# Building great talent maps with Tavi
Source: https://heytavi.com/guides/talent-maps. Last updated October 29, 2026.
A great talent map is built in conversation with Tavi, not in one go. This guide pulls together what’s worked across hundreds of maps built with Tavi in AI assistants such as Claude and ChatGPT: engineering, GTM, executive, finance and specialist searches across the US, UK, Europe and the Middle East. It walks through 12 steps, from the first brief to outreach and keeping the map live.
Each step opens with **in short**, a summary of the step. **Worth knowing** notes add context. **Say this to Tavi** lists prompts you can use in your assistant’s chat, taken from real maps and lightly cleaned up. Clients are described by type, not by name.
## Contents
- [For AI agents using Tavi](#for-agents)
- [Before you start: five things to know](#before-you-start)
- [The 12 steps at a glance](#at-a-glance)
- [Step 1 · Kick off](#step-1)
- [Step 2 · Shape the map clusters](#step-2)
- [Step 3 · Review the first build and fill the gaps](#step-3)
- [Step 4 · Refine and shape the map](#step-4)
- [Step 5 · Map all profiles](#step-5)
- [Step 6 · Rank profiles with smart sorts](#step-6)
- [Step 7 · Review the rankings, share notes, ask questions](#step-7)
- [Step 8 · Build and shape shortlists](#step-8)
- [Step 9 · Score the map and refine](#step-9)
- [Step 10 · Capture map insights and learnings](#step-10)
- [Step 11 · Kick off outreach](#step-11)
- [Step 12 · Keep the map live](#step-12)
- [Beyond LinkedIn: GitHub, research hubs and other sources](#beyond-linkedin)
- [Adapting the steps by role type](#by-role-type)
## For AI agents using Tavi
This guide is written for recruiters, many of whom are not technical. If you are an AI agent connected to the Tavi MCP server, in Claude, ChatGPT or any other client, use it to plan the conversation: what to propose at each step, what to ask, and what a good map looks like. The habits in it come from hundreds of real maps, so keep to them. Where it differs from the server’s own instructions or a tool’s description, follow those.
- The server’s instructions say which Tavi tool carries out each step. This guide stays at the level of the conversation.
- “Map all” loads the rest of a cluster beyond its first sample. It is where most of the cost is, so it comes last.
- When a user asks for a “deep search” to find missing companies (step 3), they want more research into companies and pools, not a paid deep search for one role brief. Confirm with the user before starting one of those.
- The guide’s advice on credits is for the user. Follow the server’s instructions on when to mention costs.
- A shortlist in this guide is a list in a project. Pipelines are saved deep searches, not shortlists.
### Reading the “Say this to Tavi” prompts
Each step ends with things recruiters say at that point. They are written from the user’s side of the chat, so read them as requests you will receive:
- **“Tavi” means you.** Users talk to their assistant as Tavi. “Tavi, let’s build a smart sort…” is a request for you to do it with the Tavi tools.
- **Treat each prompt as the intent behind it,** not a script. “Let’s map all in the top three clusters first” means: agree the clusters and how many profiles to load, then load them.
- **Offer the next habit when the user hasn’t asked.** New users won’t know these prompts. Study the company at kickoff, show who a cut would remove before applying it, check which clusters are worth loading in full, offer an honest score out of 10 before the user shares the map, and offer insights and client questions at the end. Suggest the next one that fits rather than all at once.
- **Be honest when scoring.** The step 9 habit only works if the score names real gaps. Don’t round a 7 up to a 9.
- **Some requests need tools Tavi doesn’t have:** researching a company, reading LinkedIn profiles, building a spreadsheet or drafting messages. Use your own tools for them if you have them and the user agrees. If you don’t, say so plainly.
- **Square brackets are placeholders,** e.g. [client], [reason] or [list].
- **Talk in the guide’s terms:** clusters, map all, smart sorts and shortlists, not tool names or IDs.
