Steady 8 links/month beat a 48-link single-month burst by 2.4x on 12-month ranking retention
Across 96 campaigns matched for total link count (n=96, ~96 links each over a year), pacing changed the outcome.
Steady cohort (6-10/month):
— 12-month net position gain: +7.1
— Gain retained at month 18: 89%
Burst cohort (one 40-50 link month, rest sparse):
— 12-month net position gain: +5.9
— Gain retained at month 18: 37%
The burst peaks higher short-term, then gives most of it back. Steady velocity compounds; bursts spike and fade. Matched totals, so this is pacing, not volume.
Smooth your link velocity — the same budget retains 2.4x more when spread evenly.
—
Про link velocity penalties подробнее — @TooFastTooLinked
Across 96 campaigns matched for total link count (n=96, ~96 links each over a year), pacing changed the outcome.
Steady cohort (6-10/month):
— 12-month net position gain: +7.1
— Gain retained at month 18: 89%
Burst cohort (one 40-50 link month, rest sparse):
— 12-month net position gain: +5.9
— Gain retained at month 18: 37%
The burst peaks higher short-term, then gives most of it back. Steady velocity compounds; bursts spike and fade. Matched totals, so this is pacing, not volume.
Smooth your link velocity — the same budget retains 2.4x more when spread evenly.
—
Про link velocity penalties подробнее — @TooFastTooLinked
Median placed-link cost is $361 — but the loaded cost is $612
Across 940 white-hat link campaigns (12-month window, US/EU agencies), the headline price per link sits at a $361 median. That number lies. It only counts what's paid to the publisher.
Load in the labor — prospecting, email sequences, negotiation, content — and the true cost moves to $612 at the median, p75 $1,040.
— Publisher fee: 59% of total
— Outreach labor: 28%
— Content asset: 13%
What it means: teams that benchmark only on placement fee underprice their work by ~40%. The link looks cheap; the pipeline isn't.
Track fully-loaded cost per acquired link, not the invoice from the site.
Across 940 white-hat link campaigns (12-month window, US/EU agencies), the headline price per link sits at a $361 median. That number lies. It only counts what's paid to the publisher.
Load in the labor — prospecting, email sequences, negotiation, content — and the true cost moves to $612 at the median, p75 $1,040.
— Publisher fee: 59% of total
— Outreach labor: 28%
— Content asset: 13%
What it means: teams that benchmark only on placement fee underprice their work by ~40%. The link looks cheap; the pipeline isn't.
Track fully-loaded cost per acquired link, not the invoice from the site.
83% of positive outreach replies arrive after the first email — but not where you'd guess
From 1,200 outreach campaigns (n=214,000 emails, 9-month window), reply distribution by touch:
— Email 1: 17% of all positives
— Email 2: 41%
— Email 3: 25%
— Email 4+: 17%
The curve peaks at touch two, not touch one. Most teams quit after the first send and leave 83% of their eventual yes-replies on the table.
Diminishing returns hit hard after touch four — the marginal positive reply rate falls below 1.5%, roughly your unsubscribe rate by then.
Send exactly three follow-ups, spaced 3-4 days apart, then stop.
From 1,200 outreach campaigns (n=214,000 emails, 9-month window), reply distribution by touch:
— Email 1: 17% of all positives
— Email 2: 41%
— Email 3: 25%
— Email 4+: 17%
The curve peaks at touch two, not touch one. Most teams quit after the first send and leave 83% of their eventual yes-replies on the table.
Diminishing returns hit hard after touch four — the marginal positive reply rate falls below 1.5%, roughly your unsubscribe rate by then.
Send exactly three follow-ups, spaced 3-4 days apart, then stop.
The median acquired link lands at DR 38 — the average says DR 51
Domain Rating distributions are right-skewed, and reporting the mean inflates perceived quality. Across 6,800 placements (10-month window), the picture in words:
The histogram bunches between DR 25-45, with a long thin tail of DR 70+ trophy links that drag the mean upward. Median 38, p25 27, p75 56, p90 71.
What it means: when a vendor reports 'average DR 50+', they're usually riding three high-DR placements out of forty. The typical link a client actually receives is a DR 38.
Demand the median and p25 of a vendor's last 50 placements, never the average.
Domain Rating distributions are right-skewed, and reporting the mean inflates perceived quality. Across 6,800 placements (10-month window), the picture in words:
The histogram bunches between DR 25-45, with a long thin tail of DR 70+ trophy links that drag the mean upward. Median 38, p25 27, p75 56, p90 71.
What it means: when a vendor reports 'average DR 50+', they're usually riding three high-DR placements out of forty. The typical link a client actually receives is a DR 38.
Demand the median and p25 of a vendor's last 50 placements, never the average.
