Quarterly backlink audit checklist (2026, 40 items)

The 2026 quarterly backlink audit is 40 items across 8 categories. The single highest-ROI workstream is lost-link recovery: reclamation outreach converts at 18-26% vs the 8.5% cold-link-building baseline, and a typical mid-market agency leaves $12-15k of already-paid-for link equity on the table per year by not running it systematically. Sources: Joshua Hardwick / Ahrefs case study (18.67%, 31 reclaims from 166 emails); IDX case study (26%, 150 from 576); Editorial.Link 2025 framing.

Your client's quarterly review is in 10 days. The retainer report is due. The last "backlink audit" anyone shipped on this account was 14 months ago, included a 12-page PDF with DA/DR trend charts and a 15,000-domain disavow file regenerated from Semrush's "toxic links" export. The client opened it once, asked one question, never referenced it again. Then a competitor outranked them last week, the CEO asked "what's going on with our backlinks," and your account lead emailed you at 4pm Friday asking for "the audit." You open Ahrefs. You see 8,000 referring domains. You see the "Lost" filter showing 412 lost backlinks since the last audit. You don't know how many of those matter, how many can be reclaimed, or how many are just normal link rot.

This checklist is for that Monday. It assumes you already know what a backlink is. It focuses on the 2026 shifts that have moved the work meaningfully off the 2018-2022 playbook: (1) the disavow file is now a conditional intervention, not a quarterly deliverable — Mueller, Illyes, Hardwick, and Gabe have spent the last 3-5 years explicitly telling the industry that "toxic backlinks" as a category is dead at Google and that disavowing tool-flagged links typically hurts more than helps; (2) lost-link recovery is the new headline workstream — converts at 2-3× cold outreach because the editor's already endorsed you once; (3) donor-health degradation is silent and continuous — SpamBrain (real-time since Penguin 4.0 in 2016) devalues links from sites that drift in quality without notifying you; the link still shows in your inventory while its equity-pass quietly drops; (4) the audit's deliverable shape has shifted from "12-page PDF" to "one-page exec summary + action-list spreadsheet with $-value attached" — and that shift is documented as a larger driver of agency client retention than ranking outcomes themselves.

Three numbers anchor what's at stake.

1. Link rot baseline: ~17% of backlinks die within 12 months; ~28% within 3 years; ~66.5% within 9 years. Two independent datasets converge — Linkody's Kaplan-Meier survival analysis (sample: "millions of links") and Ahrefs' study of 2,062,173 websites since January 2013. Loss reasons are predictable: ~50% link removed from page, ~32% donor site uncrawlable, ~17% HTTP errors. Operator implication: assume a 4-6% quarterly loss baseline against your owned link inventory; under 4% = stale data, over 10% = recent algorithm update, donor-cluster migration, or negative SEO. The quarterly audit's job is to detect which losses matter enough to recover.

2. Lost-link reclamation converts at 18-26% via Wayback + Hunter outreach. Cold link-building outreach baseline is 8.5% in 2025, falling toward 3.43% in 2026. Sources: Ahrefs case study (Antonio Gabric, 31 reclaims from 166 emails = 18.67%); IDX case study (150 from 576 = 26%); BuzzStream + Belkins cold-email benchmarks. Editorial.Link's December 2025 framing: if an agency is building 250 links a year at $300 per link and losing 40-50 of them annually, that's $12-15k of already-paid-for equity recoverable at 2-3× the efficiency of equivalent fresh acquisition. This single workstream is what turns a quarterly audit from a deliverable expense into a P&L line item.

3. John Mueller, Search Central NYC, April 8 2025: "The disavow tool is not something that you need to do on a regular basis. It's not a part of normal site maintenance." Backed by Joshua Hardwick's August 2024 Ahrefs experiment (disavowed 129 tool-flagged "toxic" backlinks, 20 days, combined traffic change −7.1%, no detectable positive impact) and Glenn Gabe's GSQi case study (a site removed its 15,000+ domain disavow file in late March 2023 and gained 140% click lift plus 31,000+ keywords into top-10 positions). Gabe's standing estimate: 99.99% of sites do not need to use the disavow tool. The quarterly audit's disavow workstream is now a 30-minute trigger check, not a four-hour spreadsheet exercise.

This checklist has 40 items across 8 categories: pre-audit baseline & scope, lost-link recovery, anchor portfolio drift, donor-health degradation tracking, indexation + attribute drift, disavow review (sparingly), audit reporting + client communication, and anti-patterns being sold as quarterly audit best practice in 2026 that do not survive scrutiny. Tier filter, browser-saved progress, no account required.

Vocabulary

If you have read our anchor-text portfolio checklist, donor evaluation checklist, disavow file checklist, and on-page SEO checklist you have most of these. New terms in italics.

Link rot: the natural decay of backlinks over time — pages get removed, sites go offline, attributes flip. ~17% per year per Linkody and Ahrefs survival studies.
Lost-link recovery / link reclamation: the workstream of detecting backlinks that have disappeared and outreach to the donor to restore them. The 2026 audit's headline workstream.
Donor-health degradation: the silent decline of donor sites that pass less equity than they used to, even though the link itself still appears in your inventory. Post-HCU absorption (March 2024) makes this continuous.
Anchor portfolio drift: the gradual change in your site's inbound anchor-text distribution over multiple quarters. Even a "safe" distribution can drift toward over-optimization if no one is watching the delta.
Indexation drift on donor pages: a backlink whose donor page got deindexed passes no equity. The link is still in your inventory; the equity-pass is silently zero.
Attribute drift: a link's rel attribute flipping (dofollow → nofollow / sponsored / ugc) without notification. Often a CMS template change on the donor side; can affect 50+ of your links overnight.
SpamBrain devaluation: Google's AI-driven spam system identifying manipulation patterns and quietly ignoring the link's ranking value — no penalty, no manual action, no recovery path. Real-time since Penguin 4.0 folded into core in September 2016, enhanced March + December 2024 spam updates.
Discovery vs monitoring: two structurally different product categories. Discovery (Ahrefs / Semrush / Moz) crawls the web to find new backlinks. Monitoring (Linkwatcher / Linkody / Monitor Backlinks / LinkGuard) re-checks a known list on 12-24h cadence. Different ROIs, different tools.
$-value-attached action-list: the 2026 audit's defensible deliverable shape — every line item has an estimated dollar value (recovered link × historical cost-per-link, or lost equity from a degraded donor). Replaces the 12-page PDF.

