Social Screening API Pricing: What It Really Costs Per Candidate Screened

Social media screening costs roughly $2 to $29 per report at published rates in July 2026. That range is real, but it is also the least useful number in this market, because two decisions move your actual bill far more than the unit price does: where in the hiring funnel you screen, and whether you screen once or continuously.

The wider confusion is that "screening" describes at least five distinct products, priced from about seven cents to $175 per candidate. Resume screening, identity resolution, social media screening, traditional background checks and AI interviews all get compared as if they were substitutes. They are not.

Below: what each product actually costs with the arithmetic shown, why the compliance wrapper is the real price driver, the funnel calculation that changes budgets by 500x, and a cost model you can run against your own volumes. This is not legal advice, and the compliance points here need review by your own counsel.

Why does cost per candidate range from $0.07 to $175?

Because five different jobs share one word. Once you separate them, the prices stop looking irrational and start looking like what they are: five markets with different cost structures.

Notice that the two cheapest categories are the two that never touch a regulated consumer report. Resume screening and identity resolution are software problems. Social screening and background checks are compliance problems with software attached, and that is most of the price gap.

What does social media screening actually cost?

Between $2 and $29 per candidate at published rates, a spread of about 14.5x for what looks like similar output. The difference is worth understanding before you optimise for the low end.

One vendor states outright that "competitors charge $15+" while pricing its own check at $2. Read that as a signal about positioning rather than about capability, then look at what the tiers actually gate. The $2 product is an add-on to an AI recruiting suite. The $29 product is a standalone screening report carrying FCRA certification and PBSA membership. Those are not the same purchase and the price difference is mostly the second list.

Why is FCRA the price driver, not the technology?

Because in the United States, using a third party's report to make an employment decision generally turns that report into a consumer report and that third party into a consumer reporting agency, with a long list of obligations attached. Those obligations cost money to meet, and the price of a screening product mostly reflects whether the vendor has met them.

You can see this in the pricing pages directly. Ferretly's free trial is labelled for non-FCRA use cases. Its Professional tier at $29 adds FCRA use cases. Same platform, same AI, different permission to use the output for hiring.

The obligations you are buying, roughly: permissible purpose documentation, candidate disclosure and authorisation, the adverse action process with pre-adverse notice and a waiting period, dispute handling and reinvestigation, accuracy procedures, and record retention. A vendor with FCRA certification and PBSA membership has built all of that. A $2 add-on inside a recruiting suite may not have.

The cheap option that is not an option

Spokeo is widely listed in background check roundups at $19.95 a month and is described in those same roundups as not FCRA-compliant, with data accuracy that varies and coverage limited to US citizens. It is a people search tool.

If you use a non-FCRA source to decline a candidate, the savings are irrelevant next to the exposure. This is the one place in this post where I would say the cheapest option is not a trade-off, it is a mistake. Check the compliance list before the price list.

Why does where you screen matter 500x more than the price?

Because unit price multiplies by volume, and the volume at each funnel stage differs by orders of magnitude. This is the single most valuable calculation in the post and almost nobody publishes it.

Take a role that draws 500 applicants and produces one hire. That ratio is the premise of the whole AI screening category.

Read across the top row and down the first column. Moving the screen from application to conditional offer cuts the cost by 500x. Negotiating the unit price from $29 to $2 cuts it by 14.5x. The stage decision is worth roughly 34 times more than the best discount you will ever get.

And here is the part that makes this easy. The cheap answer is also the compliant answer. Screening at the conditional offer stage is the standard guidance for exactly the reasons FCRA and employment law care about: you have a permissible purpose, the candidate has authorised it, the decision point is defined, and you are not running background data on 499 people who were never serious contenders. Optimise the stage first and the unit price stops mattering much.

How does continuous screening change the maths?

It multiplies rather than adds, and the multiplier is your re-screen frequency. Ferretly gates continuous screening to its Enterprise tier, which tells you it is priced as a different product, and the arithmetic explains why.

Two things follow. First, continuous screening is where per-report pricing stops working and you should be asking for a monitoring rate rather than a report rate. Second, this is the scenario where the difference between $2 and $29 stops being cosmetic: at monthly cadence across 500 workers it is $162,000 a year.

Also check the legal position before you commit to continuous monitoring. Ongoing screening of existing employees raises consent, notice and jurisdictional questions that a one-time pre-hire check does not, and the answer varies by state and by country.

What is the hidden cost nobody prices?

Profile discovery. Before any report can be written, something has to work out which accounts belong to the candidate, and that step carries both the cost and the error rate.

Vendors know it. Ferretly lists agentic AI profile discovery as a headline feature, and it sits in the free tier because it is the thing that makes everything else possible. Trestle sells an address graph with 1.79 billion name-to-address linkages precisely because identity resolution is the hard part. Pipl only charges when a search returns a positive match, which is an unusually honest way to price a step that frequently fails.