## Before you start: five things to know
| Term | What it means |
| --- | --- |
| Cluster | A segment of candidates to source from, e.g. a company type, a company type + geography, or + level |
| Sample | Until you map all, each cluster shows a sample of profiles to review, not the full cluster, and they aren’t ranked yet |
| Map all | Loads every profile in the clusters you choose. This is where most credits are used, so it comes after the map is refined. |
| Smart sort | A ranking you describe in plain language, e.g. “JD fit + gettability”, applied to every profile in a cluster |
| Shortlist | A ranked list of the best people across clusters, ready for outreach |
> **Worth knowing:** What’s loaded in each cluster at first is a sampled set of candidates, not the complete cluster, and it isn’t ranked. Review it to get a sense of the types of profiles in each cluster. Loading all the candidate data straight away would be very credit-intensive, and we believe maps should be refined before that happens.
## The 12 steps at a glance
The steps fall into four phases: **set up** (1–2), **build and refine** (3–5), **rank and review** (6–9), and **act and keep it live** (10–12).
| Step | What you’re doing | What this means | Say this |
| --- | --- | --- | --- |
| 1 | Kick off | Start a new chat, share the JD and everything you know, and ask Tavi to study the company | “Hey Tavi, let’s build a new map. Here’s the JD. Study the company, to understand their context and who and how they hire.” |
| 2 | Shape the map clusters | Tavi proposes 3–5 clusters, the filters and the first smart sorts. React before it builds. | “I like the first three clusters, they’re directionally right. The fourth is less relevant because [reason].” |
| 3 | Review the first build and fill the gaps | Study the sample, answer Tavi’s questions, and push it to go deeper | “Who are our competitors hiring from? Let’s include their feeders.” |
| 4 | Refine and shape the map | Cut the noise until every cluster matches the brief | “How do we reduce and focus the pool of candidates? I want to be efficient with my credits.” |
| 5 | Map all profiles | Load every profile, across all clusters or just the ones that matter most | “Let’s map all in the clusters that matter most first.” |
| 6 | Rank profiles with smart sorts | Describe how you want people ranked; Tavi applies it to every profile | “Tavi, let’s build a smart sort that’s a combo of JD fit and gettability.” |
| 7 | Review the rankings | Look at the top 10–20 in each cluster. Say what’s right, and ask about what isn’t. | “Why is candidate Y ranked so high?” |
| 8 | Build and shape shortlists | Turn the rankings into ranked lists across clusters, then review and refine them | “Tavi, let’s create a top 100 shortlist based on the JD fit + gettability rankings.” |
| 9 | Score the map and refine | Ask for an honest score and the gaps before you call it done | “Can you score this map out of 10? What are we missing? How can we make it better?” |
| 10 | Capture insights and learnings | Turn the map into insights, questions for the client, and a workbook | “Give me 10 core insights from building the map, and 10 questions to ask the client.” |
| 11 | Kick off outreach | Email sequences through your ATS, and LinkedIn outreach with care | “Put the shortlist into a spreadsheet, ready for our ATS.” |
| 12 | Keep the map live | The map is a living thing. Signals keep you up to date. | “What’s changed in the map since last week?” |
## Step 1 · Kick off
> **In short:** This is the beginning of your mapping journey. Start a new chat in your AI assistant, give Tavi everything you know about the role, and ask it to study the company.
Once Tavi is connected to your assistant (Claude, ChatGPT or another app), start a new chat and tell Tavi you want to build a map. Share the JD as a link, a PDF, or even rough notes. Then pass along anything else you know.
### What to share at kickoff
| Share | For example |
| --- | --- |
| The JD | A link, a PDF or rough notes |
| The hiring brief and context | What the hiring manager really wants, why the role is open, what’s confidential (e.g. “they’re also open to a VP”) |
| Companies to focus on | An exclusive list (“only map these”), or a hunting ground Tavi can add to |
| Titles to focus on | The titles you know. Tavi adds the variants and local-language versions. |
| The type of talent | E.g. direct Seed or Series A experience (current or former), non-enterprise, or hypergrowth experience |
| Locations | E.g. NYC within 50 km, or Europe excluding Germany |
No clear view on companies or titles yet? Leave it open. Tavi will propose them, and you confirm before it maps.
### Study the company
Always ask Tavi to study the company you’re hiring for. What do they do? Who have they hired, and where did those people come from? Which companies keep feeding their team? The answers shape the companies you map, the titles you match and the bar you hold candidates to.
| If you’re… | Do this |
| --- | --- |
| An agency | Share the client’s name and website, and ask Tavi to study the company to understand their context and who and how they hire |
| An internal team | Ask Tavi to study your company once. That context is saved across all your searches. |
### Get the companies and titles right from the start
Two things decide who ends up in the map: which companies you search, and which titles you match.
