The Influencer Marketing Performance Report 2026

This report, co-authored with Sherry Wu at Boston Consulting Group, is, to our knowledge, the first study that tests the influencer marketing industry's stated confidence against its actual investments, tools, and measurement infrastructure at the individual practitioner level. The findings are evaluated against BCG's precision influencer marketing methodology, which demonstrated 4-6x earned media efficiency improvement in pilot campaigns. This is not a market forecast. It is not a sentiment survey. It is a diagnostic.

The diagnosis: the industry knows it has a targeting problem. It does not yet recognise that the problem is its own.

Part 1: What does the industry think its problem is?

Why influencer marketing campaigns underperform according to marketers

Before we can understand where the influencer marketing industry falls short on targeting precision, it is worth understanding where the industry itself believes it falls short. We asked marketers three different versions of this question: why do campaigns underperform, what is the biggest data challenge, and what is the single biggest barrier to more precise targeting. The answers, taken together, reveal how the industry diagnoses itself, and where that diagnosis misses.

When campaigns fail, marketers blame the influencer, the creative, and the brief

We asked marketers to identify the single most common reason influencer campaigns underperform. The answers cluster around three execution failures:

The top three answers, wrong influencer, poor creative, and unclear objectives, are all execution variables. They describe what went wrong in the campaign itself. Combined, they account for 72% of responses.

Notice what is near the bottom of the list. Only 6% identified 'targeting the wrong consumer audience.' Only 3% cited 'lack of consumer research before campaign planning.' Combined, the two answers that point to consumer intelligence as the root cause account for just 9% of responses.

In the industry's own diagnosis, the influencer is to blame three times more often than the audience definition. The creative is to blame three times more often than the research that should have informed it. 

But the top answer is a symptom, not a cause

There is a revealing pattern beneath the surface of the top answer. 'Wrong influencer selection' is cited by 30% of marketers as the primary reason campaigns underperform. It is, by a significant margin, the industry's most common self-diagnosis.

It seems the symptoms are mistaken as root causes, while what should require double-click is taken for granted. Wrong influencer selection is not a root cause but  a downstream outcome. If you have not mapped the consumer community the influencer needs to reach, if you do not know what that community discusses, values, and trusts, then influencer selection is reduced to an educated guess based on engagement rates and audience demographics. Sometimes the guess is right. Often it is not. And when it is not, the diagnosis is 'wrong influencer' rather than 'we did not understand who we were trying to reach.'

The individual-level data confirms this. Among the marketers who identify wrong influencer selection as the primary cause of underperformance:

50% allocate nothing to consumer research before campaign execution. Only 20% use social listening data on what their target consumers actually discuss. Only 30% have tribe identification technology in their stack. Only 40% evaluate content for resonance with the target consumer tribe during review.

The marketers who say 'we keep picking the wrong influencer' are the same marketers who lack the consumer intelligence that would help them pick better ones. They are identifying the symptom and missing the cause. And because they miss the cause, the symptom repeats.

A third question approaches the challenge from a structural angle: what is the single biggest barrier preventing your organisation from implementing more precise consumer-targeted influencer marketing?

Budget is the top answer at 23%. But the cross-tabulated data reveals that all the respondents in the “budget as a barrier” spends $1M or more annually on influencer marketing. Three spend $5M+. They are not under-resourced. They are high-spend programmes that have chosen to allocate their budgets to creator fees and execution rather than to the consumer intelligence that determines whether those fees reach the right people.

Part 2: The problem the industry is not seeing (precision gap thesis)

A $40 billion channel searching for precision

The influencer marketing industry has adopted the vocabulary of precision. Influencer Marketers talk about consumer tribes, audience alignment, and data-driven selection. The language has evolved significantly from the reach-and-follower-count era. But when we examine what sits behind the language, a gap opens up between what the industry says it does and what it actually invests in, builds infrastructure for, and measures against.

We call this the precision gap. It runs through every layer of the survey data: from targeting and selection, through tools and technology, to measurement and content review. It is not a gap in ambition. It is a gap in execution. And it starts with a single, uncomfortable contradiction.

Half the industry says it understands its consumers. Half invest nothing in understanding them.

When asked how well their organisation understands the specific consumer communities their influencer campaigns are designed to reach:

If you stopped reading here, you would conclude that consumer targeting in influencer marketing is in reasonably good shape. Over half the industry appears to have done the work.

