Off-page work is the hardest part of an agency's SEO delivery to systemise. Content has a brief, a writer, an editor, a publish date. Technical audits have a crawl and a ticket list. Link building has a target list, a mailbox, a lot of silence, and a client asking why last month produced two placements instead of five.
Good link building outreach campaign management is mostly about making an unpredictable output predictable enough to staff, price, and report. You cannot control whether a journalist replies. You can absolutely control how many qualified prospects enter the pipeline each week, how fast replies get answered, and how much of an editor's time a campaign consumes before you know it's failing.
Why off-page campaigns break down inside agencies
Across agencies I've worked with, the same four failure patterns show up:
- Outreach lives in one person's inbox. When they take leave, the campaign stops. When they quit, the relationships leave too.
- No separation between prospecting and sending. The same person researches, writes pitches, sends, negotiates, and chases the client for asset approval. Throughput collapses to whatever their worst week allows.
- Retainer hours are invisible. Link building is billed as "10 hours/month" but nobody tracks that prospecting alone ate seven of them.
- Reporting is output-only. The client sees "3 links" and has no idea whether 40 or 400 emails produced them, so every slow month feels like a failure.
Fixing these is a process problem, not a tooling problem. But the process needs to be recorded somewhere both the team and the client can see.
Structure the campaign as a five-stage pipeline
Treat every prospect as a work item moving through stages, exactly like an editorial workflow. If you've already built a content production pipeline, the mental model transfers directly.
1. Prospecting
Sourcing raw candidates: SERP competitors, link intersect gaps, journalist databases, podcast lists, resource pages, industry roundups. Volume stage. Nothing here is qualified.
2. Qualification
Every prospect gets checked against fixed thresholds before anyone writes a word. Typical agency thresholds: organic traffic above 1,000/month, topical relevance to the client's category, no obvious link-selling footprint, outbound link ratio sane, contactable human identified. Expect to disqualify 50–70% of raw prospects. That rejection rate is a feature — it's what stops your senders wasting weeks on sites that would never link.
3. Outreach
First send plus a defined follow-up sequence. Two follow-ups, spaced four and nine days out, captures most of the available replies. A third adds a little; a fourth mostly generates complaints.
4. Conversation
Anything with a human reply. This is where placements are won or lost and where most agencies leak value — a reply that sits for three days is a placement you've probably lost to whoever answered faster.
5. Placement and verification
Link is live, checked for follow status, correct anchor, correct target URL, and indexed. Verification is a separate step because a shocking share of promised links go live wrong: nofollowed without warning, pointing at the homepage instead of the target page, or buried on a page nobody crawls.
Every prospect sits in exactly one stage. That single rule turns a chaotic mailbox into something you can forecast against.
Set targets by stage, not by links
Committing to "five links a month" in a contract puts you at the mercy of variables you don't control. Commit to activity and quality instead, and use conversion ratios to set expectations.
Work backwards from a real baseline. If historical data says cold outreach converts at 1.5% send-to-placement and 60% of raw prospects fail qualification, then four placements per month needs roughly 270 sends, which needs about 675 raw prospects. That's the number that drives staffing.
Rough per-hour throughput for a competent specialist:
- Prospecting: 40–60 raw prospects/hour with tooling, fewer for niche verticals
- Qualification: 25–40 prospects/hour if thresholds are written down; half that if judgement calls are involved
- Personalised sending: 15–25/hour for light personalisation, 6–10/hour for genuine research-led pitches
- Conversation handling: unpredictable, but budget 20% of total campaign hours
Those numbers make the retainer conversation honest. A 10-hour monthly link building allocation supports roughly 150–200 sends with real personalisation — not 500. Track time per stage against those estimates and you'll spot within two weeks whether a campaign is under-resourced. Any PM tool with per-task time tracking against phases works; PeakKR was built around this pattern because agencies kept discovering their link building lines were the least profitable part of the retainer.
Protect deliverability like it's an asset
Outreach programmes die from sender reputation problems more often than from bad pitches. Non-negotiables:
- Separate sending domains per campaign type. Never send cold outreach from the domain carrying your client communications.
- Volume caps per mailbox. 30–50 cold sends per day per inbox, warmed up over two to three weeks. Agencies that jump to 200/day burn the domain in a month.
- SPF, DKIM, DMARC configured on every sending domain before the first send. Non-optional in 2024 onward.
- Bounce rate ceiling of 3%. Above that, pause and re-verify the list.
- A real unsubscribe path and immediate suppression list updates.
Log these checks as recurring tasks in your project tool rather than trusting memory. Domain health reviews on a fortnightly cadence catch problems while they're still fixable.
Reply speed is the metric nobody tracks
If you measure one internal thing beyond volume, measure time-to-first-reply on inbound conversations. Journalists and editors move fast. A four-hour response window versus a next-day window is often the difference between a placement and a "we've already published, sorry."
Set a hard SLA — four business hours during campaign weeks — and make it visible. Assign a named backup for holidays. Route the outreach mailbox so a second person can see and answer conversations, because single-inbox ownership is how you lose relationships when someone leaves.
Where AI helps and where it destroys campaigns
AI is genuinely useful in the top half of the pipeline: clustering prospects by topic, summarising a target site's recent coverage, drafting three angle variations for a pitch, extracting contact details from messy pages, flagging prospects that fail your thresholds. Building repeatable prompts for these tasks — see our guide on structuring AI prompts for marketing deliverables — can cut prospecting and qualification time meaningfully.
Where it fails: fully automated personalisation and reply handling. Editors recognise generated pitches instantly, and a template that references "your recent article on [topic]" without a human having read it does long-term damage to your sender reputation and your agency's name. Draft with AI, always send human-reviewed. The same caution applies to any client-facing communication.
Report leading indicators, not just placements
The reporting structure that keeps clients calm through a slow month:
- Weekly: prospects added, prospects qualified, emails sent, reply rate, active conversations. Pure pipeline health.
- Monthly: placements with linking page URL, that page's estimated traffic, anchor text, target URL, follow status, and a one-line note on why the placement matters.
- Quarterly: referring domain growth versus the two closest competitors, plus ranking movement on the specific target URLs you've been building to.
A client who sees 240 sends and a 7% reply rate in a two-placement month understands the machine is running. A client who only sees "2 links" assumes it isn't. Same reality, entirely different renewal conversation.
Keep this data in the same workspace as the rest of the client's project rather than in a standalone outreach tool, so link building sits alongside content and technical work in one view. If you're evaluating where that should live, our roundup of PM tools for agencies covers the trade-offs between general-purpose boards and agency-specific systems.
Campaign launch checklist
- Written qualification thresholds agreed with the SEO

Nick Quirk