and 50+ numbers count?

Workable says they count reusable HR text and recruiting email drafts. Ask the vendor what each count includes before you compare libraries.

How are Indeed sponsored jobs charged?

Indeed says charges depend on the Sponsored Job budget after the credit expires. Record the date of the budget reply beside the figure.

What did Lever receive in the Hackett assessment?

Lever says it received a category rating and recognition badges. Ask the vendor which rating and badges apply to the product you would buy.

Does Lever's 73% figure apply to my firm?

As Lever notes, the figure describes customer reports rather than a result established for your firm. Test the tool on your own job orders, and compare the results with the vendor figure before you decide.

Using AI software to screen job applicants? AI Hiring Paperwork Pack is a workbook filled from your answers that shows which rules reach each tool and the wording applicants should get.

Sources