The longer definition
A fuller definition has four parts, and each one is doing work.
Collecting. Talent intelligence is built on data that has to be gathered, not merely queried. Official statistics, company registers, regulatory filings and published professional information frame a population; primary research establishes what is true now. A discipline that only queries existing datasets inherits every gap in them.
Verifying. This is the step that separates intelligence from data. A profile says someone holds a title; verification establishes whether they still do, whether the title means what it appears to, and whether a second independent source agrees. Unverified data presented as intelligence is the most common failure in this field.
Analysing. A dataset is not an answer. Analysis interprets the pattern against a specific decision — what it supports, what it rules out, what it cannot establish — and says so in writing.
To inform decisions. Talent intelligence is defined by its purpose. If the work does not change or confirm a decision somebody was going to take, it is market research rather than intelligence, and it should be bought differently.
The three layers
Almost every talent intelligence question resolves into one or two of these.
The individual layer
Who the people are: current employer, role, level, tenure, location, career history, qualifications and evidenced capability. How reachable they are in practice — notice, visa status, mobility, what would have to be true for them to move.
This is the layer a talent map is built from, and the layer with the strictest legal constraints on what may be recorded.
The organisational layer
How employers are built: function size, spans and layers, geographic distribution, what is held in-house versus outsourced, tenure and attrition patterns, hiring and leaving flows.
This is the layer that answers “how have the people who do this well organised themselves?” — the question behind most organisation design work.
The market layer
How large the total qualified population is, how it is distributed geographically, what it costs to employ on a properly normalised basis, how fast pay is moving, and whether supply is tightening or loosening.
This is the layer that official statistics contribute most to, and the one where a national average is most likely to mislead.
Internal and external intelligence
| Internal talent intelligence | External talent intelligence | |
|---|---|---|
| Looks at | Your own workforce | The market your workforce comes from and competes with |
| Typical data | HRIS, ATS, engagement, performance, pay equity, internal mobility | Market sizing, competitor structure, compensation evidence, supply trends |
| Usually owned by | People analytics or HR systems | Talent intelligence, talent acquisition, or commissioned externally |
| Answers | What is happening to us, and why | Whether that is normal, and what the market will allow |
| Blind spot | No view of what is possible outside | No view of internal capability or performance |
| Bought as | Software and internal analyst capability | Platform licence, commissioned research, or both |
Neither half is sufficient alone, and the gap between them is where most expensive mistakes live. Internal data showing 22 per cent attrition in a function is a fact without a verdict: it might be two points below the market, in which case your retention programme is working, or ten points above it, in which case something is badly wrong. Only external evidence tells you which.
How it differs from adjacent disciplines
These terms are used interchangeably and should not be.
| Discipline | Looks at | Answers | Relationship to talent intelligence |
|---|---|---|---|
| Talent intelligence | External labour markets, with internal data where relevant | Should we, where, at what cost, against whom | — |
| People analytics | Internal workforce data | What is happening inside, and why | The internal counterpart; strongest when used together |
| Talent mapping | A defined population of individuals | Who could do this role, and are they reachable | A core input; the individual layer |
| Workforce planning | Future internal capability requirements | What will we need, and when | Sets the requirement that intelligence tests |
| Labour market intelligence | Economies and skills at a macro level | What is supply and demand doing | The market layer, at lower resolution |
| Competitive intelligence | Competitors as businesses | What is their strategy and position | Overlaps on the employer dimension only |
| Market mapping | An employer landscape | Who operates in this space | A narrower input, often a first step |
| Executive search | A specific senior appointment | Who should we hire | A downstream activity; different commercial model |
Where the data comes from
Three tiers, in ascending order of effort and descending order of availability.
Official statistics and registers
National statistical offices, labour market agencies, company registers, regulatory registers and mandatory filings. Authoritative, citable and free.
Also lagged, aggregated and often blind to mid-sized employers. Excellent for framing a population; useless for establishing what is true inside a named company today.
