How to choose a link building agency, scoped against the vetting surface a sophisticated buyer reads.
Most buyers evaluate link building agencies on deliverable counts and price. The vetting questions that separate a peer-credible white-hat agency from a per-placement vendor read the methodology layer: per-URL anchor allocation against the Penguin 4.0 baseline, SpamBrain footprint-avoidance discipline at the prospect level, FTC §255 disclosure handling at the placement level, link decay accounting in the retainer math, and brand-strength signal awareness alongside link acquisition. The seven-criterion framework below is the checklist to send before the discovery call.
Seven criteria, side-by-side against the per-placement vendor pattern.
- Anchor allocation per URL
- Names the per-URL Penguin 4.0 baseline; allocates anchors against vertical exact-match ceiling (5 to 15 percent)
- DR with topical-relevance overlay
- Filters prospect lists by Domain Rating AND topical-cluster proximity; names co-citation as the signal pathway
- FTC §255 disclosure handling
- Reviews paid placements against FTC §255 (2023 revision) material-connection disclosure; verifies rel="sponsored" or rel="nofollow" attribution; treats Google ranking-hint shift (March 2020) as policy, not optional
- SpamBrain footprint avoidance
- Avoids placement sources fingerprinted by SpamBrain (shared C-class IP, content-template overlap, intra-network density, whois-history overlap)
- Link decay accounting
- Scopes monthly placement target net of the 10 to 20 percent annual decay baseline; tracks dofollow-to-nofollow flips, page removals, publisher edits
- Reactive-sourcing surface awareness
- Names the HARO discontinuation (early 2024), Connectively shutdown (December 2024), and Featured.com HARO revival (April 2025) on the timeline; monitors Featured.com and Qwoted in parallel
- Brand-mention and branded query awareness
- Names the May 2024 Content Warehouse API leak and brand-strength scoring derived from branded query volume; scopes digital PR against the Panda-patent ratio
- Outreach response benchmarks
- Quotes cold outreach 5 to 15 percent response rate, warm 30 to 50 percent, reactive-sourcing 3 to 8 percent pitch-to-link, broken-link 5 to 12 percent, resource-page 8 to 15 percent
- Anchor allocation per URL
- Surfaces site-wide allocation against the per-URL baseline. Discusses the natural-distribution baseline (70 percent or more branded, URL, topic, naked link)
- DR with topical-relevance overlay
- Surfaces topical relevance as a load-bearing filter. Names why a DR-30 link in a topically-coherent vertical publication outperforms a DR-60 link from a generic news aggregator
- FTC §255 disclosure handling
- Names the FTC §255 "clear and conspicuous" standard and the "overall net impression" evaluation. Confirms publisher applies the rel="sponsored" attribute on the placement
- SpamBrain footprint avoidance
- Names SpamBrain (2022 neural link-spam detector) and the specific footprint signals. Surfaces the prospect-vetting layer rather than the placement-marketplace layer
- Link decay accounting
- Quotes a quarterly net placement target and reports decay alongside acquisition. Treats decay as a recurring cost the retainer absorbs
- Reactive-sourcing surface awareness
- Acknowledges the platform consolidation as a real operational shift. Names which surface the agency monitors and why
- Brand-mention and branded query awareness
- Surfaces brand-mention SEO and co-citation as parallel signals to inbound links. Names branded query volume as a load-bearing input to the ratio
- Outreach response benchmarks
- Names the industry benchmarks derived from Pitchbox and Postaga aggregate data. Quotes conversion expectations grounded in the specific outreach shape
- Anchor allocation per URL
- Offers exact-match anchors as a deliverable. Quotes anchor concentration per page without naming the per-URL baseline
- DR with topical-relevance overlay
- Quotes DR-only thresholds (DR 50+, DR 70+). Treats topical relevance as a soft preference rather than a primary filter
- FTC §255 disclosure handling
- Treats disclosure as the publisher's responsibility. Quotes paid placements without specifying the rel attribute the publisher will apply
- SpamBrain footprint avoidance
- Sources placements from open link marketplaces, "high DR guest post networks," or undisclosed publisher pools. Cannot name the prospect-vetting criteria
- Link decay accounting
- Quotes gross placement count without naming decay. Reports lifetime placement count rather than active dofollow inventory
- Reactive-sourcing surface awareness
- References "HARO" as a generic verb without naming the 2024-2025 platform transitions. Cannot name which reactive-sourcing surface the campaign would route through
- Brand-mention and branded query awareness
- Treats links as the only off-page signal. Cannot name the brand-strength surface or the leaked brand-signal scoring
- Outreach response benchmarks
- Quotes inflated response rates (40 percent or above on cold outreach) or refuses to name benchmarks. Surfaces only placement counts without per-pitch conversion context
VETTING BENCHMARKS · PENGUIN 4.0 · SPAMBRAIN · FTC §255 · CONTENT WAREHOUSE LEAK
Six diagnostic reads, framed as questions to send before the sales call.