- **Companies:** go beyond the obvious names. Look at the client’s own team’s past employers, direct competitors, and companies selling to the same customer. Match the client’s business model, not just the industry: for a chain of small, leased urban fitness studios, big-box and franchise operators were the wrong pool.
- **Titles:** expand, clarify, then tighten.
- **Expand:** most roles are described several ways. A deployed engineer is also a forward-deployed, solutions or implementation engineer. Add keyword forms and local-language titles.
- **Clarify:** a title is a claim, not a job description. A “GTM Lead” was really a hands-on closer.
- **Tighten:** use exact phrases and exclude off-function words like Sales, Analytics or Support.
> **Tip: keep the map open while you work:** In assistants that can show apps, such as Claude, the map opens next to the chat. If it doesn’t pop up automatically, just ask: “Please open the map here.” Ask again any time it closes. Elsewhere, open the map from the link Tavi gives you.
**Say this to Tavi** (the user’s words at this step):
- “Hey Tavi, let’s build a new map. Here’s the JD: [link].”
- “Study the company, to understand their context and who and how they hire.”
- “Let’s focus on talent with direct Seed or Series A experience, current or former.”
- “This is confidential, but they’re also open to a VP. Useful to see whether we’re more likely to land a VP than a Head of.”
- “Here’s our hunting ground: [list]. Keep it whole, and add any others you think are great for this search.”
- “Also titles, because this role can be described in a few different ways.”
## Step 2 · Shape the map clusters
> **In short:** Before it builds anything, Tavi proposes the clusters, the filters and the first smart sorts. Your job is to react: what looks right, and what would you shape differently?
### What Tavi comes back with
| Tavi shares | What to check |
| --- | --- |
| **Insight** on the role, the company and the wider competitive space | Does it match your understanding of the brief? |
| **3–5 suggested clusters** | Are these the right segments to source from? What’s missing? |
| **Filters** used across all clusters | Location (e.g. “Europe excluding Germany”), titles and experience level |
| **Initial smart sorts** to rank people within each cluster | How people will be ranked. More on these in step 6. |
### What a cluster is
Think of clusters as segments that candidates can be sourced from. Build them in layers:
| Layer | Example cluster |
| --- | --- |
| Industry or company type | Freight startups |
| + Geography | Austria – Freight startups vs UK – Freight startups |
| + Level | Austria – Freight startups, Director vs Austria – Freight startups, VP |
Add the level layer when you have to decide which level to hire for. The talent density in each cluster often shapes the answer: in one search, the Head of pool was 1.7x deeper than the VP pool, so the client advertised the role at Head of.
### What makes a good set of clusters
- **One person, one cluster.** No one should appear twice.
- **Plain names.** Every cluster name should tell you instantly who’s in it, e.g. “Head of – NYC restaurant software”.
- **New title variants go into existing clusters.** Add a new cluster only when the company type is clearly different.
- **Give a dominant employer its own cluster** if it floods one, so it doesn’t crowd out everyone else.
- **Consolidate when there are too many.** Grouping by distance from the client’s world works well: insiders first, then adjacent, then any-industry builders.
- **Improve the map you have.** Don’t start a v2.
> **Worth knowing:** Once you confirm, Tavi kicks off the search and builds the map. Depending on its scope this takes a few minutes, and the first sample of every cluster is ready when Tavi replies.
**Say this to Tavi** (the user’s words at this step):
- “I like the first three clusters, they’re directionally right. The fourth cluster is less relevant because [reason]. Let’s replace it with [company type].”
- “Each cluster should have a VP or Head of focus, not mixed.”
- “Let’s split this cluster by geography: UK and Austria separately.”
- “Don’t add new clusters for new titles. Use these clusters unless the company type is very different.”
- “There are too many clusters. What’s a logical way to consolidate them?”
## Step 3 · Review the first build and fill the gaps
> **In short:** Study the first build, review Tavi’s notes, answer its questions, then push it to go deeper. Encouraging the agent to go the extra mile is where the best maps are made.
