How this playbook works
One role, one folder
Every role you work gets its own folder in Google Drive, named [Role] – [Client]. Workflow 01 creates it for you and fills it with the documents all the other workflows rely on. As the role progresses, you drop CVs and your pipeline tracker in there too.
+ CVs and tracker as the role runs
Workflow 01 builds all of this for you.
The prompts talk to that folder
You run these prompts in Claude Cowork with your Google Drive connected (next section shows you how). When a prompt says read the job spec in the [Role] – [Client] folder, Claude finds that folder in your Drive, reads what is inside, and saves its work back there. No uploading files into the chat every time; the folder is the shared workspace.
In a Microsoft shop? Swap Google Drive for OneDrive or SharePoint and Gmail for Outlook wherever they appear. The Microsoft 365 connector covers all of them, and every prompt in this playbook works exactly the same.
Anything in [brackets] is yours to fill
The coloured brackets are placeholders. Swap each one for your specifics before you hit enter.
Read the job spec in the “[Role] – [Client]” folder.
You typeRead the job spec in the “Senior Consultant – PeopleFirst Consulting” folder.
Set up your connections
Every connector is a one-time setup, then every prompt in this playbook can use it. Here is where they live in Claude.
Open Settings, choose Connectors
You will find it under Customize, near the bottom of the settings menu.
Hit Add and connect a tool
Browse the directory, pick the tool, sign in once. That is the whole ceremony.
Look for the tick
A tick in the Status column means you are connected and ready to run the prompts.
Three rules, then the prompts
Claude Cowork connects Claude to the tools you already run your desk on: M365, Drive, Gmail, your calendar, Slack, Notion, Apollo and loads more (check out the connectors). Each workflow in this playbook is a single prompt you paste in, fill out, and either run on demand or put on a schedule.
Start at Workflow 01
Start from here, it’ll build the role folder that all the other workflows read from. Skip it and the prompts in this playbook have nothing to go off!
Don’t send without review
These workflows are a draft for you to review. Claude can do the assembly, but you need to make the call.
Schedule what repeats
A handful of these run daily or weekly. We’ll show you how to put them on a schedule so they arrive automatically!
Run it safely
- Work in your company Claude workspace, not a personal account, so your organisation’s data settings and access controls apply.
- Candidate CVs and contact details are personal information.
- Keep sensitive details out of prompts: health information, protected attributes, anything beyond basic work eligibility.
Foundations
Role intake pack
Give Claude your brief and it populates everything, job spec, scorecards, interview questions, the whole lot.
Create a Drive folder named “[Role] – [Client]”. From this intake brief: [paste], generate and save into it: (1) a job spec, (2) an intake meeting agenda with pre-session questions for the hiring manager, (3) a structured interview scorecard, (4) a candidate-facing job ad, (5) an interview question bank mapped to the scorecard, and (6) a hiring manager briefing note. Post a summary with links to Notion, and list anything missing from the brief as open questions at the top.
Run this first, every time. Like we said, every other workflow in this playbook reads from the folder it creates.
Sourcing & Targeting
Talent pool sizing report
How many people fit this role in your market, and where to actually find them.
Uses Claude’s built-in web search, no extra connector needed
Read the job spec in the “[Role] – [Client]” folder. Using web search, size the talent pool for this role in [market]: estimated pool size with sources and a confidence level, the ideal candidate profile, a tier 1 and tier 2 breakdown by employer type and seniority, and three sourcing channels ranked by likely yield. One page, saved to the role folder. Where the data is weak or thin, say so rather than guessing.
The confidence level tells you whether to build your sourcing plan around this or go back to the drawing board.
Web searchTarget company list builder
Queries Apollo against your ICP filters and hands back a tiered, scored company list.
Requires the Apollo connector
Search Apollo for companies matching [industry], [headcount range], [location] and [tech stack]. Return [n] companies and score each against this profile: [paste]. Sort into tier 1 (strong fit), tier 2 (possible) and tier 3 (park for now), with a one-line reason each. Save to Notion as an editable table. Do not add anyone to outreach until I have reviewed the tiers.
Profile tiering for outreach
Scores an enriched candidate export against your ICP and flags who to call first.
Export enriched data from your sourcing tool to Drive first
Read the candidate CSV in [folder] in Drive. Score each profile against this profile: [paste]. Produce a priority outreach list: tier 1 first, a one-line rationale per person, and the gaps I should probe on a call. Flag anyone who looks like a false positive so I can check them by hand. Save as a new doc in the folder.
Outreach & Marketing
Recruitment marketing assets
Turns a job spec into a LinkedIn post and a populated Canva design, held as drafts.
Works from your existing Canva templates.
Read the job ad in the “[Role] – [Client]” folder. Write a LinkedIn post for the role: a hook that names the candidate’s likely frustration, a two-line role summary, three concrete reasons to apply, and a call to action. Then open [template name] in Canva and populate the text fields. Save everything as drafts; nothing is exported or posted until I approve it.
Candidate email drafter
Drafts first-touch, follow-up or decline emails built from your role brief so they’re never generic.
Read the role brief in the “[Role] – [Client]” folder. Draft a [first-touch / follow-up / decline] email to [candidate name], using the two selling points from the brief that best match their background. Under 80 words, plain language, no recruiter clichés. Save as a Gmail draft for my review.