Quick answers (the 4 questions you searched for)

How often should agencies run a backlink audit? Quarterly is the 2026 operator-default for the full audit, with continuous (12-24h cadence) automated monitoring between audits. Monthly full audits are increasingly called out as agency churn-bait — they produce 80% of the same findings 12 times a year while burning 6-8 agency hours per client per month. Reserve monthly for actively-building or high-competition niches (gambling, finance, legal); use semi-annual only as the absolute floor.

Is the disavow file still worth maintaining? Almost never as a quarterly deliverable. John Mueller, April 2025: "the disavow tool is not part of normal site maintenance." Glenn Gabe estimates 99.99% of sites do not need it. The exceptions: (a) a GSC manual action for unnatural links, (b) site has paid-link / PBN history you are walking away from, (c) clear post-negative-SEO link-spike pattern. Cross-link our disavow checklist for the workflow when one of those triggers fires.

What is the highest-ROI workstream in a quarterly audit? Lost-link recovery via Wayback + Hunter outreach. Reclamation converts at 18-26% in documented case studies (Ahrefs, IDX), vs ~8.5% for cold link-building outreach in 2025 (falling toward 3.43% by 2026). Editorial.Link's framing: a mid-market agency building 250 links/year and losing 40-50 of them at $300/link is leaving $12-15k of already-paid-for equity on the table annually. Most quarterly audits in 2018-2022 treated reclamation as a side item; the 2026 audit treats it as the headline.

Do I need a paid monitoring tool, or is the DIY workflow real? DIY is real and covers small-agency scale honestly. Spreadsheet + GSC export + manual site: checks + Wayback Machine + Hunter free tier costs $0 and takes 6-12 hours per client per quarter for ≤500 backlinks (20-40 hours for 5,000+). Paid monitoring (Linkwatcher, Linkody, Monitor Backlinks, LinkGuard) saves the hours; the question is whether your billable hour-cost across your client portfolio justifies the tool license. Above ~5 clients quarterly with non-trivial inventories, paid tooling usually wins on opportunity cost.

What success looks like

+90 days post-audit: at least 10-30% of the lost links you targeted for reclamation are back online (matching the documented reclaim baseline range); your anchor-portfolio drift report shows no exact-match cluster growing beyond safe thresholds since last quarter; your top-50 donors are still indexed and still publishing; your disavow decision is either "no action this quarter, triggers not present" (with the documented reason) or "scoped disavow for the X trigger" (with the diff). The client's review-meeting feedback is "I can see exactly what was done and what it's worth" — not "thanks for the report." The next quarterly audit takes 20% less time because the action-list spreadsheet from this quarter is the starting baseline for the next, not a fresh blank.

0 / 40 · 0%

Critical Important Nice to have

Show only undone

Every quarterly audit goes off the rails the same way: scope creep mid-execution, ambiguous 'what counts as a link' definitions, no baseline to compare to next quarter. Lock these before opening any tool.

Critical Audit scope frozen in writing — URL set, link-source set, time window A 'quarterly backlink audit' means different things to different operators. Without a written scope you spend the first 2 hours each quarter re-deciding what you're auditing, and the deliverable shape drifts every cycle. Lock: (a) the URL set (root domain only? all subdomains? specific high-value landing pages?), (b) the link-source set (Ahrefs + Semrush + GSC + monitoring tool = the union?), (c) the time window (everything ever recorded? rolling 12 months? since-last-audit only?).
How to do this
One-page scope doc per client, refreshed annually:
URL set: name the host(s) explicitly. If subdomains matter (e.g. blog.acme.com, app.acme.com), name them. If specific high-value landing pages get extra attention, list them.
Link-source set: minimum two third-party indexes + GSC. Vazoola operator framing: 'no single tool finds every backlink.' Ahrefs discovers ~35% more referring domains than Semrush on overlapping samples; the union is closer to truth than any single source.
Time window: 'since last audit' is standard for drift detection. 'Rolling 12 months' is standard for trend reporting. Pick one and stay consistent.
Scope changes mid-audit must be approved (and dated) by the DRI — see sc-04.
Important Last audit's action-list pulled up — which items closed, slipped, reclaimed If this is the third quarterly audit and you can't tell which items from the prior audit got actioned, the audit has no learning loop. The next quarter starts from a blank slate every time. Worse: client sees the same recommendations repeated and concludes the agency is generating reports, not driving outcomes.
How to do this
Open last quarter's action-list spreadsheet (per rep-02):
Count items closed (status = done) vs slipped (status = in progress / not started).
Reclamation outreach: how many of the targeted lost links were actually reclaimed? Compute conversion: reclaimed ÷ outreach-emails-sent.
Anchor / donor-health flags from last quarter: still flagged? closed? worsened?
If close-rate is < 50%, the action-list was either too ambitious or the items weren't assigned to a real owner. Diagnose at the +90-day review meeting (per rep-04) before next quarter.
Carry forward slipped items as 'open' into this quarter's action-list; never re-discover them as 'new findings.'
Critical Baseline snapshot frozen for next-quarter diff (counts are change-detection, NOT KPIs) Without a frozen snapshot you cannot detect drift next quarter — and drift detection is what justifies the cadence (per the brief's §1 framing). The numbers themselves are not KPIs (count rises mechanically with any activity); they are the baseline for calculating quarter-over-quarter change.
How to do this
Capture as a dated snapshot (Google Sheet, dated tab):
Total referring domains (across the union of sources).
Total backlinks (with explicit caveat: 'change-detection baseline, not KPI').
Dofollow / nofollow split.
Top 50 referring domains by historical traffic + link-equity contribution.
Top 200 anchor texts by frequency.
Top 50 link-target URLs by referring-domain count.
Current disavow file contents (if any — usually 'none').
The deliverable is the DIFF next quarter — not the snapshot itself. Stop reporting the absolute numbers as headline metrics to the client. Cross-link Heitzman / SEL Feb 2026 'Retire these 9 SEO metrics.'
Important Audit DRI (Directly-Responsible-Individual) named in writing — one person signs the deliverable Quarterly audits that fall apart usually trace back to fractured ownership: the SEO lead assumes the link-builder pulled the lost-links list, the link-builder assumes the analyst built the action-list, the analyst assumes the account lead is presenting to the client. One named person owns the audit end-to-end and signs the deliverable. Without this, the audit becomes 'whoever has time this week.'
How to do this
The DRI's responsibilities:
Owns the audit scope doc (sc-01) and any scope changes mid-audit.
Signs the deliverable (exec summary + action-list).
Presents at the +90-day client review meeting (rep-04).
Has authority to push back if the scope creeps mid-audit OR if the action-list grows beyond what the client team can realistically work in one quarter.
For agencies with a lead SEO + a link-builder + an analyst on the account: the DRI is usually the lead SEO. For freelance: the freelancer is the DRI by default. For in-house with an external vendor: the internal SEO is the DRI; the vendor delivers the inputs.
Important Client-side context confirmed — migration, CMS upgrade, content cull, algo update window, negative-SEO, side-channel link work An audit that ignores client-side context misattributes findings. A 40% lost-link rate is alarming in isolation but expected after a CMS upgrade that changed permalink structure. An anchor-drift spike is alarming in isolation but expected if the client's marketing team ran a brand-relaunch campaign with new exact-match anchors. Confirm context BEFORE diagnosing, not after.
How to do this
Five questions to email the client account lead before opening any tool:
Was there a site migration this quarter? (Cross-link our migration checklist.)
CMS upgrade? Template change? Permalink-structure change?
Major content cull or category restructure?
Algorithm update window in the audit period? (Cross-reference Google Search Status Dashboard.)
Any negative-SEO complaint? Client-side link-building activity outside the agency's scope?
Capture answers in the audit doc. If any answered yes, expand the corresponding workstream (lost-link recovery after migration, anchor drift after brand-relaunch, etc.).
Nice Time-box set explicitly — 6-12h SMB, 12-30h mid-market; if you blow past, the scope was wrong Industry-aggregated audit pricing maps roughly to time: $399-$1,399 SMB scope, $2,000-$5,000 deeper. That implies 6-12 hours SMB / 12-30 hours mid-market at typical agency rates. If a quarterly audit takes 50+ hours, either the scope is wrong, the tooling is wrong, or the audit is masquerading as a one-off deep audit. Time-box explicitly so you catch the drift before the budget is blown.
How to do this
Set the time-box in the scope doc (sc-01):
Small site (≤500 backlinks): 6-12 hours per quarterly audit.
Mid-market (500-5,000 backlinks): 12-30 hours.
Enterprise (5,000+ backlinks): 30+ hours; may justify monthly mini-audits instead.
Track actual hours per audit in a shared sheet. If you blow past the box by >25%, the next quarter's scope-freeze (sc-01) needs a re-look. Most over-runs trace to (a) lost-link list grew beyond the time-box's reclamation outreach capacity, (b) disavow review was triggered (one-off cost; document it), (c) the audit creeped into a one-off deep audit that should have been priced separately.

