An audience is not a volume, it is a composition. The Radar maps who follows you, who engages, who is missing. It is part of the Audience / Ecosystem / Conversations triptych: three complementary readings whose order matters.
The Audience radar works with two other inseparable readings. The audience - who follows me and how they engage. The ecosystem - which balance of power structures my sector. The conversations - which narratives circulate on my issues. Taken in isolation, each part produces a partial truth.
In practice, a radar on its own lets you judge an editorial strategy. A radar cross-referenced with the ecosystem reveals the strategic gaps. A radar cross-referenced with conversations lets you choose when to intervene. We rarely sell a radar on its own: we position it within the Understand → Act cycle.
The radar comes in after the societal audit, when the organisation needs to move from a view of issues to a fine-grained reading of its channels. It is an editorial decision tool, not an analytics dashboard.
The radar creates value when an editorial or strategic decision is on the table. Outside those moments, it is an impressive but inoperative deliverable.
To find out whether the current line speaks to the right audience and which themes generate qualified engagement.
To draw up an objective baseline and negotiate editorial priorities on data, not on intuition.
To identify the internal and external spokespeople who already drive engagement, and calibrate the targets to recruit.
To quantify current influence, compare it with peers, and pin down objectively what is missing to reach the strategic targets.
Having 100,000 followers of whom 80% are students, for a B2B banking brand, is a large but strategically poor audience. The radar's first task is to break the audience down by function, seniority, industry, geography and profile type (decision-maker, journalist, analyst, politician, student, competitor).
On LinkedIn, we use the Followers API and the declarative profile data to segment finely. On YouTube, we analyse recurring commenters. On X and Instagram, we read engagement - who likes, comments, shares - rather than the follow, which is less meaningful.
The deliverable is not a generic breakdown such as “34% managers”. It is a map of strategic profiles, with for each one: current weight, trend, relative engagement, and a qualitative reading of the fit with the ambition.
On the audience analysis of a luxury mid-cap centred on its executive's voice, the radar showed that the personal account reached more art directors and fashion journalists than the house account. Direct consequence: the editorial set-up was rebalanced so that product announcements go through the brand and substantive statements through the executive. Without this diagnosis, the house would have kept pushing its societal topics on the corporate account, where they were diluted.
A post that gets 500 likes from 80% students is worth less than a post that gets 80 likes from 60% target decision-makers. The radar crosses content performance with the composition of likers and commenters to produce a qualified engagement score.
In practice: each post is scored by theme (CSR, innovation, product, culture, recruitment, opinion piece), by format (image, video, carousel, long text), by author (brand versus executive voice), and by time slot. We then look at the composition of those engaging for each combination.
Output: a theme × profile matrix that says clearly which editorial axes work with which targets. This matrix feeds editorial line decisions, not just monthly reporting.
For a CAC 40 executive for whom we built a continuous Social Impact Score, the matrix showed that substantive posts on decarbonisation reached 60% investor/analyst profiles, whereas more institutional posts circulated in a closed loop among HR peers and employees. This gap justified doubling the cadence on the decarbonisation pillar and cutting internal announcements to the strict minimum - a trade-off that stands up at committee because it rests on the real composition of those engaging, not on intuition.
A brand speaks, but it is people who are heard. The radar systematically measures the comparative performance of brand accounts and personified accounts (executive, communications department, business leaders, ambassador employees).
For each spokesperson we qualify: own audience, engagement rate, composition of those engaging, and thematic alignment with the brand's strategic pillars. We then identify the gaps: one executive reaches 3x more journalists than the brand, one business director reaches prospects but not regulators.
This dimension feeds the advocacy and spokesperson recommendations: who to push, who to coach, who to leave in a supporting role, who to activate first on which issues.
On engagements where the executive's voice is central, the radar feeds directly into the Social Impact Score editorial strategy: house messages aligned with the strategic pillars, an editorial calendar calibrated by author and by theme, and an internal validation grid. The writing remains the work of a senior consultant in direct dialogue with the executive - no automated ghostwriting, and several rounds of exchange before publication. This editorial discipline is what separates a credible voice from a stream of standardised posts.
The right target but passive. E.g. public decision-makers who follow without commenting. Action: more engaging formats, personification by a credible spokesperson.
The right target and active. E.g. sector HR directors who comment and share. Action: capitalise, name them in the content, activate them in advocacy.
Noise audience. Students for a B2B brand, sector enthusiasts. Little value, little effort. Action: let it be, do not over-optimise.