When asked what percentage of their influencer marketing budget is allocated to consumer research and analytics before campaign execution:

What kind of research is the other 52% actually doing? When asked whether they had conducted dedicated consumer research specifically to inform influencer campaign targeting in the past 12 months, the responses add texture. 36% have conducted a quantitative consumer survey of 500 or more respondents, which approaches the minimum BCG benchmark. 27% rely on social listening and analytics only, without the primary research that identifies which communities to listen to in the first place. 18% use existing audience data from other channels, meaning the influencer team inherits someone else's consumer model. 9% say targeting is handled entirely by their agency.

36% doing quantitative research at scale is more encouraging than the budget data alone suggests. But 27% relying on social listening without the upfront research to know which communities to listen to is a methodology gap, not an investment gap. Social listening can map conversations once you know which communities matter. It cannot, by itself, identify which communities exist.

The aggregate data is striking. But aggregates can hide nuance. Perhaps the marketers who report strong tribe understanding are the ones who invest in research, and those who spend nothing are the ones who acknowledge weaker understanding. Perhaps the contradiction resolves at the individual level.

It does not.

What happens when you look at individual marketers

When we cross-referenced the individual-level data, matching each practitioner's tribe understanding claim against their actual research budget, the paradox sharpened rather than resolved:

Among those who report detailed, research-backed tribe profiles:

Half of the marketers who say they have detailed, research-backed tribe profiles allocate nothing to research. 71% invest 5% or less. Not a single one invests more than 10%.

We then stress-tested the same group against the capabilities and behaviours you would expect to see if genuine tribe intelligence were present:

At every checkpoint, the confidence exceeds the evidence. Only a third use social listening for selection. Fewer than half have tribe identification technology. Barely more than a third check whether content resonates with the target tribe during review. And only 11% track the efficiency metric that would tell them whether their tribe targeting is actually producing better results.

When BCG uses the word 'tribe,' they mean an emotionally connected consumer community defined through primary research: what the community discusses, what it values, who it trusts, what content formats resonate, and what emotional and functional attributes drive their preferences. Producing a tribe profile in this sense requires a consumer survey of 2,000-5,000 respondents (approximately $50K per market) combined with social listening analysis that maps actual conversations, not just sentiment scores.

When most marketers use the word 'tribe,' they almost certainly mean something different: an influencer platform audience tab that shows age, gender, location, and interest categories. This data is valuable for media planning. But it is not a tribe profile. It tells you who might see the content. It does not tell you whether the content will enter the conversations that drive the community's decisions.

The distinction matters because it determines everything downstream. A brief informed by conversation-level tribe intelligence produces different influencer selections, different creative angles, and different measurement criteria than a brief informed by demographic overlap. The industry has adopted the language of the first while operating the infrastructure of the second.

Everyone says the industry has a targeting problem. Nobody says it is their targeting problem.

The confidence gap does not just show up in budgets. It shows up in self-awareness.

As Part 1 documented, only 9% of marketers and senior leaders in influencer marketing identify targeting the wrong audience or lack of consumer research as the primary cause of underperformance. The top answers are all execution variables: wrong influencer selection (30%), poor creative (21%), and unclear objectives (21%).

Now compare this to the same marketers' assessment of the industry's overall targeting maturity:

Only 3% say the industry targets with high precision. 44% describe it as immature or essentially guessing. These are the same people who, in the previous question, did not identify targeting as the reason their own campaigns underperform.

The industry says targeting is broken. But almost nobody says it is their targeting that is broken.

This is a well-documented psychological pattern called the above-average effect: the tendency to rate oneself as better than average, even in areas where the self-assessment is unsupported by evidence. In this dataset, it is not merely directional. It is quantifiable.

Among marketers who describe the industry as immature or essentially guessing, 47% simultaneously report detailed, research-backed tribe profiles for their own organisation. 60% of those same marketers allocate nothing to consumer research. Only 13% use social listening for selection. Nearly half of those who say the industry is guessing believe they themselves are the exception, while investing nothing in the intelligence that would make them one.

The agency relationship provides the answer. When asked whether their agency provides consumer intelligence as part of influencer campaign planning, the responses reveal a remarkably strong correlation between what the agency provides and what the client believes it has.

31% say their agency provides detailed consumer tribe analysis before campaign planning. Another 31% say their agency provides basic audience demographics but not deeper intelligence. 10% are not sure what consumer intelligence their agency uses. 7% say consumer intelligence is not part of their agency's service.