Tavi returns an initial build of the talent map, along with notes and questions for you. Study the map, read the notes and respond to the questions. Then push Tavi to fill the gaps. It will kick off deeper research and come back with proposals to fill them.
> **Remember:** Each cluster shows a sample of profiles, not the full cluster, and they aren’t ranked yet. Use the sample to judge whether each cluster holds the right *types* of people.
### Where to look for more pools
| Pool | What it is |
| --- | --- |
| Competitors | Companies selling the same product to the same buyer |
| Feeders | Where the client’s own team, and their competitors’ teams, came from |
| Comparable industries | A different sector with the same kind of product or complexity |
| Same final customer | Industries that sell to the client’s customer, e.g. delivery, procurement and food-safety tools all sell to restaurants |
| AI-native companies | Companies built around AI from day one, where people already work AI-first |
| Similar growth trajectory | Companies at the client’s stage or just ahead, e.g. a fast-scaling Series B |
| Alumni | People who’ve left any of the above |
| Lookalikes and outliers | Companies structurally like the feeders, and smaller, lesser-known companies with high talent density |
Also look at cluster sizes. A cluster that’s much smaller than you’d expect is probably too tightly defined. A huge one probably needs focusing (step 4).
**Say this to Tavi** (the user’s words at this step):
- “Can you do a deep search to make sure we’re not missing any great companies within [vertical]?”
- “I want to make sure candidates with experience at hypergrowth startups are included, regardless of sector. Can you do that?”
- “Who are our competitors hiring from? Let’s be sure to include their feeders.”
- “Why are there so few candidates in this cluster? It feels a little light.”
- “Why are there so many candidates in this cluster? This is way too many. How do we focus it more?”
- “What gaps in the talent pool coverage are there that you recommend we fill?”
## Step 4 · Refine and shape the map
> **In short:** Maps get bigger before they get better, and that’s natural. Before you spend credits loading every profile, cut the noise until every cluster reflects people who actually match the brief.
If you’re seeing thousands of profiles, that’s okay. The aim is to reduce each cluster until it accurately reflects candidates who match the brief. Ask Tavi how to reduce and focus the pool, review its suggestions, and approve the ones that make sense.
This might take one round or a couple. It depends on how clearly you see the role, and how open you’d like to keep things. Work with Tavi until each cluster is the size you want, e.g. from 5,000 people to a focused 2,500.
### Levers Tavi will typically suggest
| Lever | For example |
| --- | --- |
| Titles | Exact phrases only; exclude off-function, too-senior or too-junior titles |
| Experience | A floor and/or a cap set from the role |
| Company size | Size bands (e.g. 51–5,000 people), or dropping the top-paying giants the client can’t compete with on pay |
| Company type | Software companies only; remove consultancies, staffing firms and contractors |
| Industries | For business roles, exclude whole industries that don’t fit, such as pharma, banks or insurers |
| Geography | A tighter radius, or splitting a big region into its own clusters |
| Common-word company names | Exact matches only, or limited to the right region, so namesakes don’t creep in |
### Exclusions to apply everywhere
- The client’s current staff, across every cluster.
- Alumni, if the client doesn’t want rehires.
- Blacklisted employers, applied to *current* employer only.
- Your other clients, as source companies.
- Duplicates, so each person sits in one cluster.
**Don’t filter blind.** If you’re not sure what a cut would remove, ask Tavi to show you who it would drop first.
> **Remember:** You’re still looking at a sample. Every cut you make now means fewer profiles to load, and fewer credits used, when you map all.
**Say this to Tavi** (the user’s words at this step):
- “How do we reduce and focus the pool of candidates? There are way too many right now, and I want to be efficient with my credits.”
- “We have around 5,000 people in the map. How do we get it to a focused 2,500 before I map all?”
- “What company types should we kick out?”
- “Who would that cut remove? Show me a few examples before we apply it.”
- “Have you made sure you’ve removed anyone currently at [client]?”
## Step 5 · Map all profiles
> **In short:** When the map feels right, click “map all”. Tavi loads every profile, across all clusters or just the ones you care about most.