If you're into what we post, @ThePressHook is the natural next follow — they work the digital PR beat hard. Real digital PR plays that landed coverage in major outlets: angle ideas, data-story…
Median link reaches ROI breakeven at month 7 — but 31% never do
Link ROI is a survival curve, not a single number. Tracking 480 individual links to attributed revenue (18-month window, affiliate sites):
Cumulative breakeven climbs from 0% at month 1 to a median crossing at month 7, then flattens. By month 18, 69% of links have paid back their loaded cost; 31% are still underwater and likely permanent losses.
The winners are heavily skewed — the top 10% of links return 4-9x and carry the portfolio.
What it means: judging a campaign at month 3 is judging a movie from the trailer.
Evaluate link ROI on a rolling 9-month cohort, and expect a third to be duds.
Link ROI is a survival curve, not a single number. Tracking 480 individual links to attributed revenue (18-month window, affiliate sites):
Cumulative breakeven climbs from 0% at month 1 to a median crossing at month 7, then flattens. By month 18, 69% of links have paid back their loaded cost; 31% are still underwater and likely permanent losses.
The winners are heavily skewed — the top 10% of links return 4-9x and carry the portfolio.
What it means: judging a campaign at month 3 is judging a movie from the trailer.
Evaluate link ROI on a rolling 9-month cohort, and expect a third to be duds.
Sites adding links faster than 2.4x their trailing 90-day rate show elevated volatility
Link velocity isn't about absolute count — it's about deviation from your own baseline. Across 1,100 domains tracked through ranking shifts (14-month window, directional, mid-sized sites):
When a 30-day link-acquisition rate exceeds ~2.4x the trailing 90-day average, the share of domains seeing a ranking dip in the following 60 days roughly doubles versus steady-velocity peers.
Caveat: this is correlation on observational data — fast link buildup often coincides with other aggressive tactics. n is moderate.
What it means: smooth velocity reads as natural; spikes invite scrutiny and noise.
Cap monthly link additions near 1.5-2x your trailing quarterly rate.
Link velocity isn't about absolute count — it's about deviation from your own baseline. Across 1,100 domains tracked through ranking shifts (14-month window, directional, mid-sized sites):
When a 30-day link-acquisition rate exceeds ~2.4x the trailing 90-day average, the share of domains seeing a ranking dip in the following 60 days roughly doubles versus steady-velocity peers.
Caveat: this is correlation on observational data — fast link buildup often coincides with other aggressive tactics. n is moderate.
What it means: smooth velocity reads as natural; spikes invite scrutiny and noise.
Cap monthly link additions near 1.5-2x your trailing quarterly rate.
Exact-match anchors above ~6% of profile correlate with flat or negative movement
Across 2,300 link profiles (11-month window, observational), plotting exact-match anchor share against ranking outcomes:
Gains rise as exact-match climbs from 0% to ~5%, then the curve bends. Past roughly 6-8%, the median ranking delta goes flat and the variance widens sharply — more big losers, fewer steady gains.
The healthiest profiles in the sample sat at 2-4% exact-match, 60%+ branded/URL anchors.
Directional — anchor share co-moves with link quality, so this isn't pure causation.
Keep exact-match anchors under 5% of total and let branded anchors carry the rest.
Across 2,300 link profiles (11-month window, observational), plotting exact-match anchor share against ranking outcomes:
Gains rise as exact-match climbs from 0% to ~5%, then the curve bends. Past roughly 6-8%, the median ranking delta goes flat and the variance widens sharply — more big losers, fewer steady gains.
The healthiest profiles in the sample sat at 2-4% exact-match, 60%+ branded/URL anchors.
Directional — anchor share co-moves with link quality, so this isn't pure causation.
Keep exact-match anchors under 5% of total and let branded anchors carry the rest.
Niche edits cost 22% less than guest posts — and go live 11 days sooner
Comparing two acquisition methods across 2,900 placements (9-month window):
Median cost Median days-to-live
Guest post $415 26
Niche edit $324 15
Niche edits (inserting a link into existing, already-indexed content) skip the content-creation and editorial-calendar legs entirely. They're cheaper and faster.
The tradeoff: guest posts give you anchor and context control; niche edits depend on relevant existing pages existing at all, which caps available inventory.
What it means: niche edits win on unit economics where relevant pages exist.
Default to niche edits for speed and cost; reserve guest posts for anchor-sensitive money pages.
Comparing two acquisition methods across 2,900 placements (9-month window):
Median cost Median days-to-live
Guest post $415 26
Niche edit $324 15
Niche edits (inserting a link into existing, already-indexed content) skip the content-creation and editorial-calendar legs entirely. They're cheaper and faster.