When cross-referenced: among those whose agency provides detailed tribe analysis, 89% of clients report detailed tribe understanding and only 13% allocate nothing to research. The intelligence is there, the agency is providing it, and the client's confidence is grounded.

Among those whose agency provides nothing, or whose clients are not sure what intelligence is used, the confidence persists anyway. 100% still claim detailed tribe understanding. 100% allocate nothing to research. The confidence is entirely disconnected from the evidence.

The precision gap is, to a significant degree, a capability gap. In many cases where brands rely on external services / do not have inhouse analytical capabilities, it’s an agency capability gap. The quality of consumer intelligence available to the brand is determined by what the agency provides. Where the agency provides depth, brands have depth. Where it does not, brands have nothing, and most of them do not know it..

Where does the unfounded confidence come from?

There is one more layer to the precision gap. 53% of marketers say their first step when planning an influencer campaign is to 'define target audience / consumer segments.' 34% start with KPIs and measurement. 11% start with the creative brief. Zero percent say their first step is to identify potential influencers.

This sounds encouraging. Nobody admits to starting with the influencer. The consumer-first principle appears to have been internalised, at least in stated workflow.

But when cross-referenced at the individual level, the gap between the stated sequence and the actual practice reopens. Among those who say they start by defining the audience: 30% allocate nothing to consumer research. Only 15% use social listening for influencer selection. Only 30% have tribe identification technology.

70% invest something in research, which is better than the aggregate. But only 15% use social listening, the tool that converts a demographic audience definition into conversation-level tribe intelligence. Most are defining the audience from existing CRM data or media plan segments, not from the consumer intelligence that precision targeting requires.

Everybody starts with the audience. Almost nobody researches the audience (or believe existing customer profiles or paid media audience is enough)..

The marketers' own self-assessment confirms this. When asked what percentage of their campaigns were 'precisely targeted' to a well-defined consumer segment or tribe in the past 12 months, 58% said half or fewer. Only 19% said 76-100%. If 51% have detailed tribe profiles that 'inform every campaign,' the precisely-targeted rate should be much higher. Even marketers who claim strong consumer understanding acknowledge, when asked directly, that most of their campaigns do not meet the standard of precise targeting.

Part 3: How the industry actually selects influencers (capability gap)

How the industry actually selects influencers

Part 1 showed what the industry thinks its problem is: wrong influencer selection, attribution difficulty, and budget constraints. Part 2 revealed a deeper issue: half the industry claims consumer tribe intelligence it has not invested in producing.

This section answers the question both parts raise: if not consumer intelligence, then what? When marketers sit down to decide which influencer to work with, what data do they actually use? What tools do they have? And what capabilities have they built, or not built, to make that decision?

The answers explain why the precision gap persists. The industry has built an extraordinarily effective infrastructure for finding and managing influencers. It has not built the infrastructure for understanding whether those influencers reach the right consumers.

94% use engagement rate. 25% use data on what consumers actually talk about.

When asked which data inputs they use when selecting influencers, marketers revealed a hierarchy that tells the story of the precision gap in a single table:

The top five inputs, used by 72-94% of marketers, all describe the influencer: how much engagement they generate, who follows them, how they performed last time, and how large their audience is. These are channel metrics. They tell you about the vehicle, not the destination.

Only one input in the list measures the consumer directly: social listening on what target consumers actually discuss. It sits at the bottom, used by 25%. There is a 3.8-to-1 gap between the most-used input (engagement rate, 94%) and the one BCG identifies as the foundation of precision targeting (social listening, 25%).

One data point in this table deserves particular attention. 72% of marketers say they use 'consumer tribe alignment' when selecting influencers, matching influencer audiences to identified consumer segments. This sounds like genuine consumer-centric selection. It suggests nearly three-quarters of the industry is already doing what BCG recommends.

But only 25% use social listening data on what those consumers actually discuss. And social listening is the input that converts demographic alignment into genuine tribe resonance. Without it, 'consumer tribe alignment' is a proxy activity: matching the age, gender, and location profile of an influencer's audience against a brand's target segment. That is demographic matching, not tribe alignment. It tells you whether the influencer reaches people who look like your target. It does not tell you whether the influencer reaches people who are in the conversations your brand needs to enter.