Published professional information
Public professional profiles, company disclosures, job postings, conference programmes, patents and publications.
Broad and current-ish, but self-reported and unevenly maintained. This is the tier most software products are built from, which is why they are good at shape and weak at fact.
Primary research
Analysts researching named employers and populations directly, corroborating each material finding against a second independent source.
Slow and expensive, and the only tier that establishes what is actually true now. It is the difference between a report that describes a market and one you can act on.
A report built only from tiers one and two will be confidently out of date, because both tiers describe the past. The question to ask any provider is simple: which of these three tiers did this number come from, and when was it collected?
What it is used for
- Headcount and workforce planning — testing whether a plan is deliverable in the markets it assumes, before the budget is committed.
- Location and site strategy — comparing candidate locations on supply, fully-loaded cost, competitor saturation and real catchment.
- Reward and pay positioning — establishing what the market actually pays, on a properly normalised basis, against the right comparators.
- Organisation design — evidencing spans, layers and capability placement against named comparators rather than generic benchmarks.
- Competitor analysis — understanding how a rival has built the function that is out-executing yours.
- Succession and leadership bench — mapping the external bench alongside the internal one. See succession planning.
- Market entry — establishing whether a new geography can be staffed at all, and on what timeline.
- Due diligence — testing whether a target’s claimed capability and bench actually exist, and who would leave on a change of control.
- Diversity strategy — setting targets against the representation that exists in the population you are hiring from.
- Build versus buy — deciding whether a capability can be hired at the pace required, or has to be developed.
How organisations mature into it
Most organisations move up this ladder one rung at a time, and there is no requirement to reach the top.
Anecdotal
- What it looks like
- Market views come from recruiters, candidates and whoever spoke most recently. No data, no record of what was believed last quarter.
- Typical failure
- A hiring plan built on an assumption nobody can source.
Reactive research
- What it looks like
- Research commissioned per decision, usually late, usually when something has already gone wrong.
- Typical failure
- The evidence arrives after the decision has been made.
Planned research
- What it looks like
- Intelligence commissioned ahead of known decision points, with a budget line of its own.
- Typical gain
- This is the rung where most of the value appears, and it is reachable without hiring anyone.
In-house capability
- What it looks like
- A small dedicated team, usually with a platform licence, handling continuous questions and commissioning depth externally.
- Typical failure
- Buying the platform and not the analyst, then wondering why the dashboard has not changed a decision.
Embedded
- What it looks like
- Intelligence is an input to annual planning, reward cycles, location decisions and board papers as a matter of course.
- Typical gain
- Decisions get made earlier and with less argument, because the evidence is already on the table.
What talent intelligence cannot do
Every provider should publish this list. Most do not.
- It cannot predict an individual. It establishes that a population exists and how it behaves in aggregate. It cannot promise a named person will accept an offer.
- It cannot see inside a company. Internal performance ratings, confidential individual pay and undisclosed restructuring plans are not observable, and any provider claiming otherwise is either inferring or doing something you do not want associated with your name.
- It cannot fix a flawed value proposition. If the role, the package and the leadership are not competitive, better evidence will tell you that faster — it will not make the hire happen.
- It does not age well without maintenance. Compensation data in a fast-moving function can be materially stale within two quarters.
- It cannot make the decision. A good report narrows the options and makes the risks explicit. The judgement remains yours, and a provider who claims to remove it is selling certainty rather than evidence.
- It cannot lawfully profile people on protected characteristics. Diversity work is aggregate-level only. Inference from a name or a photograph is not evidence and is not lawful as a basis for individual records.
