Brief
The whole campaign in one place β purpose, audience, background, and the phase-by-phase plan across Email Β· Facebook groups Β· PS swaps. Plan shows these phases on a calendar; anyone can fill in the detail, so leave a cell blank for the person doing it.
Plan Β· what's planned vs what went out
Each campaign's phases with the plan and what actually happened per channel, as it's logged. Build and edit the detail in the Brief.
Research & audience
Build the audience for each channel β filter what we already hold, or find new contacts/groups. Pick a channel below.
Audience filters
Segments aren't linked to contacts yet β Org type / Region / Keyword are the real filters for now. Suppressed and inactive contacts are excluded automatically.
Preview (first 50 matches)
| Organisation | Region | Type | Freshness |
|---|
The preview shows 50; open the full list to review everyone and untick anyone to leave out. Exclusions carry into the send. Then head to Execute β Email.
Show & targeting
Add / research groups
| Group | Followers | Category | Area | Join | Posts done Β· add a phase | Admin / contact |
|---|
Find swap partners
Add / refresh a partner
| Partner | Reach | Policy | Exchange Β· click to log |
|---|
Execute & log
Do the thing and record it. Email sends; Facebook is drafted then posted by hand and logged.
Compose & send
Sends to the audience you built in Research β Email (current filters).
Fires automatically at that time (checked every ~10 min). Recipients are re-resolved when it sends, so contacts added in the meantime are included.
Upcoming scheduled sends
Loadingβ¦
Draft a post Claude Haiku
A starting draft. Groups dislike spam β personalise it, and never post the same text everywhere.
The PS to send partners
One blurb, reused for every swap β trim to each partner's spec (e.g. Pleasance: 40 words + link + square image).
Results & insights
Every send with its opens/clicks, plus sector and personalisation breakdowns β the history the next campaign learns from.
Overall stats
Open rate by org type targeted
| Org type | Delivered | Open rate |
|---|
Personalised vs. generic β open rate
| Delivered | Open rate | |
|---|---|---|
| With {FirstName} merge | β | β |
| Generic (no merge) | β | β |
All sends
| Date | Campaign | Recipients | Delivered | Opens | Clicks | Open rate | Personalised |
|---|
Click a row for opens/clicks detail, if any exist yet.
Insights
Every campaign we've run β including the imported EmailOctopus + Mailchimp history β so the next one is smarter than the last. Read-only; no client data.
Which sectors engage best β open rate by org type
Loadingβ¦
Open rate over time
Personalised vs generic
Audience Γ show β which list performs for which production
Subject-line patterns β by open rate
| Pattern | Sends | Delivered | Open rate |
|---|
Data & sources
Feed the database β upload research or find a contact. Every record is checked and deduped on the way in.
Upload new research (CSV)
Accepts a CSV with an "Email" column plus any of Organisation/First name/Surname/Location/Org
type/Tags (any order, extra columns ignored) β or a headerless list where each row is
Organisation, ..., email@somewhere.org, .... Same forgiving parser used to
rescue the original research spreadsheets, so an ad hoc file still works β but the template
gets First name/Surname right (for the {FirstName} merge) instead of guessing from a single
"Contact name" column.
Research assistant Claude Haiku + web search β costs a small amount per query
Look up a real contact for one organisation via grounded web search. Never invented β if nothing is found, it says so. Review before adding; this does not write to Airtable by itself.
Neighbourhood targeting
Guidelines
The house rules that keep us deliverable and compliant. Most are enforced by the tool β they're here so everyone works the same way.
Sending responsibly
Never blast unverified addresses
Dud addresses are what got the old Mailchimp account blocked. Uploads are deduped and suppressed emails hard-blocked; keep lists clean.
Unsubscribes & bounces are permanent
Anyone who unsubscribes or hard-bounces goes to Suppressions and is never re-added, whatever a future spreadsheet says.
Legitimate interest, not "subscribed"
These are organisational / community contacts reached with a relevant offer, with one-click unsubscribe on every send β not consumers we've claimed opted in.
Per-channel rules
Schools: no attachments, ever
Attachments trip spam filters β send links. Mind the calendar: next-year shows AprβMay, festive Sep/Nov, never in holidays, 12pm/3pm.
Facebook groups: post by hand, respect the group
No posting API exists and auto-posters risk bans β we never use them. Read each group's rules, don't spam, and log what you posted.
PS swaps: match the programme & reciprocate
Only pitch shows in line with a partner's programme, respect each list's spec, and always run theirs in return.
Keeping the data clean
Edit freely β it's a shared, living database
Found a real name behind an info@ address, or a venue's new marketing contact? Fix the record β it's there for everyone on the next show.
Upload, don't duplicate
One database, one record per org β every upload auto-checks against it, so re-adding last year's research updates the record instead of cloning it.