In a marketing context, a wrong match is a data quality issue. In screening it is something else entirely: an adverse action taken against the wrong person. That is the risk that should shape your vendor questions, and it is why match confidence, false positive rate and human review on non-clean results are worth more than a lower per-report price. Atlantic sells human verification on non-clean results as its differentiator for exactly this reason.

So budget for the resolution layer separately. At Trestle rates it is roughly $0.09 to $0.28 per lookup, and at PDL Pro roughly $0.28. Small next to a $29 report, and the reason that report is either accurate or not.

How do you model this against your own volumes?

Thirty lines of Python. Run it with your funnel, your cadence and your real screen rate before you take any vendor quote seriously.

# screening_cost.py

def per_hire(applicants, hires, screen_stage_ratio, rate,
             resolution_rate=0.0):
    """screen_stage_ratio: fraction of applicants who reach the
       stage where you screen. 1.0 = everyone, 0.002 = offer only."""
    screened = applicants * screen_stage_ratio
    cost = screened * (rate + resolution_rate)
    return cost / hires, screened

def continuous(population, per_year, rate, resolution_rate=0.0):
    return population * per_year * (rate + resolution_rate)

# --- your funnel ---
APPLICANTS, HIRES = 500, 1
RESOLUTION = 0.22          # identity lookup, priced separately

for label, ratio in [("every applicant", 1.0),
                     ("shortlist", 0.05),
                     ("final stage", 0.006),
                     ("offer only", 0.002)]:
    for rate in (29.00, 2.00):
        cost, n = per_hire(APPLICANTS, HIRES, ratio, rate, RESOLUTION)
        print(f"{label:16} @ ${rate:>5.2f}  "
              f"{n:>6.0f} screened  ${cost:>9,.2f} per hire")

# --- workforce monitoring ---
for cadence, n in [("annual", 1), ("quarterly", 4), ("monthly", 12)]:
    print(cadence, continuous(500, n, 29.00, RESOLUTION))

Change one input and the ranking moves. Set the stage ratio to 1.0 and no unit price saves you. Set it to 0.002 and the difference between vendors becomes a rounding error against a single recruiter hour. That is the actual finding.

Are you buying a report or an API?

Two different products for two different buyers, and the pricing models reflect that rather than competing on it.

If you resell screening, the per-report price is your cost of goods, and that is where this gets uncomfortable. At $29 COGS, reselling at $35 leaves a 17.1% gross margin. At $49 it is 40.8%. Build the report yourself on an API layer and the same $35 sale carries a very different margin. That calculation, not the sticker price, is why BGV firms and ATS platforms buy APIs rather than reports.

Where does Phyllo fit, and where does it not?

We are the data layer, not the report. Phyllo's social screening gives platforms structured signals across social platforms so they can build their own screening product, with identity resolution to establish which accounts belong to a subject, and coverage across 25+ platforms from one integration. It powers background verification for BGV resellers, influencer vetting for brand safety, and visa and immigration checks. We hold GDPR compliance and SOC 2 Type II.

Where we are the wrong purchase, plainly. If you are an HR team hiring a few dozen people a year and you need a finished, FCRA-compliant report you can file, buy from a CRA such as Ferretly, Checkr or Sterling. That is what they are built for and we are not a substitute. We are the right purchase when screening is something your product does, at volume, and you need the data underneath rather than the PDF on top.

And a distinction worth being precise about, because it is the one people get wrong: a consented data layer and a public-data screening report answer different questions. Consent gives you verified depth on someone who has connected an account. Public screening gives you breadth on someone who has not. Screening usually needs the second, which is why our screening products sit on public signals while our creator products sit on consented ones. We wrote up that split in consent-based versus public social APIs, and the underlying data economics in social data API pricing.

What should you ask before buying?

1. Which of the five products is this? Resume screening, identity resolution, social screening, background check or AI interview. Vendors blur this deliberately in demos.
2. Are you a consumer reporting agency, and is this output an FCRA consumer report? Get the answer in writing. If the answer is no, you cannot use it for employment decisions in the US.
3. At what funnel stage does your pricing assume I screen? Then do the arithmetic in this post with your own numbers.
4. What is the price for monitoring rather than for a report? If continuous screening is quoted at the per-report rate, that is not a monitoring product.
5. How is the subject matched, and what is the false positive rate? Ask what happens on an ambiguous match and whether a human reviews non-clean results.
6. Is identity resolution included or billed separately? If it is bundled, ask how it is done. If it is separate, add $0.09 to $0.28 per subject to every quote.
7. What is the international coverage? Several vendors in this market are US-only in practice regardless of what the marketing says.
8. What are the retention and deletion terms? Screening data is sensitive personal data and you will be asked about this in every enterprise procurement you enter.