If you apply through Indeed, some employers now see a 0 to 100 Smart Fit Score on your application, plus a written summary of your strengths and "areas to review." That weakness line is built only from what you submitted, so you can write it yourself before a recruiter ever reads it.
The score is part of Indeed's Premium Plus tier, which is still in beta. Money.com's November 2026 roundup flagged it around the same time Indeed Hiring Lab reported the first year-over-year gain in US postings in nearly four years. More postings mean more of your applications will pass through this ranking layer.
What is Indeed's Smart Fit Score?
The Indeed Smart Fit Score is a 0 to 100 rating that Indeed's AI gives applicants, measured against the employer's screening criteria. It comes with a candidate summary "highlighting fit, strengths, and areas to review," and it's only available to Indeed Premium Plus employers, a tier Indeed's own pricing page labels Beta.
What's public so far:
- Tier: only Premium Plus has it. The three lower tiers on Indeed's pricing page don't.
- Price: Indeed hasn't published one. For comparison, sponsored job budgets start at $5 a day or $150 a month per job.
- Reach: Money.com puts Indeed at more than 3.5 million employers and 685 million job seeker profiles.
- Launch context: Indeed showed the score at FutureWorks 2026, next to Smart Screening and credential checks.
What isn't public: how many employers are in the beta, how scores are distributed, or whether you can see your own score. Indeed hasn't published any of that. Don't trust anyone who claims to know your number.
How employers rank applicants on Indeed under Smart Screening
Under Smart Screening, deal-breaker questions filter applicants first. Then a weighted score is built from your resume and screening answers, and an AI Recruiter asks follow-up questions about criteria your resume didn't prove. Aptitude Research's June 2026 analysis described Smart Screening as moving the phone screen into the application itself.
Here's the sequence, based on Indeed's employer documentation:
- Criteria get extracted. Indeed pulls screening criteria from the job description automatically. The employer can edit them and mark each one must-have or nice-to-have.
- Deal-breakers filter. Employers set must-pass questions. If you fail one, you can get a rejection and never receive a score.
- The score gets built. It uses only applicant-supplied details: resume, screening answers, location and similar.
- The summary gets written. It covers fit, strengths and areas to review.
- The AI Recruiter follows up. It asks questions by chat, phone or video, in whatever formats the employer allows, and skips the ones your resume already answers.
Indeed tells employers Smart Screening takes 26% fewer applications to produce a hire. Put plainly, the pitch is that recruiters get to read less. Whatever the summary says about you may be most of what they read.
Your weakness line is written by omission
The "areas to review" line is mostly a list of requirements you never backed up with evidence, because the score sees nothing except what you submit. Indeed says the Smart Fit Score uses no third-party data and doesn't compare you with other candidates.
That design sounds fair, and in one way it is: you aren't being judged against a stranger's resume. But look at the mechanism. If the model can only read your submission, then a skill you have and didn't write down is indistinguishable from a skill you lack. Your portfolio, your recommendations and the manager who would vouch for you don't exist to the scorer unless they show up in the resume or a screening answer.
Indeed's own example shows the stakes. If management is a must-have and budgeting is a nice-to-have, missing management "will have a bigger impact on the score." So if you led a team but described it as "coordinated cross-functional work," you've turned a strength into a gap through phrasing alone.
A skill you have but never wrote down looks exactly like a skill you lack.
That's why the headline says to write your weakness line first. Before you apply, list the must-haves you can't point to in your own document. That list is your likely "areas to review" line. Either fix each item with evidence or accept it knowingly.
The job description is the scoring sheet
Indeed pulls criteria straight from the job description and weights must-haves above nice-to-haves, so tailoring is now arithmetic, not style. One missing must-have can cost you more than several missing nice-to-haves.
Smart Fit works like a resume match score, except the weights come from the posting's own language. Work through it like this:
- Split the posting. Copy every requirement into two columns: Required (must have, minimum, X+ years) and Preferred (nice to have, bonus, plus).
- Find proof for each required item. One line per item: a title, a scope, a tool, or an outcome from your actual history.
- Use the posting's noun. If it says "stakeholder management," don't write "partnered with leadership" and hope the model makes the connection.
- Put required proof high. Your most recent role and your summary are where a reader looks first.
- Add preferred items only where they're true. The weight is low, and a padded claim invites a follow-up you can't answer.
| Requirement type | Weight in score | Where your proof belongs |
|---|---|---|
| Must-have | Higher | Summary and most recent role, in the posting's words |
| Nice-to-have | Lower | Wherever it's genuinely true, briefly |
| Deal-breaker question | Pass or fail, before scoring | Screening answer, read literally |
| Location and logistics | Scored as submitted | Profile and screening answers, consistent |
Doing this by hand for every posting is the tedious part, and it's the exact work Refolk takes off you. Paste the posting, and Refolk rewrites your own history against it with the required items proven first. Then it scores how well you actually fit, so you see the gap before Indeed's summary does.
A resume that proves the must-haves shortens the AI interview
Every must-have you prove in writing is a follow-up question you won't have to answer live. Indeed's documentation says the AI Recruiter's questions come from the scoring instructions, and "certain questions will be skipped if evidence can be found in the resume."