### Before you map all, check that you’ve:
- [ ] Locked in the clusters, including any new ones you added along the way
- [ ] Pushed Tavi to fill the gaps in companies, geographies and titles
- [ ] Shaped and refined each cluster to cut the noise and focus on the profiles that matter
### What “map all” does
Up to now, each cluster has shown a sample of the profiles that exist. When you click “map all”, Tavi accesses *all* the profiles in each cluster. You choose the scope:
| Option | When to use it |
| --- | --- |
| The full pool, across all clusters | You’re confident in every cluster and want the complete picture |
| Specific clusters only | Start with the clusters you care about most, and load the rest later if you need them. This is the most credit-efficient route. |
It may take a couple of minutes.
> **Worth knowing: how credits work here:** Mapping all is where most credits are used, because every extra page of profiles is loaded and billed. A cluster of 3,000 people is about 30 more pages. That’s why steps 3 and 4 come first: every cut you made there makes this step cheaper.
**Say this to Tavi** (the user’s words at this step):
- “Before I map all, how many more profiles match in each cluster? Which ones are worth loading in full?”
- “Let’s map all in the top three clusters first.”
- “It should be ranking from the full 950, not just the first 100 loaded.”
## Step 6 · Rank profiles with smart sorts
> **In short:** Now that every profile is loaded, decide how you want candidates ranked. Describe it to Tavi in plain language, and it builds a smart sort and applies it to every profile in every cluster.
You can rank profiles in whatever way you wish. A good starting point is how well each person matches the JD. From there, add what matters most for this role: how likely you are to land someone, or direct experience of the context you’re hiring for. Each smart sort reshuffles the ranking within every cluster. It’s a powerful way to cut and rank a large pool.
### Smart sorts that work well
| Smart sort | What it ranks for |
| --- | --- |
| JD fit | How closely the profile matches the JD’s must-haves and nice-to-haves. The baseline for every map. |
| JD fit + gettability | Fit first, then how realistic someone is to land: time at their company, pay band vs their company’s tier, equity lock-in, not a job-hopper or a lifer. The default main sort. |
| JD fit + AI-first | Fit, plus hands-on evidence of working AI-first: building AI workflows into their team, or shipping AI features |
| JD fit + hypergrowth | Fit, plus time at a fast-growing, top-VC-backed startup during its scale-up phase |
| JD fit + 0→1 | Fit, plus having built something from scratch: first hire in a function, founding team, a new product or market launch |
| JD fit + talent density | Fit, plus a current company with 150 or fewer people and a $1B+ valuation or $100M+ ARR |
| JD fit + startup fit | Fit, plus 1–5 years at a company with under 250 people |
| JD fit + relocation likelihood | Fit, plus evidence they’d move: past moves, regional ties, languages |
> **Worth knowing: gettability:** If you want to know exactly how Tavi defines gettability, just ask in the chat. It’s at least partly driven by how long candidates have been at their current company or in their current role. To deprioritise people who’ve only just joined, tell Tavi the rule and the date it should be measured to.
### Keep sorts honest
- **Use the same names on every map,** e.g. “JD fit” and “JD fit + gettability”, so everyone knows what they’re looking at.
- **Don’t reward brand names.** Ask Tavi to leave employer reputation and school prestige out of the wording, or famous names float up regardless of fit.
- **Measure tenure at the company, not the role.** A recent promotion shouldn’t count against someone.
**Say this to Tavi** (the user’s words at this step):
- “Tavi, let’s build a smart sort that focuses on JD fit.”
- “Tavi, let’s build a smart sort that’s a combo of JD fit and gettability.”
- “Tavi, let’s rank candidates on the JD plus their direct experience with hypergrowth.”
- “Tavi, let’s build a smart sort that ranks candidates on how AI-first they are.”
- “Tavi, let’s build a smart sort that ranks candidates on how capable and experienced they are in the 0→1 phase.”
- “Tavi, let’s make sure that overall, we’re deprioritising candidates with under 12 months at their current company by December 2026.”
- “How do you define gettability?”
## Step 7 · Review the rankings, share notes, ask questions
> **In short:** Review the top-ranked candidates in each cluster, say the top 10 to 20. How do they feel to you? Share your notes with Tavi. Your feedback shapes the clusters and the rankings from here.