A noisy but non-strategic audience. Risk of pulling the editorial line towards content that is popular but inoperative. Action: adjust, do not be carried along.
An audience radar taken in isolation answers “am I speaking to the right people?”. Supplemented by a peer benchmark, it answers “do I carry more or less weight than my competitors with those audiences?”.
We select 2 to 4 relevant peers (not always direct competitors: sometimes they are neighbouring organisations in the issue ecosystem). For each: the same analysis grid as the client brand. We produce a comparative reading of audience quality, themes that perform, active spokespeople and cadence.
The benchmark serves two purposes: defending an editorial ambition internally (“look, peer X does this, and reaches this audience”), and detecting blind spots (“nobody in the sector speaks to this audience that we could capture”).
On the engagements we support, the benchmark is what secures the decision at committee. A communications department can argue internally on the basis of intuition; it cannot defend a budget without comparing its reach with that of its peers. The radar produces this comparison in a uniform way - same grid, same scope, same tools - which no hand-made benchmark can guarantee.
The radar draws on a proprietary toolbox linked to the official APIs. LinkedIn Followers API for audience composition. YouTube Data API for commenters and channel history. Python scripts developed in-house for profile segmentation and the detection of strategic profiles by keywords and company patterns.
The raw output is a set of multi-tab Excel files processed in Notebook SV for automatic classifications (by function, sector, geography, theme). The final PowerPoint is generated through automated templating to ensure consistency and reduce layout time.
On regulatory issues, we cross-reference with Follaw.sv to check whether the audience contains the parliamentary decision-makers followed on the relevant files. A cross-check that does not exist in standard market tools.
On conversational issues, the radar works with Connexia: the radar is the snapshot of the audience, Connexia is the film of the conversations. Together they feed the Social Impact Score, the barometer we run on executives and brands to track over time the resonance, polarity, engagement and authority of relays. On influence ecosystems, Connexia completes the set-up by mapping the environment of actors that weighs on the issue - taking the radar beyond its audience logic alone and plugging it into a reading of the balance of power.
The radar follows a sequenced process: each step validates the previous one and conditions the next. The same team that calibrates stays in charge until delivery, to avoid losing intent between extraction and presentation of findings.
Client brief, choice of platforms, selection of peers, definition of the strategic profiles to detect. This is where the depth is calibrated.
Multi-channel API collection, extraction of followers and engagers, retrieval of content history over the defined period.
Each platform has its own grid. LinkedIn: jobs, sectors, geographies, engagement by theme. YouTube: commenters, dominant topics, retention. X, Instagram: qualified engagement.
Grouping of profiles into strategic segments. Content × audience matrix. Comparative reading with the selected peers.
Detailed Excel file, annotated PowerPoint, 2-hour presentation workshop. Actionable recommendations, not raw indicators.
The radar is never sold on its own. It sits within a cycle that turns the reading of the audience into an activation strategy.
The societal audit sets out the reading of the issues across five dimensions. The radar goes deeper on the communications / audiences dimension. Without this perspective, the radar remains an isolated editorial audit.
The stakeholder ecosystem (Connexia method) maps the balance of power. Cross-referenced with the radar, it reveals the influential actors that the audience does not yet reach - the famous strategic gaps that structure the roadmap.
Editorial strategy, employee advocacy programme, parliamentary advocacy. The radar conditions the trade-offs: which pillars to invest in, which spokespeople to push, which channels to activate.
Connexia takes over with real-time monitoring. The Social Impact Score scores the trajectory month after month. The radar is rerun at 12 months to measure the change in audience composition, not just volumes.
Because LinkedIn's declarative profile data is the richest on the market for identifying B2B decision-maker audiences: function, seniority, company, sector. No other platform offers this level of structured information at scale.
For the majority of B2B and public affairs issues, no. TikTok becomes relevant in consumer, luxury, tech and lifestyle sectors, or when the aim is to reach an audience under 30. The choice follows from the issue, not from fashion.
LinkedIn: current data via the API (no retroactive history on audience composition, only on published content). YouTube: full history since the channel was created. X and Instagram: limited retroactive data. The analysis is therefore often a snapshot at a point in time, plus a reading of change from the start of tracking.
Follaw completes it on the regulatory side: if the aim is to influence parliamentarians, we check whether the key decision-makers' accounts are present in the audience. Connexia tracks conversations over time, as an extension of the radar. The radar is a snapshot, Connexia is the film.
Let's discuss your strategic challenge. No-commitment first conversation.