Inbound triage + call booking
Screens replies against your criteria and queues the right next message for every candidate.
Read the “[Role] – [Client]” folder: job spec, FAQ and screening questions. Check my inbox for replies to the [campaign] outreach. For each reply, draft the right next step: screening questions if unanswered, a Calendly booking invitation for [talent partner] if they meet the criteria, or a warm decline if they do not. Never send anything. Leave each as a Gmail draft with a one-line reason, and summarise the batch in Notion for my morning review.
Spend a few minutes reviewing these each morning. It’s crucial that every message out is still yours.
Screening & Interviews
Candidate shortlist brief
Reads every CV in the role folder and returns a ranked shortlist against the scorecard.
Read every CV in the “[Role] – [Client]” folder and score each candidate against the scorecard. Produce a ranked shortlist: name, score, one-line rationale and a suggested interview order. Flag hard stops (visa, location, salary) separately, and where a CV is ambiguous, say so rather than marking the candidate down. Save to the role folder for my review.
Interview scheduling pack
Finds the availability, drafts the invite, and briefs your interviewer from the scorecard.
Check my calendar for three available [30/60]-minute slots next week, avoiding [constraints]. Draft a scheduling email to [candidate] offering the slots. Then build an interviewer briefing from the scorecard in the “[Role] – [Client]” folder: what to assess, questions per dimension, and what a strong answer sounds like. Save the briefing to Notion and leave the email as a draft.
Post-interview debrief summary
Turns rough interviewer notes into a clear, structured and evidence-based debrief!
Here are my notes from today’s [role] interview: [paste]. Read the scorecard from the role folder and write a structured debrief: a score per dimension with the evidence behind it, strengths, risks, and a hire / no-hire recommendation with reasoning. Where my notes do not cover a dimension, mark it not assessed rather than inferring. Save to the role folder.
If you don’t have the evidence, write “not assessed”.
Close & Reporting
Offer letter generator
Builds a personalised offer letter from your template and the candidate’s details.
Candidate: [name]. Role: [title]. Package: [$ base + super + variable]. Start date: [date]. Read the offer template in the [client] folder in Drive and generate a complete, personalised offer letter. List every number at the top for my confirmation before I send anything. Save to the role folder and leave as a Gmail draft.
Weekly client pipeline update
Reads your tracker and drafts the client update, with blockers and next actions.
Read my pipeline tracker for [role] in Drive. Write a client update: candidates in play and their stage, movement since last week, blockers, and recommended next actions with owners. Direct tone, no padding. Save to Notion and leave as a Gmail draft addressed to [client contact].
Slack hiring update
Posts a formatted weekly hiring update to your internal channel, after you approve it.
Scheduled posts, not real-time triggers
Read the pipeline tracker for [roles] in Drive. Write a concise hiring update formatted for Slack: candidates in play, stage, week-on-week movement, with any role that needs urgent attention flagged at the top. Share it with me as a draft first; once I approve, post to [#channel].
Recruiter activity dashboard
Turns weekly activity data into a team performance summary with one action per person.
Read this week’s activity data from [spreadsheet] and our benchmarks from Notion. Produce a team summary: output per recruiter, pipeline conversion by stage, RAG status against target, and one specific action per person. Flag any data that looks wrong or incomplete before drawing conclusions. Save to Notion for my review before it goes anywhere near the team.
This is the workflow recruitment teams outgrow first. If the spreadsheet becomes a bottleneck, see the back page ;)
Put it on a schedule
Five steps and the workflow runs on cadence, ready in your morning review queue.
Open Cowork
In Claude, switch to Cowork mode.
Connect your tools
Link the connectors the workflow needs. You can do this once per tool (not once per prompt).
Paste and fill
Drop the prompt in and replace every [bracket] with your specifics. Vague brackets in means you’ll get vague drafts out, but put the time in for this and you’ll see the rewards.
Type /schedule
Trigger the scheduler inline.
Set the cadence and save
Choose the frequency. From here it’ll automatically arrive.
Which ones to schedule
Daily: 07 Inbound triage. Weekly: 12 Client update, 13 Slack update, 14 Activity dashboard. At role intake: 01 Intake pack, 02 Pool sizing. The rest fire when you need them.
When you want to level up from prompts
These workflows will genuinely give you back hours every week. What we see next, in almost every recruitment business we work with: the pipeline lives across spreadsheets, reporting is stitched together by hand, and somebody has built a homegrown dashboard on a local server that collapses or gets a bug the moment the whole team starts relying on it.
Neoma builds bespoke recruitment dashboards on secure, hosted infrastructure: pipeline forecasting, placement analytics, consultant trackers, ratio modelling and finance summaries, connected live to your CRM through its API. We have already built this for recruitment businesses, and we design against the CRMs recruiters actually use, including JobAdder, Bullhorn and Vincere.
Start with a scoping day
We spend a day on site with you: review what you already have, assess your CRM’s API and your data security, and hand you a fixed cost and timeline. From there, commit to the full build or take it module by module.
Download the PDFTake the whole playbook with you
neoma.com.au · AI capability uplift and bespoke builds for recruitment businesses · neoma.com.au