The vetting framework reads the agency at the methodology layer rather than the deliverable layer. Each diagnostic below is a question the buyer can send in writing before the discovery call; the agency's written response is the first-pass signal of whether the methodology depth is real.
Ask for the per-URL anchor allocation model, not the site-wide average.
Penguin 4.0 integrated link-spam evaluation into the core algorithm in September 2016 and shifted anchor-text scoring to real-time, URL-specific discounting. An agency that quotes site-wide anchor averages cannot scope the per-URL baseline that triggers the real-time discount. A peer-credible vetting question is "What exact-match anchor ceiling do you scope per URL for our vertical?" The answer should name the 5 to 15 percent vertical-specific ceiling, the per-URL allocation method, and how the agency reads the existing per-URL exposure before recommending new anchors.
Ask which reactive-sourcing surface the campaign monitors and why.
The reactive-sourcing platform layer changed twice in eighteen months. Cision shut HARO down in early 2024 and migrated the audience to Connectively. Connectively shut down in December 2024. Featured.com bought the HARO brand and revived the platform in April 2025. A peer-credible agency knows the timeline, monitors Featured.com and Qwoted in parallel because the journalist audience split during the transition, and can name which queries are running on which platform for the buyer's vertical. An agency that references HARO as a generic verb without naming the platform transitions is reading from a pre-2024 playbook.
Verify the SpamBrain footprint-avoidance discipline at the prospect level.
SpamBrain (Google's 2022 neural link-spam detector) flags PBNs and link farms through shared C-class IP and ASN clustering, whois-history overlap, content-template overlap, intra-network linking density, and rendering-level fingerprinting (shared analytics codes, AdSense IDs, custom CSS). A peer-credible agency names these signals when asked how it sources placements. A per-placement vendor sources from open marketplaces where SpamBrain has already mapped the footprint, treats the placement as the deliverable, and cannot name the prospect-vetting criteria that would have flagged the marketplace publisher's footprint exposure.
Confirm FTC §255 disclosure handling at the placement level.
The FTC §255 Endorsements and Testimonials Guides (2023 revision) require material-connection disclosure on any paid placement that contains an endorsement element. The disclosure must be "clear and conspicuous" per the Dot Com Disclosures guide, judged on proximity to the claim, prominence, absence of distracting elements, and understandable language. Google separately mandates rel="sponsored" or rel="nofollow" on the link attribute, treated as a ranking hint since March 2020. A peer-credible agency reviews paid placements against both layers, names the publisher's disclosure pattern, and verifies the attribute the publisher will apply before the placement runs. An agency that treats disclosure as the publisher's problem exposes the buyer to "Unnatural links to your site" manual action risk and FTC compliance exposure.
Verify the decay-aware retainer math, not the gross placement count.
Acquired links decay at 10 to 20 percent per year as publishers edit pages, remove content, or flip dofollow to nofollow. A retainer scoped against gross placement count loses signal each quarter as the back-catalog erodes. A peer-credible agency scopes the monthly placement target net of the decay baseline and reports both gross acquisitions and the net active dofollow inventory. Vetting question: ask for the previous quarter's gross placements, the count of those placements that flipped or removed, and the net active inventory at quarter close.
Verify the brand-strength signal awareness alongside link acquisition.
Google's May 2024 Content Warehouse API documentation leak surfaced brand-strength scoring tied to branded search query volume. A domain accumulating inbound links without generating branded queries triggers the Panda-patent ratio that flags the link profile as unnatural. A peer-credible agency scopes digital PR and brand-mention work alongside link acquisition because the branded query volume mathematically supports the inbound link velocity. An agency that treats links as the only off-page signal cannot read the brand-strength surface and ships campaigns that look unnatural to the ratio even when every individual placement is clean.
From vetting questions to single-quarter pilot.
Send the vetting questions before the sales call.
Email the seven vetting questions above to the agency before the discovery call. A peer-credible agency answers in writing with methodology-grounded specifics. A per-placement vendor either ignores the questions, sends a sales deck in response, or sends generic answers stripped of the actual benchmarks. The written response IS the first-pass signal.
Ask for a sample link audit on a non-confidential domain.
A link audit reads the inbound profile, segments by anchor category and topical-cluster proximity, surfaces per-URL exact-match concentration against the baseline, and names the placements that are passing signal versus the placements Penguin 4.0 is discounting. A peer-credible agency can produce a sample audit on a competitor or public domain. The output reads as a diagnostic document, not a sales pitch.
Verify references inside the buyer's vertical.
Off-page work is vertical-specific. Legal carries state-bar advertising review overhead. Medical carries YMYL and FDA-adjacent compliance. Penalty recovery requires conservative-velocity rebuild discipline. Ask for references inside the buyer's vertical, not generic SaaS references. A peer-credible agency provides references that name the vertical-specific compliance overhead the engagement carried.