Reclamation converts at 18-26% vs 8.5% cold-outreach baseline. A mid-market agency leaves $12-15k of already-paid-for link equity on the table annually by not running this systematically. This is the workstream that turns the quarterly audit from a deliverable expense into a P&L line item.

Critical Pull 'Lost backlinks' report from Ahrefs / Semrush for the audit window; cross-reference with monitoring tool The lost-link inventory is the input to the entire workstream. Ahrefs' Lost filter + Semrush's equivalent are the two third-party indexes most commonly used. Cross-reference with your monitoring tool's lost-link alerts (or your DIY spreadsheet) — alerts catch losses earlier (12-24h) but third-party indexes catch losses your monitoring missed (e.g. from links you didn't know you had).
How to do this
Workflow:
Ahrefs → Site Explorer → Backlinks → Lost. Set date filter to audit window (e.g. last 90 days).
Semrush → Backlink Audit → Lost. Same date filter.
Export both to CSV. Dedupe by donor-domain + target-URL.
If you have a monitoring tool (LinkGuard, Linkwatcher, Linkody, or DIY), pull its lost-link alerts for the same window. Add to the union.
Expect 4-6% quarterly loss against your owned inventory (per the link-rot baseline). Significantly under 4% = stale data. Over 10% = recent algorithm update, donor-cluster migration, or negative SEO; investigate before proceeding to reclamation.
Critical Apply 'best links only' filter — narrow from all lost to lost actually worth recovering Joshua Hardwick / Ahrefs published a real-world filter cut: 6,285 lost links narrowed to 594 worth recovering. The other 5,691 are either low-quality donors not worth reclaim outreach, or links from pages that were dead before they linked to you. Reclamation outreach time is finite; spend it on links that mattered.
How to do this
Apply Editorial.Link's December 2025 vetting heuristic (also matches Hardwick's filter):
Donor DR ≥ 40 (or DA ≥ 30 if you use Moz).
Donor monthly organic traffic ≥ 1,000.
Donor language matches your target audience (usually English).
Donor is niche-relevant to your site.
The lost link's target URL still exists on your site (if it was a 404 cause of the loss, that's a different fix — see ll-04).
For higher-stakes clients: also filter for donors that themselves ranked in the top-10 for at least one commercial query in the audit window. These are the donors whose equity-pass is still hot.
Important Categorize each remaining lost link by REASON — different reasons need different fixes Lost-link causes are not all equal. Treating them all the same wastes outreach time on un-recoverable losses (donor page itself deindexed) and misses self-fix opportunities (your own 404 caused the loss — restore the page, equity is back without any outreach).
How to do this
Five categories with different remediation paths:
(a) Link removed from page — donor still has the page, just edited out your link. Highest reclaim potential. → Outreach (see ll-05).
(b) Destination page now 404 (your fault) — your page got deleted or moved without a 301. Restore the original URL OR add a 301 to the closest live equivalent. Recovers equity instantly without outreach. → ll-04.
(c) Donor page itself deindexed / 404 — donor's content is gone. Lost forever. Log and move on.
(d) Attribute flipped dofollow → nofollow / sponsored / ugc — donor template change usually (sitewide) or editorial change (single page). Sitewide = log and move on. Single page = outreach.
(e) 301 redirect chain inserted — link still mostly works; low priority. Document for the chain-cleanup workstream (cross-link migration #6 ll-03).
Critical For category (b) — restore your 404'd target URLs or 301 to closest equivalent Category (b) lost links are the lowest-effort recovery available: no outreach, no donor cooperation needed. Your page got deleted (or moved without a 301), and the donor's link is now pointing at a 404. Restore the page OR add a 301 to the closest live equivalent — equity is back the next Googlebot crawl.
How to do this
For each category (b) lost link:
Check if the URL existed before. Wayback Machine snapshot will show.
If the page was deleted intentionally (e.g. discontinued product), 301 the URL to the closest live equivalent (parent category, similar product).
If the page was deleted accidentally (e.g. CMS migration dropped a permalink), restore the URL from your archive.
NEVER 301 to homepage as a catch-all — Google treats that as a soft 404 (cross-link our migration checklist red-01).
Verify after fix: curl -ILso /dev/null -w "%{http_code}\n" https://your-domain.com/restored-url should return 200 (direct) or 301 → 200 (one hop).
Critical For category (a) — Wayback-confirm original context, Hunter the editor, broken-link-style outreach Suganthan Mohanadasan's documented pattern: outreach that references WHAT the editor originally endorsed converts 2-3× higher than context-less outreach. The Wayback Machine snapshot of the donor page (from when your link was live) shows the surrounding context — quote it back in the outreach so the editor remembers the value.
How to do this
Per-link workflow:
Wayback Machine → enter the donor URL → find a snapshot from when your link was live → screenshot the context (the paragraph around your link).
Hunter.io → enter the donor domain → find the editor / page-owner contact. If editor has moved (LinkedIn shows new role), find the CURRENT editor of the same section.
Draft a 4-sentence outreach naming: (a) the linking page, (b) the anchor text, (c) the new URL, (d) the reason for the reader — usually a small UX improvement, e.g. 'the link now 404s, which adds friction to your readers.' Never just 'please update.'
Sign with your full name + role. Never first-name-only on cold outreach.
Realistic conversion: 18-26% per Hardwick + IDX case studies. If your team is below 10%, debug personalization or value-prop framing before scaling.
Important Track reclaim conversion rate per outreach cycle — target 10-30% Without measurement you cannot iterate. The documented baseline range is 10-30% (Ahrefs 18.67%, IDX 26%). If your team is below 10%, the issue is usually one of: (a) low personalization (template-style outreach), (b) low context (no Wayback reference), (c) generic value-prop ('please update' vs 'broken link affects your readers'), (d) wrong contact (generic editorial@ instead of the current editor).