A short glossary
The vocabulary, used consistently.
| Term | Definition |
|---|---|
| Talent intelligence | Collecting, verifying and analysing data about people and labour markets to inform workforce decisions. |
| Talent mapping | Identifying and profiling the individuals who could perform a defined role, as a named and verified population. |
| Market mapping | Establishing which employers operate in a defined space, usually as a precursor to mapping the people inside them. |
| Labour market intelligence | Macro-level supply, demand and wage analysis for skills in a geography. |
| Compensation benchmarking | Establishing what the market pays for a role, on a stated and normalised basis. |
| Talent pipelining | Building and maintaining relationships with identified talent ahead of a defined future need. |
| Name generation | Producing a list of individuals matching a profile, without the verification or analysis of a map. |
| Fully-loaded cost | Total employer cost including statutory on-costs, not base salary. |
| Addressable population | The share of a qualified population that is realistically reachable, after notice, mobility and visa constraints. |
| Confidence grading | A stated level of certainty against a finding, with the basis for it recorded. |
Talent intelligence by market
Each market page is written for that market rather than translated from a template.
| Market | What changes there |
|---|---|
| United Kingdom | Employer National Insurance and pension change the real cost; ASHE anchors the benchmark; notice periods decide the timeline. |
| United States | No national market — thirty metros with their own wage curves, equity-heavy senior packages and state pay-transparency rules. |
| United Arab Emirates | Free zone versus mainland employers, Emiratisation as a parallel market, and packages split into basic, allowances and gratuity. |
| Germany | Collective agreements set the pay, works councils set the timeline, and the Mittelstand holds capability the big names do not. |
| Netherlands | The CAO sets the scale, holiday allowance sits outside salary, and part-time working makes headcount the wrong unit. |
| Belgium | Three working languages, automatic wage indexation, joint committees and a Brussels institutional market with its own logic. |
| Austria | Fourteen salary payments a year, near-universal collective agreements, and a capital that runs Central and Eastern Europe. |
| Australia | Superannuation inside the package, modern award floors, five capital markets across three time zones and visa-dependent supply. |
Research coverage extends well beyond these eight. These are the markets where we publish a dedicated page because the structural differences are large enough to change how the research itself has to be built. See all eight markets →
Talent intelligence FAQs
What is talent intelligence?
Talent intelligence is the practice of collecting, verifying and analysing data about people and labour markets — inside and outside an organisation — to inform decisions about hiring, pay, location, organisation design and workforce strategy.
In its external form, which is what a research firm supplies, it answers questions about the market: how large a qualified population is, which employers hold it, what it costs to employ, where it is concentrated, and how competitors have structured the same function.
What is the difference between talent intelligence and people analytics?
People analytics looks inward at your own workforce data — engagement, attrition, performance, internal mobility, pay equity. Talent intelligence looks outward at the market your workforce is drawn from and competes with.
The two answer different halves of the same question. People analytics tells you that attrition in a function is running at 22 per cent. Talent intelligence tells you whether that is high for that market, who the leavers are going to, and whether the replacements exist at the salary you are offering.
Is talent intelligence the same as talent mapping?
No, although they are closely related and often bought together. A talent map identifies and profiles the individuals who could perform a defined role — a named, verified population.
Talent intelligence is the broader discipline that uses mapping data, market statistics, compensation evidence and competitor analysis to inform a decision above the role: whether to hire, where to build, what to pay, or whether to develop the capability internally instead. See what talent mapping is.
What are the three layers of talent intelligence?
The individual layer: who the people are, what they have done, where they sit and how reachable they are. The organisational layer: how employers are structured, how big their functions are, what they hold in-house and how long people stay. The market layer: how large the total population is, how it is distributed, what it costs and which way the supply curve is moving.
A decision usually needs two or three of them. A pay review needs the market and organisational layers; a succession plan needs the individual and organisational layers.
Where does talent intelligence data come from?
From three tiers. Official statistics and registers — national statistical offices, labour market agencies, company registers, regulatory registers and published filings. Published professional information — public profiles, company disclosures, job postings, conference and publication records. And primary research — analysts researching named employers and populations directly.
The first two frame a population. Only the third establishes the current picture, which is why a report built entirely from the first two will be confidently out of date.
Who uses talent intelligence?
Talent acquisition and HR leadership, reward and remuneration committees, workforce planning teams, executive and board committees, strategy and transformation functions, private equity and corporate development teams, and increasingly dedicated in-house talent intelligence teams inside large employers.