This is where your judgement goes into the map. Tavi turns each piece of feedback into a rule, re-runs the sorts, and the next pass is closer to how you’d rank people yourself.
### What works
- **Say what’s right, not just what’s wrong.** Pointing to someone who’s exactly right tells Tavi what to look for.
- **Ask why.** “Why is this person ranked so high?” usually reveals a missing factor, or a keyword match that isn’t real evidence.
- **Name the pattern, not just the person.** “Too much enterprise experience” or “only one month in the role” each become a rule that catches everyone like them.
- **Look across clusters, not just within them.** If the strongest people keep coming from one cluster, ask for more companies like those. If a cluster rarely produces anyone good, cut or merge it.
- **Ask for a line or two on each of the top 10.** It’s a quick way to see where a sort is drifting.
**Say this to Tavi** (the user’s words at this step):
- “Candidate X is exactly what we’re looking for. Great stuff.”
- “Why is candidate Y ranked so high?”
- “I feel like a lot of these top-ranked candidates have too much enterprise experience. I’m far more interested in startup experience.”
- “I feel like we should be ranking candidates with experience in a hypergrowth setting higher. How should we tackle that?”
- “Who are the top 10 in each cluster? Tell me in a sentence or two about each.”
- “Across the clusters, the strongest people are coming from [cluster]. Let’s find more companies like those.”
## Step 8 · Build and shape shortlists
> **In short:** Once you’re happy with the rankings, create shortlists. They’re the path from the map to action. Start simple, e.g. the top 100 across all clusters, ranked by JD fit.
Just tell Tavi in the chat which shortlists you want, and what to call them. Tavi builds an overall list of the best candidates from across the clusters. Once a shortlist is made, review it and keep giving Tavi feedback on the rankings and the order. It all helps with deeper refinement.
### What works
- **Fewer lists, named exactly.** One main list is often enough. Split by level or location only when the client’s question needs it, e.g. Top 100 London vs Top 100 rest of UK.
- **No quotas per cluster.** Take the best across all clusters, and skip duplicates.
- **Keep the reviewed order.** Once people have been checked, the list should reflect that, not just the raw sort.
- **Flag before outreach:** the client’s customers, competitors, and anyone who’d need a visa.
- **Don’t pad.** If only 62 people truly fit, ship 62.
- **Archive old lists,** so only the current ones show in the map.
> **Worth knowing: validate with your assistant’s browser:** If your assistant can browse the web (for example Claude in Chrome, or ChatGPT agent), it can visit each person’s LinkedIn page and review it in real time, to validate Tavi’s findings. Tavi should always ask if that’s something you’d like to do. Reads typically catch recent job moves, contractors, wrong locations and keywords that aren’t real experience. Grade as you go (A / B / C), and move at a reasonable pace on LinkedIn: one profile at a time.
**Say this to Tavi** (the user’s words at this step):
- “Tavi, let’s create a top 100 shortlist, based on the JD fit rankings.”
- “Tavi, let’s create a top 100 shortlist based on the JD fit + gettability rankings, but focused on candidates in London only. Call it Top 100 JD fit + Gettability London.”
- “Read the full LinkedIn profiles of the top 30 on the shortlist before we finalise it, at a reasonable pace.”
- “Keep the shortlist in the reviewed order.”
- “Archive the old shortlists, so when I click on lists I only see the new ones.”
## Step 9 · Score the map and refine
> **In short (the habit that matters most):** Before you assume the work is done, ask your assistant and Tavi to score the map out of 10, name what’s holding it back, and fix it. Don’t share until it’s a 9.
Your assistant and Tavi are honest about the state of a map, including what’s weak. So check yourself against what they think. Maps often start around 6 or 7 out of 10. The fixes that take them to a 9 are usually concrete:
- New clusters for employers that were missed
- Profile reads that cut false positives
- Merging or rebuilding thin clusters
- A signal the first sort missed, e.g. people who built US teams for European scale-ups
- A final check of the top of each shortlist against the client’s latest criteria
**Ask at every stage,** not just at the end: after the first build, after the gap-filling, and before you share anything.
### Bring in client feedback
The best rankings come after the client reacts to real people. Paste their feedback in raw, then tell Tavi which point matters most.