How to do this
Track per outreach cycle (one row per cycle in a shared sheet):
Date sent.
Number of outreach emails.
Replies (any reply, including 'no thanks').
Reclaims confirmed (link restored, verified live).
Conversion = reclaims ÷ emails. Reply-rate = replies ÷ emails (separate, useful as leading indicator).
Report quarterly in the audit deliverable. The conversion number is the proof the workstream is real — and the trend over multiple quarters is the proof the team is improving.
Important Report reclaim outcomes in $-value-attached terms — reclaimed links × historical cost-per-link The line a client's CFO can defend is 'we reclaimed $4,200 of equity that we had already paid to acquire' — not 'we reclaimed 14 backlinks.' Use the client's historical cost-per-link (CPL) from the link-building budget, or the industry baseline ($508.95 average per acceptable backlink per Editorial.Link's 518-SEO survey). This is the framing that turns the audit into a P&L line item.
How to do this
Compute per quarter:
Historical CPL: pull from client's link-building invoices over the last 12 months. (avg cost paid per acquired link). If unavailable, use $300-$500 as industry-baseline placeholder.
Reclaimed equity $-value = reclaimed links × CPL.
Year-projected upside = quarterly reclaimed equity × 4.
Report in the one-page exec summary (rep-01). Example line: 'Q3 reclamation: 14 links restored at avg $375 CPL = $5,250 reclaimed equity. Projected annual ≈ $21k.'
This number is what justifies the quarterly audit retainer line item to the client's finance team.

Penguin folded into core September 23, 2016 — SpamBrain devalues over-optimized anchor clusters in real time, silently. An anchor profile that drifted into over-optimization gradually is the most-missed quarterly audit finding. This category DETECTS drift; cross-link sibling #2 for the target distributions themselves.

Critical Pull current anchor distribution for site-level + top-50 linked-to URLs; diff against last quarter Drift only matters relative to a baseline. The quarter-over-quarter delta is what reveals over-optimization patterns SpamBrain looks for. A single snapshot is useless; the diff is everything. The top-50 linked-to URLs are where over-optimization concentrates because they're the commercial targets the link-building team focuses on.
How to do this
Workflow:
Export current anchor distribution from Ahrefs or Semrush: site-level + per-URL for the top-50 linked-to pages.
Open last quarter's snapshot (sc-03 baseline).
Compute the delta per anchor text per URL: percentage points change quarter-over-quarter.
Flag any anchor whose share grew > 5 percentage points QoQ on a single URL.
Flag any single anchor that now exceeds 40% of links to a single URL.
Drift detection tool: our anchor distribution analyzer handles the per-URL distribution snapshot. For the diff, manage the baseline + current in a sheet (one row per anchor per URL).
Audit anchor distribution free
Important Flag any exact-match commercial anchor cluster growing — even within 'acceptable' thresholds The SpamBrain trap is not the absolute number — it's gradual growth toward over-optimization. A page that goes from 3% exact-match anchor share in Q1 to 8% in Q4 is still inside 'acceptable' thresholds at any single snapshot but flagged as a manipulation pattern by SpamBrain when the trajectory is what's evaluated. Cross-link sibling #2 (anchor-text portfolio) for the actual safe-distribution targets.
How to do this
Quarterly trend check:
For each top-50 linked-to URL, plot the exact-match commercial anchor share over the last 4 quarters.
Flag URLs where the share has grown for 2+ consecutive quarters, even if still under the 10% absolute threshold.
For flagged URLs: diagnose. Is the agency's own outreach inflating the cluster? Client running their own campaign? Natural growth from genuine brand mentions?
If the agency's own outreach is the cause, the fix is upstream — at the link-building campaign brief, not in the audit. The audit's job is to surface the pattern; the campaign team's job is to course-correct.
Important Cross-check anchor drift against link-building campaigns this quarter — process bug at the brief level, not link-removal When an anchor cluster drifts due to the agency's own link-building, the fix is NOT removing the recently-built links (defeats the purpose). The fix is fixing the link-building campaign brief to require anchor variety. Catching this in the quarterly audit means the next campaign's brief gets the constraint — not next year's audit catches the same pattern.
How to do this
If an-01 + an-02 flagged a cluster growing on a specific URL:
Pull the agency's link-building campaign log for the audit window. What anchors were specified in the briefs for outreach targeting this URL?
If the briefs allowed exact-match anchors → fix the brief template for next quarter. New rule: max 10% of any campaign's outreach uses exact-match anchors; the other 90% are branded / descriptive / partial-match.
Document the brief-template fix as an action-list item with the campaign lead as owner.
This is the kind of process improvement the quarterly audit should produce — not a 'remove these 14 links' panic move.
Nice Cross-link to anchor-text portfolio checklist for the actual safe distribution targets This category's job is drift DETECTION, not setting the targets. The targets (30-50% branded / 20-30% partial-match / ≤10% exact-match) live in our anchor-text portfolio checklist. Don't re-litigate them here — reference the sibling article and move on.
How to do this
For the actual safe-distribution targets and the reasoning behind them, see our anchor-text portfolio checklist. That article covers:
The 30-50% branded / 20-30% partial-match / ≤10% exact-match target distribution.
Why exact-match anchor stuffing triggers SpamBrain devaluation.
How to build outreach anchor diversity into campaign briefs.
Edge cases (e.g. fresh sites with < 15 referring domains where the distribution math breaks down).
This quarterly-audit category just detects DRIFT against those targets. Setting the targets is sibling-article territory.