This 47-percentage-point gap is the operational version of the 51% vs 48% paradox from Part 2. The same definitional collapse is at work. When marketers say 'tribe alignment,' most mean 'the influencer's audience demographics broadly match our target segment.' When BCG says 'tribe alignment,' they mean 'the influencer's audience participates in the conversations, values, and communities we have identified through research and social listening.' Both sides use the same phrase. They mean different things.

What social listening actually changes

The small group of marketers who do use social listening for selection show a meaningfully different capability profile than those who do not:

Marketers who use social listening are 4.2 times more likely to have tribe identification technology. They are 3.7 times more likely to evaluate content for tribe resonance. They are 6 times more likely to track EMV per working spend. Social listening is not just one more data input. It is a marker for an entirely different approach to influencer marketing: one built on consumer intelligence rather than channel metrics.

This does not prove that social listening causes better practice. It may be that the marketers who invest in social listening are simply more sophisticated across the board. But the correlation is strong enough to suggest that social listening functions as a gateway capability: once an organisation invests in understanding what consumers actually discuss, it naturally begins to build tribe identification tools, evaluate content for tribe resonance, and measure efficiency rather than activity.

Sherry Wu's observation adds an important nuance. Even among the 25% who report using social listening for selection, the depth of usage varies. Some may be running genuine conversation analysis that shapes tribe profiles. Others may be monitoring share of voice or campaign mentions, which is social listening in a narrow, reactive sense. BCG's definition requires proactive, strategic use of social listening to identify consumer tribes before any influencer selection begins. The true precision marketers are likely a subset of the 25%, which means the gap may be even wider than the headline number suggests.

For most teams, social listening is occasional, not structural

How often marketers use social listening before briefing influencers reinforces the picture. We asked how frequently teams use social listening data to identify trending topics or cultural conversations before briefing:

Only 24% say social listening is always a core part of their planning. The largest single group, 33%, uses it 'sometimes, when time and budget allow.' For a third of the industry, social listening is a discretionary activity, something that gets done when the schedule permits rather than something the workflow requires.

This has a direct consequence for the quality of influencer selection. If social listening is occasional, the tribe intelligence it produces is occasional. Campaigns planned during a busy period, when time and budget are tight, default to demographic selection without consumer conversation data. The precision of targeting becomes a function of the team's calendar, not the organisation's process.

The tools are built for finding influencers, not for understanding consumers

If the data inputs explain what marketers look at, the technology stack explains what they are equipped to look at. And the pattern is clean: every tool that supports influencer execution is near-universal. Every tool that supports consumer intelligence is adopted by fewer than half.

There is a clean dividing line at approximately 50%. Above it: the tools that help you find, manage, and measure influencers. Below it: the tools that help you understand consumers, identify tribes, and monitor content quality in real time.

100% have discovery tools. 44% have tribe identification. The industry has universalised the execution layer and left the intelligence layer optional.

This is not a criticism of the platforms available. Social listening technology (Brandwatch, Talkwalker, Meltwater, Sprinklr) is mature and widely accessible. Tribe identification is a feature in several enterprise platforms. The tools exist. The issue is procurement priorities: influencer marketing teams budget for the platforms that manage their workflow, not the platforms that inform their strategy.

What happens inside a single-tool stack

28% of marketers manage their entire influencer marketing workflow with a single all-in-one platform. When we examine what capabilities these single-tool users actually have, the execution-intelligence divide appears within their own stack:

Even within a single platform, execution capabilities (64-91%) outpace intelligence capabilities (27-45%) by roughly two to one. The tool is designed for the workflow, not the intelligence. This is not surprising: the market's dominant platforms, CreatorIQ, GRIN, Aspire, Upfluence, were built to solve influencer management problems, not consumer research problems. They do what they were designed to do. The gap is in what the industry has chosen to buy.

Intelligence capability scales with tool count, but not linearly

One of the most revealing patterns in the data is how intelligence capability changes as organisations add tools to their stack:

The two-tool tier is the most striking finding in this table. Zero percent have tribe identification. Zero percent have social listening. Adding a second tool to an all-in-one platform does not add intelligence capability. The second tool is almost always workflow or reporting, another execution tool layered on top of the first.

At three tools, the numbers recover to roughly the single-tool level: 33% tribe identification, 44% social listening. These marketers have likely added a specialised tool, but the intelligence gap remains for two-thirds of them.