- **Turn new criteria into a sort.** A CEO’s “hypergrowth plus complex implementations” became the main ranking for one search.
- **Rank, don’t filter.** A signal like fast growth should move people up, not remove people who aren’t on a list.
- **Write in the realism,** e.g. “they won’t attract people from major AI labs” or “they want a startup vibe”.
- **Keep unknowns open.** Don’t let Tavi assume an answer the client hasn’t given. Save the question for the next meeting.
**Say this to Tavi** (the user’s words at this step):
- “Hey Tavi, can you score this map out of 10? How good is it? What are we missing, what are the gaps? How can we make it better?”
- “I don’t want to share this with the client until you score the overall map at 9 out of 10 at least.”
- “Which thin clusters haven’t we touched yet? Let’s fix them now.”
- “Here’s what the CEO said: [feedback]. That’s the main sort we’ll use.”
- “I don’t want to remove candidates just because they’re not on your company list. There’s got to be a better way to tackle this.”
## Step 10 · Capture map insights and learnings
> **In short:** Clients hire on decisions, not lists. Turn the map into insights, questions for the client, and a clear answer to the question you started with.
The map is the evidence. What the client remembers is the answer it gives, e.g. “Head of had 1.7x more strong fits than VP overall, and 2.1x outside NYC, so advertise the role as Head of.”
### What to ask for
- **10 core insights + 10 questions for the client,** before every client meeting. Good questions also raise the map’s score: the client’s answers often close the last gaps.
- **An answer to the question from step 1,** with exact numbers in a table, e.g. VP vs Head of, or pool sizes by country.
- **A “companies you may not know” list,** for clients who want to learn the market.
- **Five short points on how the map was built,** for the meeting.
Pressure-test every insight: does it deserve its place, and is it about *this* client?
### Package it in a workbook
Assistants that can create files, such as Claude and ChatGPT, can put the whole map into one clean Excel workbook. Settle on a template once, then point your assistant at your last good one.
| # | Sheet | What’s on it |
| --- | --- | --- |
| 1 | Overview | How we got here: the brief, cluster definitions, exclusions, and a two-sentence definition of each smart sort |
| 2 | 10 core insights | What the map tells the client |
| 3 | 10 core questions | What to ask the client next |
| 4 | Shortlist | The top 100 in reviewed order, with each person’s cluster, LinkedIn link, A/B/C grade and a one-line note |
| 5+ | One sheet per cluster | Every profile, ranked with the smart sort applied |
- **Mirror the map.** The order in the workbook should match the order in Tavi.
- **Put the LinkedIn link right next to each name.**
- **Build it once, at the end.** Finish the reviews and re-ranks first.
**Say this to Tavi** (the user’s words at this step):
- “Can you give me 10 core insights from building the map, and 10 questions to ask the client to make the map better?”
- “Let’s build on this as a tight analysis. Use bolding, put the data in a table, and compare in-NYC with outside NYC.”
- “And 5 points, not long, on how we built the map.”
- “Let’s build the Excel output for this role: same logic as the last one, so I don’t have to repeat myself.”
- “Describe each smart sort in two sentences, so I can explain it to the client.”
## Step 11 · Kick off outreach
> **In short:** With a shortlist you trust, it’s time to reach out. Run email sequences through your ATS, and use LinkedIn outreach with care.
### 11A · Email sequences
- **Get the list out of the map.** Your assistant can put the shortlist into a spreadsheet, with each person’s cluster, LinkedIn link and notes, ready to import.
- **Connect your ATS.** An integration with your ATS keeps candidates tracked in one place and lets you kick off sequences from there.
- **Personalise from the map.** Each person’s cluster and profile notes tell you why they’re a fit, which is the best opening line you’ll get.
### 11B · LinkedIn outreach
An assistant that can control a browser can help with LinkedIn outreach, but use it with caution.
> **Worth knowing: outreach with care:** Keep volumes low and move at a reasonable pace. Review every message before it’s sent, and send it from your own account. Your LinkedIn account and your reputation are on the line, so a person should always be in the loop.
**Say this to Tavi** (the user’s words at this step):
- “Put the shortlist into a spreadsheet, with the LinkedIn link next to each name, ready to import into our ATS.”