SpamBrain devalues silently when donor sites degrade in quality. HCU and core updates re-classify whole sites in ways that change the equity-pass on links you already own. This is the forensic, not janitorial, workstream that justifies the quarterly cadence.

Critical Top-50 referring domains by historical traffic + equity contribution — quick health check on each Not DR / DA alone (cross-link donor-evaluation sibling #3 for why DR-only is broken). Rank by historical referral traffic + link-equity contribution to your site. The top-50 by this measure are where any health degradation hurts most — a top-5 donor losing 50% of its organic visibility is the highest-impact silent loss in your portfolio.
How to do this
For each of the top 50 donors:
(a) Indexation: site:donor-domain.com — is the donor's homepage still indexed? (Most are; flag any that aren't.)
(b) Organic trend: Ahrefs or Semrush organic-traffic graph — is the donor's traffic trend positive over 12 months? Or in clear post-update decline?
(c) Content focus drift: manual eyeball — does the donor still publish content in the niche they were known for, or have they pivoted?
(d) Activity: latest published content < 90 days old? Or is the site abandoned?
Tools: our backlink-health-snapshot handles the per-URL spot check for ad-hoc donor investigation.
Run the health snapshot free
Critical Flag any donor with a clear HCU / core-update visibility drop in the last 12 months Glenn Gabe's standing observation: links from HCU-impacted donors silently lose equity-pass per the donor's site-level quality re-classification. The link still appears in your inventory; the equity it passes drops, sometimes to zero. ~22% of HCU-impacted sites have shown 20%+ lift after subsequent core updates; the other 78% remain meaningfully below baseline. Flag the donors in your portfolio that are still down so the client knows the equity erosion is real and silent.
How to do this
For each top-50 donor (from dh-01):
Pull the donor's organic-traffic graph in Ahrefs/Semrush.
Overlay the major algorithm-update windows from the Google Search Status Dashboard (HCU Aug 2022, Sep 2023, Mar 2024 absorption, plus subsequent core updates).
Flag any donor that lost >30% of organic traffic in the 60-day window after a confirmed update AND hasn't recovered.
For flagged donors: log the lost-equity in the quarterly report. The link still exists; the equity-pass is silently degraded.
Action: do NOT remove the link or 'disavow the donor.' The action is awareness + reallocation of new link-building budget toward fresher healthier donors.
Important Flag deindexed or sandboxed donor pages — the link is in your inventory but passes nothing Industry monitoring data: over half of paid/built backlinks reportedly never get indexed OR get deindexed over time. A backlink from a deindexed donor page is in your monitoring tool's 'live' count but passing zero equity. The quarterly audit catches this; weekly monitoring may miss the deindex event because the link technically still exists.
How to do this
Batch-check the donor PAGE (not the donor domain) for indexation:
Pull the list of top-200 backlinks (ranked by equity contribution).
Use our backlink-index-checker or a manual site:donor.com/specific-url check per URL.
For deindexed pages: classify as (a) likely permanent (page deleted, donor moved) → lost-link recovery via ll-03; (b) likely temporary (transient 5xx, crawl error) → re-check next monitoring cycle before acting.
Track aggregate indexation rate of donor pages quarter-over-quarter — a falling rate is a leading indicator of equity attrition across the portfolio.
Check if it is indexed in Google
Important Flag donors with paywall, robots block, or noindex on the page linking to you — equity-pass = zero A donor page behind a paywall (subscriber-only article) or with a noindex directive or robots.txt block on Googlebot is structurally invisible to Google's link-graph processing. The link is real; the equity-pass is zero. These are common silent equity losses on news / media donor sites that switched to paywalls mid-year.
How to do this
For each top-100 donor backlink:
Fetch the donor page with a normal user-agent and check for paywall content gates.
Check the donor page's robots meta tag and X-Robots-Tag header: curl -sI https://donor.com/page | grep -i x-robots-tag.
Check the donor's robots.txt for any path-level blocks affecting the linking page.
For confirmed paywalled / blocked / noindexed donor pages: log as silent equity loss in the quarterly report. Action depends on relationship — if you have a good relationship with the donor, outreach asking to make the page indexable; otherwise log and reallocate budget.
Nice Re-rank the top-50 donor list itself — the equity contribution ranking drifts quarter-over-quarter The donors that mattered most 12 months ago may no longer be the top 50. Equity contribution drifts as donor sites grow / decline, as your own page-level traffic shifts, as new high-quality donors join your portfolio. Re-rank quarterly so the dh-01 health check spends time on the donors that actually matter NOW, not the donors that mattered when the report template was first built.
How to do this
Quarterly:
Pull historical referral traffic per referring domain over the last 90 days (Ahrefs / Semrush organic referral, or GA4 referral channel).
Rank by referral traffic + a weighting for backlink count from each donor (signal that the donor passes multiple equity nodes, not just one).
Compare to last quarter's top-50 list. Note donors that entered + left.
For donors that left the top-50: don't drop entirely; deprioritize health-check frequency to semi-annual.
For donors that entered: add to next quarter's full health-check rotation.