Do we need a talent intelligence function or can we buy it?
Both models work, and the honest answer depends on volume. If you face continuous questions across many markets, an in-house team with a platform licence is more economical over time. If your questions are periodic, high-stakes and specific, commissioned research costs less and delivers more depth per question.
Many mature organisations run both: a small in-house team with a platform for continuous monitoring, commissioning external research for the decisions that have to withstand scrutiny. We compare the options, including the costs of each, on our platforms and providers page.
What can talent intelligence not do?
It cannot tell you what an individual will do. It can establish that a population exists, how it is paid and how it behaves in aggregate; it cannot promise that a specific person will accept an offer.
It cannot verify what is not observable — internal performance ratings, confidential pay for a named individual, or an undisclosed restructuring. And it cannot make the decision for you. A good report narrows a decision and makes the risks explicit; it does not remove judgement.
What is labour market intelligence?
Labour market intelligence is the market layer of talent intelligence: supply and demand for skills in a geography, wage levels and movement, participation and migration patterns, and the structural forces behind them.
It is usually built from official statistics and modelled data. Talent intelligence adds the organisational and individual layers on top, which is what makes it actionable for a specific employer rather than descriptive of an economy.
What is competitive talent intelligence?
Competitive talent intelligence is the analysis of named competitors as employers: how their functions are structured, how large they are, where they are located, what they pay, what they have stopped doing, who is joining and who is leaving.
It is built from lawful sources and primary research into publicly observable professional information. It does not involve approaching a competitor’s staff under a pretext, and any firm willing to do that is a liability rather than an advantage.
How often does talent intelligence need refreshing?
It depends on the volatility of the market and the decision it supports. Compensation data in a fast-moving function can be materially out of date within two quarters. Structural market data — where capability is concentrated, how competitors are organised — usually holds for a year or more.
A sensible default is quarterly refresh for pay and supply, annual for structure. What matters more than the cadence is that every figure carries the date it was collected, so you can see what has gone stale rather than assuming none of it has.
Is talent intelligence lawful?
Yes, when it is scoped and documented properly. In the UK and EU it runs on a legitimate-interests basis under GDPR, with a documented balancing test. Other markets apply their own regimes — the federal PDPL and the DIFC and ADGM frameworks in the UAE, the Privacy Act and Australian Privacy Principles in Australia, and a patchwork of state laws in the US.
What matters in practice is discipline: professional information only, sources recorded per field, no special-category data compiled against named individuals, and deliverables transferred securely. Our methodology page sets out the standard in full.
What is the difference between talent intelligence and workforce planning?
Workforce planning is the internal process of forecasting what capability an organisation will need and when. Talent intelligence is the external evidence that tells you whether that plan is deliverable.
A workforce plan built without market evidence is a wish list with dates on it. The two are designed to be used together: the plan sets the requirement, the intelligence tests it against the market that actually exists.
What does a talent intelligence report contain?
A written market definition, the method and sources used, the analysis itself, a confidence level against each material finding, the structured dataset behind it, and a statement of what the research could not establish.
The last two are the ones that distinguish a research deliverable from a slide deck. Without the dataset you cannot interrogate the conclusion, and without the limitations you cannot judge how much weight it will bear.
How mature does an organisation need to be to use this?
Not very. The most common first engagement is a single question — is this plan staffable, or are we paying the right amount — bought as a one-off brief for a few thousand pounds or dollars.
Maturity matters for what you do next: whether you build an internal function, licence a platform, or keep commissioning research per decision. All three are legitimate end states.
What is the difference between talent intelligence and recruitment?
Recruitment fills a role. Talent intelligence informs a decision, which may well be the decision not to open the role at all.
The commercial models differ accordingly. Recruitment is normally paid on placement, which creates an interest in concluding that a hire is available. Research is paid as a fixed fee for the work, which is what allows it to report that a market is thinner than hoped.