A thin resume doesn't only score lower. It also triggers more questioning by chat, phone or video. Each live answer is another chance to:
- Contradict a date or title on your resume
- Undersell something under time pressure
- Give a vague answer that the summary then quotes back as a weakness
Indeed's job seeker explainer on Smart Screening is the one official source written for applicants, and it's worth reading before you apply anywhere that uses it. The takeaway from the employer docs is simpler, though. The best way to handle the AI interview is to make most of it unnecessary.
Deal-breaker questions come before any score
For some applicants, the Smart Fit Score is never calculated at all. Employers set must-pass questions, only applicants who pass get a score and summary, and Indeed can send a rejection when an answer fails.
No amount of resume work helps if you misread a yes/no knockout. The common traps:
- Licenses and certifications. Answer the question as written ("currently hold" is not "in progress").
- Location. Know whether it's asking where you live now or whether you'll commute or relocate.
- Work authorization. Read whether it asks about now or about future sponsorship.
- Years of experience. Count the way the question defines it, not the way your resume summary rounds it.
One more point to check yourself. A third-party review by Culturro reads Indeed's terms as saying Smart Screening switches on for all current and future jobs on an employer's account, with the AI Recruiter on by default. That's one reviewer's interpretation, and Indeed hasn't confirmed it in the sources I found. If it's accurate, you'll hit knockouts and follow-ups on postings from employers who never consciously chose them.
Who actually reads the Smart Fit Score summary
The summary goes to recruiters, hiring managers and small-business owners, and few of the recruiters list tooling skills. In Refolk's index of professional profiles, 137,924 US profiles currently hold Recruiter, Talent Acquisition Specialist or Technical Recruiter titles, and only 9,133 of them list Applicant Tracking Systems as a skill.
| Segment | Count | Note |
|---|---|---|
| US recruiters, TA specialists, technical recruiters | 137,924 | Refolk's index |
| UK, same titles | 9,824 | Likely undercount; excludes "Recruitment Consultant" |
| US, same titles, listing ATS skill | 9,133 | Refolk's index |
| US-to-UK multiple | ~14.0x | Directional, given the title caveat |
| Share of US recruiters listing ATS skill | ~6.6% | 9,133 / 137,924 |
The most common current US title in the sample is Technical Recruiter. Employers in the sample include Experis and K2 Partnering Solutions, plus Snowflake and Faire in the ATS-skill subset.
9,133 of 137,924 profiles holding Recruiter, TA Specialist or Technical Recruiter titles.
What follows is inference, not measurement. If so few recruiters list ATS skills, and many of Indeed's 3.5 million employers are hiring managers and owners with no recruiter at all, it's plausible that a lot of readers take the summary at face value. Indeed itself tells employers to use their own judgment "regardless of their Smart Fit Score," which tells you what they expect some people to do otherwise. Culturro's list of questions for buyers asks whether anyone outside Indeed has audited the score for bias. I found no public audit.
So write for a literal reader. Assume the "areas to review" line will be read as fact.
Why AI candidate scoring matters more as postings recover
Postings are growing again, so more applications will pass through scoring layers like this one. Indeed Hiring Lab reports US postings up 3.1% year over year to October 23, the first positive year-over-year trend in nearly four years.
The first positive year-over-year trend in nearly four years, per Indeed Hiring Lab.
The recovery is uneven. Software development postings are up 23.8%. But more than a quarter of unemployed workers had been searching for six months or more in October, unemployment duration is at its highest since December 2021, and Hiring Lab says employers have "tilted their hiring toward experience." Economist Laura Ullrich has pointed to stronger demand for senior talent. When employers are picky, a weighted rubric that punishes missing must-haves is a natural tool for them to reach for.
And the silence is already the norm. Indeed's own 2025 FutureWorks material says almost 75% of job seekers never hear back after applying. Gem's 2026 benchmarks, as summarized in a third-party review, show job boards and company marketing bringing in about 90% of applications but only about half of hires, while sourced candidates are about 8x more likely to be hired. Even a perfect fit score keeps you in the channel that produces most of the noise.
If you do apply through Indeed, keep your profile current. Indeed says job seekers with detailed, current profiles are 2.8x more likely to appear in employer searches and 4.5x more likely to be contacted.
FAQ
Can I see my own Indeed Smart Fit Score?
There's no public evidence that you can. Indeed lists the score and summary as features of an employer pricing tier, Premium Plus, and nothing published says applicants get to see their number. The practical substitute is to audit the posting yourself: any required item you can't point to in your resume is a likely "area to review."
Does Indeed compare me against other applicants?
Indeed says no. The Smart Fit Score is built from your own resume, screening answers, location and similar details, with no third-party data and no comparison to other candidates. An employer reviewing a stack of applications may still see your number next to other people's, even though it wasn't calculated against theirs.
Will stuffing keywords raise my score?
Not in a way that holds up. The score weights must-haves, and the AI Recruiter asks follow-ups about criteria the resume doesn't clearly prove, so a keyword without evidence behind it tends to invite a question rather than skip one. Use the posting's exact terms, but attach each one to a real title, scope or outcome. The summary is written for a human who will read it as fact.