A backlink whose donor page got deindexed passes no equity. A backlink whose rel attribute flipped dofollow → nofollow passes far less. Both are silent in any system that just counts 'is the link still there.' This category catches the silent equity changes.

Critical Batch-check rel attribute on owned/built links — has dofollow flipped to nofollow / sponsored / ugc? Real-world case (BacklinkDog): a SaaS company lost 50 dofollow backlinks overnight when a publisher upgraded its CMS and switched the template default to nofollow. Undetected for 3 months, rankings dropped 20+ positions. Quarterly batch-check on the rel attribute catches this within one cycle.
How to do this
Per-quarter workflow:
Pull the list of owned/built links (your imported inventory, or the top-200 most valuable per dh-01 ranking).
Use our nofollow bulk checker for the batch check, OR a custom script that fetches each donor page and inspects the rel attribute on the link to your domain.
Diff against last quarter's attribute snapshot.
Flag any link whose rel attribute changed.
For single-link flips (one link on one page) → outreach to the donor. For sitewide flips (multiple links on one donor with same rel change) → almost certainly a donor CMS template change; log and move on (outreach won't help).
Bulk nofollow check free
Critical Batch-check donor-page indexation — link from deindexed page passes zero equity Already in dh-03 framing — flagged again here because the indexation drift is the indexation-drift workstream's responsibility, not just donor-health. Donor health asks 'is the donor SITE healthy.' This asks 'is the SPECIFIC PAGE the link sits on still indexed.' Different question, different cadence (this is per-link, donor-health is per-donor).
How to do this
Same tool as dh-03 (backlink-index-checker), different scope:
Inventory: ALL owned/built backlinks (not just top 200).
Cadence: quarterly batch-check.
Classification of deindexed pages: (a) permanent (lost-link via ll-03), (b) transient (re-check next cycle).
Aggregate KPI: percentage of owned/built backlinks whose donor page is currently indexed. Track quarter-over-quarter. A falling rate is the leading indicator of equity attrition; the per-link drill-down identifies WHICH links to action.
Check if it is indexed in Google
Important Handle CMS-template-flip vs single-page-flip differently When 50 links on the same donor all flip dofollow → nofollow on the same date, that's a CMS template change — outreach to the editor about one of those links won't help (the template change is sitewide). When a single link on a single donor flips, that's editorial — outreach can recover it.
How to do this
From the dr-01 flagged list, group by donor:
Sitewide pattern (3+ links on same donor, same flip, same date): CMS template change. Document for the donor's CMS pattern log; do not outreach. If the donor is in your top 50 (dh-01), flag for future re-evaluation.
Single-link flip on a donor where other links are unchanged: editorial change on that specific page. Outreach the editor: 'I noticed our link on [page] changed to nofollow — was that intentional? Happy to discuss if there's a concern.'
Track outreach success rate on single-link flip recovery — usually higher than lost-link recovery (15-30%) because the donor relationship is intact.
Nice Aggregate indexation rate quarter-over-quarter — falling rate is leading indicator of equity attrition Per-link diagnosis tells you WHICH links to action. The aggregate rate tells you whether your portfolio is HEALTHY overall. A falling indexation rate across quarters is the leading indicator that the link-building team is acquiring donors that don't pass equity reliably — fix at the donor-evaluation step (cross-link donor-evaluation sibling #3).
How to do this
Quarterly KPI to add to the audit deliverable (rep-01):
Compute: (donor pages currently indexed) ÷ (total owned/built backlinks) × 100%.
Compare to last quarter's number.
If falling consistently for 2+ quarters, the donor-evaluation upstream needs review — donors being acquired don't index reliably or get deindexed quickly.
Cross-link our donor evaluation checklist for the criteria upstream link-building should be filtering against.

Mueller, Illyes, Hardwick, Gabe consistently since 2017: the disavow tool is a conditional intervention, not a quarterly deliverable. Three triggers; otherwise skip and document 'no action.' This category exists specifically to LOG the no-action decision as a defensible deliverable.

Critical Trigger check — only proceed to disavow review if one of three triggers is present Per Mueller (April 2025), Illyes (multiple), Hardwick (Aug 2024 experiment), Gabe (15K-domain removal case study). The disavow file is a conditional intervention. Three triggers warrant review; absent all three, the audit's disavow workstream is a 30-second logged 'no action this quarter.'
How to do this
Check, in order:
Trigger 1 — GSC manual action for unnatural links present? GSC → Security & Manual Actions → Manual Actions. If 'Unnatural links to your site' = ACTIVE, proceed to disavow review.
Trigger 2 — site has paid-link / PBN history we are walking away from? Confirm with the client (or your own records if you were the agency that placed them).
Trigger 3 — clear post-negative-SEO link-spike pattern in the last 30-90 days? Sudden spike of low-quality referring domains with abusive anchors. Look for: foreign-language donors with English-anchor commercial keywords, sudden cluster of low-DR donors, anchor cloud full of pharma / casino / adult terms.
If ALL THREE answer no: log 'no disavow action this quarter, no triggers present' + date + reviewer name + the disavow-file state (probably empty / unchanged) in the audit deliverable. This IS the deliverable — a documented defensible decision.
Important Vendor toxicity scores used as TRIAGE only — NEVER piped directly into a disavow file Gabe's 15,000-domain disavow case study removed → 140% click lift, 31k+ keywords into top-10. The disavow file was actively suppressing the site. The mechanism is that vendor toxicity algorithms generate false positives at a rate where the ranking-damage from disavowing-good-links materially exceeds the upside from disavowing-actual-bad-links (Google already algorithmically ignores actual bad links). Vendor scores are TRIAGE: 'which 100 of these 50,000 domains warrant a manual eyeball look first.'
How to do this
If a trigger fired (dv-01) and you need to manually review a portion of the link profile:
Use Semrush Toxic Score or Moz Spam Score to narrow the universe — e.g. flag the top 200 highest-scored domains for human review.
Manually review each of the 200. Decide based on YOUR judgment with context: is this manipulative? Does it match one of the dv-01 triggers? Was it in our control to place?
Disavow only the ones a human reviewer (not the score) deems actually bad.
Document the methodology in the audit deliverable: 'X domains reviewed manually, Y disavowed based on [reasoning].' Never pipe a tool's full toxicity export into a disavow file unchanged.
Important Log the no-action decision explicitly when no triggers are present An audit that omits the disavow workstream entirely looks like an oversight to a client who has been trained by 5 years of agency reports to expect a 'toxic link report' as a deliverable. The logged 'no action this quarter, no triggers present' line educates the client AND documents a defensible decision against any future second-guessing.