- “Draft a short, personal first message for each of the top 10, based on why they’re a fit. I’ll review them before anything is sent.”
## Step 12 · Keep the map live
> **In short:** A map isn’t finished when the shortlist goes out. It’s a living thing. Keep coming back to it as the search moves, the market shifts and new roles open.
### Signals and map intelligence
Tavi watches the map for you. It notifies you about meaningful shifts and updates in the map on a daily basis, by email, by text or in the map itself, whichever you prefer.
### Ways to keep using the map
- **Revisit the shortlist** as people move, so it stays current.
- **Re-run the smart sorts** whenever new profiles come in, so everyone is ranked.
- **Reuse it for the next role.** The company study, clusters and sorts carry over to the next search for the same client or team.
- **Use it as talent intelligence,** not just for hiring: where the talent pools are, who’s growing, and which companies feed which.
**Say this to Tavi** (the user’s words at this step):
- “What’s changed in the map since last week?”
- “Re-run the smart sorts across all clusters, so the new profiles are ranked.”
- “We have a new role at the same client. Let’s start from this map’s company study and clusters.”
## Beyond LinkedIn: GitHub, research hubs and other sources
Some of the best candidates have thin LinkedIn profiles, or none at all. For the right roles, ask your assistant to look where their work actually lives, then bring what it finds back into the map.
| Source | Best for | How it’s used |
| --- | --- | --- |
| GitHub and open source | Engineering roles | Mine the commit histories of the client’s own dependencies, and of small to medium projects close to their stack |
| Research hubs (papers, preprints, conference proceedings) | Research and ML roles | Find authors on the topics the role needs, then match them to profiles |
| Competitions and hackathons | Early-career and engineering roles | Use wins and selective programmes as evidence of excellence, rather than prestige badges |
| Registries and public records | Regulated industries | Use e.g. medical-device registries to find the right companies, then map their people |
| Funding and growth data | Stage-specific searches | Research raises and stage outside the map, then add the companies to clusters |
People with a LinkedIn profile go into the map as their own cluster, e.g. “GitHub contributors”, so they’re ranked with everyone else. People without one go on a separate sheet in the workbook.
**Say this to Tavi** (the user’s words at this step):
- “Which GitHub repos are most relevant for [client]? Small to medium-sized repos and communities where great engineers hang out. Find the people we’d miss if we only looked at LinkedIn.”
- “For the open-source folks, put whoever has a LinkedIn profile in the map. Those without one can go in a separate spreadsheet.”
## Adapting the steps by role type
The 12 steps hold for every search. What changes is where the talent lives, what makes someone gettable, and what to watch for.
| Role type | Where the clusters come from | What makes someone gettable | Watch out for |
| --- | --- | --- | --- |
| Engineering | Feeder companies, the client’s own team’s history, talent-density companies, GitHub and open source | Pay tier vs their company, equity lock-in, visa status, 12+ months at the company | Many title variants; brand bias in sorts; big-tech pay the client can’t match |
| GTM / sales | Competitors (fair game), adjacent sellers to the same buyer, feeder and lookalike sales teams | Wins new logos vs manages accounts; deal size; company size and stage | Core city first, then exceptional commuters, then relocators |
| Executive / leadership | Peers by level, and step-ups one level down | Step up vs sideways move; no job-hoppers and no lifers | Current SVPs may see a VP role as a step down; get the client reacting to real profiles early |
| Founder / co-founder | Former founders, founders of stalling companies, operators with 0→1 experience | Founders of healthy companies rarely move | Load the full cluster before cutting; use a calibration sheet with the client |
| Finance / fractional | Industry finance teams, fractional and outsourced CFOs | Experience band, but role fit first | Tenure rules need exceptions for fractional people |
| Specialist / medical / scientific | Alumni and feeder employers, registries, non-English titles | Tiny pools, so coverage matters more than gettability | Domain checks; home country first, then a global mirror |
| Early-career / programmes | Students, new grads, early founders, side-project builders | Excellence relative to peers | Reward selection events, not prestige badges |
| Non-hiring maps | Job seekers: the people who’d hire them. Sales lists: buyers vs competitors. Market maps: leaders at target companies. | Sorts rank companies or buyers, not candidates | Research funding and stage outside the map, then add the companies |