Nick QuirkGuest posting fails in agencies for a boring reason: the work is invisible. A spreadsheet with 400 rows tells you nothing about where deals are stuck, who owes whom an email, or whether next month will produce four links or fourteen. By the time you notice a problem, the campaign is six weeks behind.
A guest posting pipeline fixes this by treating every opportunity like a deal in a sales CRM — a card that moves left to right through named stages, with an owner, a date, and a next action. You can see stalls. You can forecast. You can staff correctly. Here's how to build one that your team actually maintains.
Why a spreadsheet stops working at ~50 prospects
Spreadsheets are fine for a one-off campaign of 20 sites. They break the moment you're running guest posting for five clients at once, because:
- Status becomes a text field. "Emailed 2x", "waiting", "maybe?" — three people write three different things and no filter works.
- Nothing surfaces aging. A prospect that went quiet 19 days ago looks identical to one you pitched yesterday.
- Handoffs vanish. The outreach specialist gets a yes, drops a note in Slack, and the writer never sees it.
- Client reporting is manual. Someone rebuilds the same pivot table on the 1st of every month.
A visual board solves all four because stage is position, aging is a timestamp, handoffs are assignee changes, and reporting is a card count.
The nine stages that actually matter
Resist the urge to model every nuance. Nine stages covers almost every agency workflow:
- Prospect Pool — raw finds from Ahrefs, SERP scraping, competitor backlink exports, conference speaker lists. No qualification yet.
- Vetting — someone checks traffic, topical relevance, outbound link patterns, whether the site sells links, and whether you've pitched them in the last 6 months.
- Pitch Queue — vetted and approved, with a contact found and 2-3 topic angles drafted. This is your ammunition stockpile.
- Pitched — first email sent. Follow-up cadence running.
- In Conversation — they replied. Negotiating topic, format, links, or fee.
- Topic Approved — you have a written yes on a specific title, word count and deadline.
- Brief & Draft — handed to writing. This is where the card crosses into your content workflow.
- Submitted — sent to the editor, awaiting response.
- Published — live, link verified, added to the client's backlink log.
Two extra buckets sit off to the side: Rejected (with a reason field) and Nurture (good site, wrong timing — revisit in 90 days). Do not delete rejections. After a year you'll know which prospect sources produce the highest rejection rate and can stop feeding them into the pipeline.
What goes on the card
Keep it under ten fields or nobody fills it in. The ones that earn their place:
- Domain, contact name, contact email, and the URL of the page where you found the contact
- Client (if you're running a shared board across accounts)
- Approved topic and target anchor/URL
- Owner — one named human, never a team
- Next action date — the single most useful field on the board
- Cost, if it's a paid placement (track it; it's a client-billable line item)
- Thread link — a permalink to the email conversation
Run it with WIP limits, not effort
The classic failure mode is a team that pitches 300 sites in week one and then drowns in replies they can't service. Put a work-in-progress cap on the middle stages.
A practical set for one full-time outreach specialist: 25 cards max in Pitched, 20 max in In Conversation, 10 max in Brief & Draft per writer. When In Conversation hits 20, you stop pitching and start closing. It feels wrong for about a week and then your close rate climbs, because every conversation gets a real reply within 24 hours instead of three days.
WIP limits also make capacity planning honest. If a client wants eight links a month and your writers can absorb ten drafts, the board tells you whether that's feasible before you sign the retainer.
Know your conversion math
Once the pipeline has run for 60-90 days, the stage-to-stage ratios become your forecasting engine. A reasonable mid-market benchmark for unpaid, relationship-based guest posting:
- 200 raw prospects → 120 pass vetting (40% junk rate is normal)
- 120 pitched → 40 replies (33% reply rate with a personalised first line and a real topic angle)
- 40 replies → 18 topics approved
- 18 approved → 13 published (ghosted editors, scope changes, sites that go dark)
That's roughly 15 raw prospects per live link. Cycle time from first pitch to published URL typically runs 45-60 days. Both numbers matter enormously for client conversations: if you sign a client on the 1st of March promising links, the first placements land in mid-April at the earliest. Say that out loud during onboarding, not during the month-two review.
Your numbers will differ. Track them per client and per prospect source. One agency I worked with found that prospects sourced from competitor backlink gaps converted at 2.4x the rate of prospects from generic "write for us" searches — so they killed the second source entirely and reallocated eight hours a week.
The handoff that breaks everything
Stage 6 to stage 7 — Topic Approved to Brief & Draft — is where most pipelines leak. Outreach and writing are usually different people, sometimes different contractors, occasionally different time zones.
Make the handoff a hard gate with a definition of done. A card cannot move into Brief & Draft until it has: confirmed title, word count, deadline, target URL and anchor preference, the publication's style guide link, and a note on internal-link rules (many sites cap external links at one). Without those six items, the writer will guess, the editor will reject, and you'll lose two weeks.
This is also where the guest post should enter your normal editorial process rather than living in a parallel universe. Same brief format, same review stage, same QA. If you use contractors for the writing, your standard onboarding and freelance writer workspace setup applies here too — guest posts aren't a special category that justifies a separate process.
Reporting the pipeline, not just the links
Clients see one number — links live this month — and panic when it's low. Show them the board instead. A monthly snapshot like "31 pitched, 22 in conversation, 9 topics approved, 6 drafts in review, 5 published" tells a completely different story than "5 links."
Add a simple forecast: 9 approved topics × 72% publication rate ≈ 6-7 placements landing in the next 30-45 days. That's a defensible number, derived from your own ratios, and it buys you patience during slow months.
The same stage-count view is what you use internally to decide whether to hire. If Pitch Queue has 140 cards and Pitched is capped at 25, you don't have a prospecting problem — you have an outreach capacity problem.
Tooling: what the board needs to do
You need four things: kanban stages, per-card owners, date-based automation (auto-flag cards untouched for 7+ days), and the ability to filter by client. Trello and Notion handle the visual part fine but get awkward once guest posting has to connect to billable hours and client retainers — you end up maintaining two systems. Purpose-built agency tools keep outreach, content production and time tracking on the same record; if you're weighing options, the