How to do this

One line in the audit deliverable (rep-01):
Disavow review: no action taken this quarter. None of the three triggers were present (no manual action; no paid-link history walk-away; no post-negative-SEO link spike detected). Reviewer: [name]. Date: [date]. Disavow file state: [unchanged / none].
For the client's first quarter learning this is a defensible workflow: include a short note explaining WHY this is the right call per Mueller / Illyes / Gabe — and cross-link our disavow file checklist for the full reasoning.
This documented decision is the deliverable. Activity isn't always action.
Build the disavow file free

Poor reporting drives more agency client churn than poor SEO performance — multiple 2026 agency-retention analyses converge on this. The audit's deliverable shape determines whether the work is recognized as value or filed under 'activity.'

Critical One-page exec summary — reclaimed $-value + drift flags + decisions The client's exec / CFO does not read 12-page PDFs. They read the first page. The exec summary must answer in 60 seconds: 'what did this audit produce that justified the cost?' Reclaimed-equity $-value is the answer; everything else is supporting context.
How to do this
One page, this structure:
Reclaimed equity this quarter: $-value + count of links restored (per ll-07).
Anchor drift flags: count of URLs flagged for over-optimization risk (per an-01 / an-02).
Donor-health flags: count of top-50 donors with degradation patterns (per dh-02).
Indexation / attribute drift: count of links with confirmed equity loss (per dr-01 / dr-02).
Disavow decision: 'no action' OR 'scoped disavow for X trigger' (per dv-03).
Next-quarter action-list: count of open items + estimated $-value-at-stake.
The full detail goes in the action-list spreadsheet (rep-02). The exec summary is what the client opens; the spreadsheet is what the client's team works.
Critical Action-list spreadsheet — one row per item with type, owner, status, $-value-at-stake, due-date A 12-page PDF cannot be worked. A spreadsheet can. The deliverable shape matters as much as the content — the action-list is what survives the quarter and becomes next quarter's input. Columns: item, type, owner, status, $-value, due-date.
How to do this
Required columns:
Item: short description (e.g. 'Reclaim outreach to editor@example.com for /article-X').
Type: reclamation-outreach / page-restore / donor-monitor / anchor-fix / etc.
Owner: agency-team-member OR client-team-member name (NOT 'team' — a person).
Status: not-started / in-progress / done / blocked.
$-value-at-stake: estimated equity at stake (for reclaim: target-link × CPL; for donor-health: donor's referral traffic × CTR × CPC equivalent).
Due-date: realistic deadline within the quarter.
Shared (Google Sheet / Notion / equivalent) so client team can update status in real-time. Reviewed at the +90-day client meeting (rep-04).

Important Diff against last quarter's baseline — what changed quarter-over-quarter The audit's value compounds across quarters. The single-quarter snapshot is interesting; the trend is actionable. Track the same metrics from sc-03's baseline snapshot quarter-over-quarter so the client sees direction, not just position.

How to do this

Add to the exec summary as a delta table:

Metric	Last Q	This Q	Δ
Total referring domains	...	...	...
Dofollow / nofollow ratio	...	...	...
Top-anchor cluster max %	...	...	...
Donor-health flags	...	...	...
Owned-link indexation rate	...	...	...
Reclaimed equity $	...	...	...

Direction matters more than absolute values. A growing reclaimed-equity number, a falling drift-flag count, a stable indexation rate = the audit is working.

Important Quarterly executive review meeting — 30 min walking through the action-list, not the PDF The audit deliverable becomes real when the client team commits to actions. A 30-minute review meeting walking through the action-list spreadsheet (NOT presenting the PDF) is where commitment happens. Without the meeting, the audit deliverable is filed and forgotten.
How to do this
Standing meeting structure (30 min):
5 min: exec summary highlights (reclaimed $-value, top 3 flags).
15 min: walk the action-list. For each open item: who owns it, what's the deadline, any blockers.
5 min: agree on cross-team handoffs (agency does X, client team does Y).
5 min: scope discussion for next quarter — any context changes the agency should know about?
Calendar invite locked at audit kick-off. The DRI (sc-04) presents. The client-side counterpart should be the SEO lead / marketing-ops lead, not just an account manager.
Nice Time-spent per audit tracked — if > 30h or < 6h, the scope was wrong The scope-freeze in sc-01 + time-box in sc-06 only work if you actually measure actual hours. Track per-audit, per-client. Multi-quarter trend shows whether the audit is delivering value at the time-cost OR whether the scope is creeping (or shrinking past the threshold of usefulness).
How to do this
Track in a shared sheet:
Date.
Client.
Scope tier (SMB / mid-market / enterprise).
Hours spent (per audit-DRI; capture from Toggl / Clockify / agency time-tracking).
Reclaimed equity $-value (from ll-07).
Quarterly review: ratio of reclaimed-equity ÷ hours-spent. If trending down across multiple quarters, debug. The ratio should improve over time as the team gets better at the workflow.

Six things vendor blogs, agency proposals, and dashboard SaaS products still sell as quarterly-audit best practice that do not survive scrutiny against 2025-2026 evidence. If you've planned around any of these, you're not behind — you're early enough to fix the deliverable shape.

