White-hat link building, scoped to the algorithms that score it.
Manual outreach with relationship-led editorial placement. Anchor allocation per URL. SpamBrain footprint avoidance at the prospect-list layer. Placements meant to compound through the next algorithm update, not reverse out inside 12 to 18 months.
Editorial placement, scoped per URL, paced against decay.
The link-spam-detection layer compounds every algorithm update. White-hat methodology means each placement clears the editorial threshold and the campaign architecture renders the same fingerprint as organic outreach.
Editorial placement, never rented inventory.
Every placement is editorial. The agency does not maintain a rented-link inventory, a private blog footprint, or any other link-scheme apparatus. Outreach is manual, content-fit is verified against the publisher's editorial guidelines, and the placement either lands on its merits or it doesn't. The volume cap is a feature; the link-spam-coded models that scale past it trigger the algorithmic detection layer.
Anchor-text distribution scoped per URL.
Each landing page gets its own anchor allocation against the per-vertical Penguin 4.0 baseline. Exact-match commercial anchors stay under 15 percent of the URL's inbound profile. Branded anchors run 40 to 55 percent. URL and topic and naked-link anchors fill the rest. The allocation lives in the campaign-scope document and the prospect-list assignment maps anchor recommendation to placement target.
SpamBrain footprint avoidance at the outreach layer.
Mailbox identity segmented per campaign cohort. Outreach paced against the 30-day moving average the link-graph network evaluates over. Prospect set vetted against TrustRank seed-set proximity. The neural detection layer reads patterns at the source-domain plus destination plus anchor plus timing level; the campaign architecture renders the same fingerprint as organic editorial outreach.
Decay-net placement counts in the retainer scope.
10 to 20 percent annual link decay is the baseline industry attrition. The retainer scope names a net placement target after decay, not a gross delivery number. Quarterly review surfaces lost placements and refreshes the prospect list with new vertical-cluster surfaces. The branded query volume movement is the secondary signal we track against the Panda-patent ratio surfaced in the May 2024 Content Warehouse leak.
A quarterly white-hat link-acquisition campaign, audit to placement.
Inbound link profile segmented and scored.
We pull the full inbound profile (Ahrefs + Google Search Console + Majestic for triangulation), segment by anchor category and topical-cluster proximity, model the URL-level distribution against the per-vertical baseline, and identify the placements passing signal versus those discounted, neutral, or actively dragging. Disavow candidates flagged. Anti-Trust Rank exposure surfaced. Output is the diagnostic spec the campaign builds against.
Vetted prospects scoped to the campaign target.
Prospects sourced from competitor backlink analysis, vertical-citation layer (Avvo / Healthgrades / 1-800-Dentist depending on vertical), curated resource pages, broken-link surfacing via Wayback Machine and Check My Links, and HARO query streams across the Featured.com platform. Each prospect vetted on Domain Rating, traffic, topical-relevance overlap, outbound-link density, and SpamBrain footprint risk. The list typically runs 200-400 prospects per quarter.
Manual outreach paced against decay.
Outreach runs through segmented mailbox identity per campaign cohort. Cold outreach converts at 5 to 15 percent. Warm outreach to existing relationships converts at 30 to 50 percent. HARO pitches convert at 3 to 8 percent pitch-to-link. Resource-page acquisition runs 8 to 15 percent. Broken-link building runs 5 to 12 percent. Placements scoped against measured monthly decay, with anchor allocation handled at the URL level rather than the root-domain average.
Net placement target sustained.
10 to 20 percent annual decay tracked at the placement level. Quarterly review surfaces lost placements (publisher edits, page removals, dofollow-to-nofollow flips), refreshes the prospect list with new vertical-cluster surfaces, and re-models anchor distribution against the updated profile. Branded query volume tracked as the Panda-patent ratio input that supports inbound link velocity.
Methodology questions we get during the audit conversation.
What makes a link white-hat vs. grey-hat?
The distinction is editorial vs. transactional. Editorial placement is a publisher choosing to link to a page because the content fits an editorial decision. Grey-hat placement is paid or coerced placement that the publisher would not have made on merit (paid guest posts without rel="sponsored", niche-edits at scale, citation-pack networks). Penguin 4.0 evaluates the anchor profile at the URL level in real time and discounts placements whose pattern reads as paid. SpamBrain reads the link-graph network at the network level and demotes domains whose inbound profile fingerprints as scheme-coordinated. Both algorithms reward placements that meet the editorial threshold.
What's a realistic link target for a quarterly retainer?
A quarterly white-hat retainer typically targets 6 to 14 placements net of decay across the campaign mix (link acquisition, digital PR, citation building, HARO). The exact target depends on the inbound profile diagnosed in the audit and the vertical-citation gap surfaced for the niche. Below 4 placements per quarter the maintenance retainer doesn't compound against the 10-20 percent annual decay baseline. A serious off-page SEO agency scopes the placement target net of the 10-20 percent decay attrition rather than gross delivery, because the gross number reverses inside 18 months as the decay surface absorbs the wins.
How do you handle a domain with manual-action recovery work pending?
Recovery work runs first, link acquisition second. A domain carrying an active manual action or a SpamBrain demotion has its inbound profile audited for the placements driving the signal. Disavow file assembled and submitted. Reconsideration request drafted if a manual action applies. New link acquisition pauses until Search Console signals the demotion has lifted. Building over an active signal compounds the demotion rather than recovering it.
How is anchor text decided per placement?
Anchor recommendation maps to the placement target at prospect-list assembly time. The URL accumulating the inbound link gets its current anchor profile pulled and the new anchor selected from the bucket that's underweighted (branded if exact-match is approaching the ceiling, topical-variation if branded is saturated, etc.). The publisher gets a recommended anchor in the pitch and the final placement either uses it or the publisher's own editorial choice; either way the anchor is recorded against the URL profile and the next placement adjusts.
The exact-match ceiling is 15 percent per URL. Most operators we audit are at 40.
The audit pulls the inbound profile, segments by anchor and topical-cluster proximity, models the URL-level distribution against the per-vertical baseline, and names the placements Penguin 4.0 is discounting in real time. Inside two weeks.