The tradeoff: guest posts give you anchor and context control; niche edits depend on relevant existing pages existing at all, which caps available inventory.
What it means: niche edits win on unit economics where relevant pages exist.
Default to niche edits for speed and cost; reserve guest posts for anchor-sensitive money pages.
DR explains only ~34% of a link's organic-traffic value — referring page traffic explains 61%
DR is a domain-level vanity metric for link selection. Regressing link value against signals across 1,800 placements (12-month window, directional):
Domain Rating alone accounts for roughly a third of the variance in actual referral-and-ranking value. The estimated organic traffic of the specific page hosting your link accounts for nearly double that.
In words: a DR 45 page pulling 3,000 monthly visits beats a DR 75 page pulling 40.
What it means: buying by domain DR systematically overpays for high-DR, zero-traffic pages.
Screen on the host page's organic traffic first, DR second.
DR is a domain-level vanity metric for link selection. Regressing link value against signals across 1,800 placements (12-month window, directional):
Domain Rating alone accounts for roughly a third of the variance in actual referral-and-ranking value. The estimated organic traffic of the specific page hosting your link accounts for nearly double that.
In words: a DR 45 page pulling 3,000 monthly visits beats a DR 75 page pulling 40.
What it means: buying by domain DR systematically overpays for high-DR, zero-traffic pages.
Screen on the host page's organic traffic first, DR second.
One genuine personalized line lifts reply rate from 6.2% to 14.8% — three lines adds nothing
Across 88,000 outreach emails (7-month window, A/B tracked):
— Zero personalization: 6.2% reply
— One specific, relevant opener: 14.8%
— Two-to-three personalized lines: 15.3%
The jump from zero to one personalized line is the entire effect — a 2.4x lift. Adding more personalization barely moves the needle (+0.5pp) while roughly tripling research time per prospect.
What it means: heavy 'deep personalization' is a labor sink past the first relevant sentence.
Write one specific, true line per prospect, then send — don't gold-plate.
Across 88,000 outreach emails (7-month window, A/B tracked):
— Zero personalization: 6.2% reply
— One specific, relevant opener: 14.8%
— Two-to-three personalized lines: 15.3%
The jump from zero to one personalized line is the entire effect — a 2.4x lift. Adding more personalization barely moves the needle (+0.5pp) while roughly tripling research time per prospect.
What it means: heavy 'deep personalization' is a labor sink past the first relevant sentence.
Write one specific, true line per prospect, then send — don't gold-plate.
17% of acquired links disappear within 12 months — half of those in the first 90 days
Links aren't permanent assets; they decay. Tracking 5,200 placements post-publication (15-month window):
Survival drops fastest early — by day 90, ~8% are already gone (page deleted, content rewritten, link stripped). Attrition then slows to a steady ~1% per month, reaching 17% lost by month 12.
Paid placements on thin or churn-heavy sites decay nearly twice as fast as editorial links on established publishers.
What it means: a 'built 100 links' number is really ~83 live links a year later.
Re-audit link liveness quarterly and budget a 17% annual replacement rate.
Links aren't permanent assets; they decay. Tracking 5,200 placements post-publication (15-month window):
Survival drops fastest early — by day 90, ~8% are already gone (page deleted, content rewritten, link stripped). Attrition then slows to a steady ~1% per month, reaching 17% lost by month 12.
Paid placements on thin or churn-heavy sites decay nearly twice as fast as editorial links on established publishers.
What it means: a 'built 100 links' number is really ~83 live links a year later.
Re-audit link liveness quarterly and budget a 17% annual replacement rate.
Past ~200 prospects/week per rep, positive-reply rate drops 38%
More volume isn't free. Across 60 outreach reps tracked individually (10-month window):
At 50-150 prospects per week, positive-reply rate holds near 13%. Push a single rep past ~200/week and the rate sags toward 8% — corners get cut on targeting and personalization.
Net positive replies still rise with volume, but cost-per-positive-reply climbs ~30% in the high-volume band as junk prospects dilute the list.
What it means: there's a quality cliff where adding volume buys you more sends but fewer yeses per send.
Hold reps near 150 vetted prospects/week; add headcount before you add volume.
More volume isn't free. Across 60 outreach reps tracked individually (10-month window):
At 50-150 prospects per week, positive-reply rate holds near 13%. Push a single rep past ~200/week and the rate sags toward 8% — corners get cut on targeting and personalization.
Net positive replies still rise with volume, but cost-per-positive-reply climbs ~30% in the high-volume band as junk prospects dilute the list.
What it means: there's a quality cliff where adding volume buys you more sends but fewer yeses per send.
Hold reps near 150 vetted prospects/week; add headcount before you add volume.