Important Don't ship a 12-page monthly backlink report with DA/DR trend lines as the deliverable Three independent problems: (a) DA and DR aren't Google metrics — Heitzman / SEL Feb 2026: 'Stop putting these in executive dashboards'; (b) total backlink count rises mechanically with any link acquisition regardless of quality — a vanity number; (c) monthly delivery of high-effort reports most clients don't read is a documented churn driver. Ship a quarterly audit with an action-list ($ recovery candidates, anchor-drift items, donor-health flags) plus a 1-page monthly monitoring summary — not 12 pages of trend lines.
How to do this
Replace the 12-page monthly report with:
Quarterly audit deliverable (this checklist): 1-page exec summary (rep-01) + action-list spreadsheet (rep-02). Reviewed at a 30-minute meeting (rep-04).
Monthly monitoring summary: 1 page max. Alerts triggered, links lost, links recovered, any drift flags worth raising before the quarterly. NOT a full backlink report.
The format change typically meets initial client resistance ('but the report was so detailed') and produces measurably better retention because the client actually engages with the deliverable.
Critical Don't pipe Semrush or Moz 'toxic' lists directly into a recurring disavow file Mueller and Illyes have called this out specifically as harmful. Vendor toxicity algorithms generate false positives at a rate where ranking-damage from disavowing-good-links materially exceeds upside from disavowing-actual-bad-links (Google already ignores actual bad links algorithmically). Gabe's 15,000-domain disavow file removal → 140% click lift case study is the canonical anti-example.
How to do this
The right workflow (from dv-01 + dv-02):
Disavow workstream only proceeds if one of three triggers fires (manual action / paid-link walk-away / post-negative-SEO spike).
If triggered: use vendor toxicity scores as TRIAGE (narrow universe for manual review), never as disavow input.
Disavow only what a human reviewer with context deems actually manipulative AND matches a trigger pattern.
Stop selling 'monthly disavow file refresh' as a deliverable. It's at best wasted work; at worst it's the Gabe pattern of silent suppression.
Important Don't bill quarterly audit by report PDF page count The whole point of the audit is to produce an action-list with $-value attached — reclaimed-equity targets, anchor-drift flags, donor-health degradations. A 50-page PDF documenting every backlink in detail is the opposite of useful: it dilutes the action-list and trains the client to ignore future audits. 'Report-page-count = price' is one of the most-cited 2026 agency red flags.
How to do this
The deliverable is:
One-page exec summary (rep-01).
Action-list spreadsheet (rep-02) — typically 20-50 rows.
That's it. No 50-page PDF. No appendix of every backlink. The action-list IS the deliverable; the exec summary IS the executive view.
Pricing should map to scope (sc-01) and time-box (sc-06), not page count. Industry-aggregated 2026 pricing: $399-$1,399 SMB, $2,000-$5,000 deeper scope, 10-30 hours typical at $100-$250/hr.
Critical Don't run an automated 'disavow service' on a recurring schedule against third-party toxicity exports Industrialised form of anti-02. Several SaaS 'auto-disavow' services pitch monthly disavow-file regeneration as a feature. Mechanism is exactly what Illyes warns against, at machine scale; false-positive rate is multiplied by automation; failure mode (Gabe-pattern silent suppression) is hidden because the deliverable looks like activity. Marie Haynes — who used to run a public toxic-link blacklist — shut hers down because she no longer believes the workflow helps.
How to do this
If a tool / service you use pitches 'automated disavow file updates' as a recurring deliverable, that's a red flag, not a feature. The disavow workflow REQUIRES human judgment per dv-02; any 'auto-disavow' is by definition piping vendor toxicity scores into the disavow file, which is the exact pattern Mueller, Illyes, Hardwick, and Gabe have warned against.
If you've been running auto-disavow as part of a service tier:
Audit the current disavow file (Gabe-pattern: many sites discover the auto-disavow has accumulated thousands of false-positive entries).
Run a controlled test: REMOVE the disavow file for one site and monitor for 60-90 days. If the site lifts (Gabe-pattern), the auto-disavow was actively suppressing.
Discontinue the automated workflow. Replace with the conditional dv-01 trigger-based review.
Important Don't set monthly backlink-count goals as a KPI for the link-building team Velocity-only goals are the proximate cause of the over-optimized anchor profiles, low-quality donor portfolios, and unnatural growth patterns SpamBrain silently devalues in 2024-2026. Editorial.Link's 2026 518-SEO survey: 67.2% cite 'scaling without quality loss' as the primary obstacle in link building — a number that goes UP when link-count is the team's KPI.
How to do this
Replace 'links acquired' as a KPI with:
Link-quality distribution: % of acquired links from donors meeting your defined quality bar (e.g. DR≥40 + 1K+ traffic + niche-relevant + English).
Donor-diversity index: % of new referring domains this quarter vs % of repeat donors (some repeat is fine; pure-repeat signals manipulation).
Reclaimed equity $-value (from ll-07) — the audit-side KPI that measures recovered value, not new acquisition.
Average donor health: average organic-traffic trend across the top 50 referring domains.
These are slower-changing metrics than 'links/month' — but they don't reward the manipulation patterns SpamBrain catches.
Important Don't treat 'backlink discovery' and 'backlink monitoring' as one workstream in one tool They are structurally different product categories with different ROIs. Discovery (Ahrefs / Semrush / Moz) crawls the web to find new backlinks, refreshes on multi-day cadence, excels at competitor analysis and prospecting. Monitoring (Linkwatcher / Linkody / Monitor Backlinks / LinkGuard) re-checks a known list on 12-24h cadence, excels at fast drift detection. Trying to do continuous monitoring out of a discovery tool means alerts lag the actual event by 3-7 days.
How to do this
The 2026 agency stack is usually three layers:
Discovery: Ahrefs OR Semrush. The all-in-one index. Refresh cadence ~15min (Ahrefs) to 2-4 weeks (Semrush).
Ground truth: Google Search Console. Sampled, capped at 1,000 rows UI, but the only source from Google itself. Always cross-reference.
Monitoring: a dedicated monitoring tool (Linkwatcher, Linkody, Monitor Backlinks, LinkGuard) OR a DIY spreadsheet + cron for the imported list of owned/built links. 12-24h cadence on a known list.
One tool covering all three is rare and usually weak in monitoring. Pick the right tool per job, not the cheapest all-in-one.

About the Author

Andrei

SEO and digital marketing professional with 13+ years of experience. Started as a website administrator in 2011, transitioned to SEO, and achieved top-3 rankings for competitive keywords. Co-founded a consulting firm specializing in marketing audits for companies in Ukraine and internationally. Built LinkGuard to solve the problem he experienced firsthand: most SEO teams purchase links but never monitor their survival. Based in Kyiv, Ukraine.