Scope a single-quarter pilot against measurable signal movement.
Quarterly retainer engagements compound across quarters because the journalist and editorial relationship layer rebuilds the warm-list conversion math. A single-quarter pilot establishes the agency's operational discipline (anchor allocation, footprint avoidance, decay accounting, FTC §255 compliance) and produces enough signal to evaluate the methodology fit before committing to annual scope. The pilot should ship the audit, the campaign mix scope, and the first quarter of placements with the net-of-decay reporting.
Methodology questions we get during the audit conversation.
What separates a peer-credible link building agency from a per-placement vendor?
The vetting surface comes down to whether the agency prices the campaign architecture or just the placement. A peer-credible agency scopes anchor allocation per URL against the Penguin 4.0 baseline, vets prospects against SpamBrain footprint signals, handles FTC §255 disclosure review at the placement level, accounts for the 10 to 20 percent annual link decay in retainer math, and scopes brand-mention work alongside link acquisition because branded query volume mathematically supports the link velocity. A per-placement vendor sources placements from open marketplaces, quotes anchor concentration without the per-URL baseline, treats FTC disclosure as the publisher's problem, and reports gross placement counts without the decay surface. The full off-page SEO services mix the agency runs against (link acquisition, digital PR, citation building, HARO, resource-page placement) is the signal of campaign-architecture pricing rather than per-placement pricing.
What anchor-text distribution should a link building agency target?
A natural inbound anchor profile across a domain runs 70 percent or more branded, URL, topic, and naked-link anchors. Commercial-intent exact-match anchors run between 5 and 15 percent of total inbound links, scoped per URL rather than at the root-domain average. The Penguin discount triggers when commercial exact-match concentration on a specific URL exceeds the per-vertical baseline. A peer-credible agency allocates anchor text per URL against this baseline; an agency that quotes site-wide anchor averages cannot scope the per-URL exposure that Penguin 4.0 evaluates in real time.
How does a link building agency price the campaign?
Quarterly retainer pricing is the standard structure. The retainer absorbs the campaign architecture: link audit, prospect-list assembly, outreach paced against SpamBrain footprint avoidance, anchor allocation per URL, FTC §255 compliance review, citation-layer maintenance, decay tracking, and quarterly reporting. The placement target net of the 10 to 20 percent annual decay baseline is the deliverable. Per-placement pricing covers the placement and the vendor's margin without the campaign architecture; for one-off placement need on a known outlet, per-placement is structurally cheaper. For a sustained campaign that has to clear Penguin 4.0 and SpamBrain across the quarter, the campaign-architecture overhead is the load-bearing layer the retainer prices for.
How long does a link building engagement take to produce ranking movement?
The first quarter ships the audit and the initial placement batch. Ranking movement typically surfaces in the second and third quarters as the placements accumulate, Penguin re-evaluates the per-URL anchor profile, and the brand-strength signal compounds through branded query volume growth from digital PR. The compounding curve depends on the vertical (legal and medical run longer than SaaS because the editorial-surface conversion-rate baseline runs lower), the existing inbound profile (penalty recovery runs longer than greenfield), and the campaign mix. A peer-credible agency scopes signal expectations against the vertical and the inbound starting state rather than promising universal timelines.
What questions should I ask a link building agency about FTC compliance?
Three questions read the FTC §255 surface specifically. First: "How do you verify the publisher's disclosure on paid placements?" A peer-credible answer names the "clear and conspicuous" standard and the proximity, prominence, and understandable-language criteria. Second: "Which rel attribute do you confirm on the placement and when?" The answer should name rel="sponsored" for compensated placements and rel="nofollow" as the fallback, with the verification happening before the placement runs, not after. Third: "How do you handle a placement where the publisher refuses to apply the attribute?" The peer-credible answer is the agency walks away from the placement; the per-placement vendor ships it anyway and exposes the buyer to "Unnatural links to your site" manual-action risk.
How do I evaluate an agency's case studies and references?
Vertical-specific references carry the load. Off-page work in legal, medical, dental, real estate, penalty recovery, and SaaS each runs against different compliance baselines and editorial-surface conversion rates. Ask for references inside the buyer's vertical and ask the references about the agency's operational discipline (anchor allocation, footprint avoidance, decay accounting, FTC §255 handling) rather than just the placement count. A peer-credible reference can name the vertical-specific compliance overhead the engagement carried; a sales-deck reference names the placement count without the methodology context.
The audit produces the vetting evidence the framework reads against.
The audit reads the inbound profile against the seven vetting criteria, names the per-URL anchor exposure, the SpamBrain footprint signals in the existing back-catalog, the FTC §255 compliance state, and the decay rate against the current acquisition cadence. Inside two weeks.