At four or more tools, the picture changes entirely: 83% have tribe identification and 83% have social listening. Intelligence capability only becomes the norm when organisations build a diversified stack with dedicated tools for consumer analytics alongside their influencer management platform.

Sherry Wu's observation from BCG's direct experience confirms what the data shows: the intelligence layer and the execution layer require different tools, and no single platform covers both adequately. BCG uses a minimum of three. The survey shows 44% of marketers use one or two.

The industry's weakest capability is its biggest emerging opportunity

When marketers ranked their own team's capabilities from strongest to weakest across eight areas, the results produced a clear hierarchy:

How would you rate your team's (or agency's) capability in the following areas?

Highest ranked

Influencer discovery is ranked first, with a mean of 2.28. Keyword and search strategy is ranked dead last, with a mean of 6.97. The gap between the industry's strongest and weakest self-assessed capability is 4.69 points on an 8-point scale. Not a single practitioner ranked keyword strategy as their strongest capability.

The top three capabilities are the execution skills the industry has been building for a decade: finding influencers, briefing them, and reviewing their content. The bottom four are the intelligence and emerging capabilities: consumer research, social listening, monitoring, measurement, and search.

Keyword and search strategy sits at the bottom by a wide margin. And this matters because search, both social search and AI-powered discovery, is the fastest-growing channel through which consumers find influencer content. TikTok search, Instagram search, and AI chatbot recommendations are increasingly how consumers discover products and creators. The industry's weakest capability is directly in the path of the channel's biggest growth vector.

32% see it coming. 13% are building it.

When asked what they include in influencer briefs regarding keyword and search optimisation:

Only 13% include keyword strategy that covers both social search and AI/LLM discoverability. 34% provide social search keywords but not AI optimisation. The largest group, 34%, provides general hashtag guidance only: a set of tags appended to the post, not a structured keyword strategy embedded in the content itself.

32% of marketers expect search and SEO optimisation of influencer content to be a top-3 effectiveness driver in the next two years. But of those who see it as a priority, 36% provide only hashtag guidance or have not considered search optimisation at all. They recognise the opportunity but have not built the capability.

BCG's approach treats keyword strategy as part of the content brief engine, not as an afterthought. Keywords and emotional elements are researched before the brief is written, then embedded into the caption structure so the content surfaces in both social search and AI-powered recommendations. This is a fundamentally different approach from appending hashtags after the content is produced.

The scale of the opportunity makes the capability gap consequential. An Influencer Strategists research synthesis of 80 academic studies found that influencer content can be engineered to appear in AI recommendations with 91% success rates when properly structured. 63 million US consumers used AI chatbots for purchase decisions in 2026. The teams that build keyword and search capability into their influencer briefs now are building a structural advantage that compounds with every AI model update and every increase in social search usage. The rest of the industry ranks this capability dead last.

From selection to measurement: does the rest of the workflow compensate?

The data in this section shows that the industry selects influencers primarily through channel metrics: engagement rate, demographics, and past performance. Only 25% incorporate social listening. Only 44% have the technology to identify consumer tribes. Only 13% embed AI-optimised keyword strategy in briefs.

This raises a natural question. Even if the selection process is imprecise, could the measurement layer compensate? If the industry tracks the right KPIs, monitors in real time, and uses sophisticated attribution, it might catch targeting errors early, learn from them, and improve over time. A feedback loop could gradually close the precision gap even without upfront consumer research.

How the influencer marketing industry measures success

Part 3 ended with a question: if the selection process runs on channel metrics rather than consumer intelligence, can the measurement layer compensate? Can good measurement catch targeting errors, create a feedback loop, and gradually close the precision gap even without upfront consumer research?

The data in this section says no. The measurement architecture mirrors the selection architecture. It is built to count activity, not to evaluate efficiency. And its two most critical gaps, EMV per working spend and consumer tribe resonance, are precisely the metrics that would tell the industry whether its targeting is working.

91% track engagement. 9% track whether the spend was efficient.

When asked which KPIs their organisation tracks for influencer campaigns, marketers revealed a measurement hierarchy that tells a clear story about what the industry values and what it overlooks:

The top three KPIs, tracked by 78-91%, are activity metrics. They answer: how many people saw the content, how many engaged, how many clicked. These are useful data points. They confirm that a campaign ran and generated activity. But they do not answer the question a CMO ultimately needs answered: was this spend efficient relative to what we could have achieved through other channels or with better targeting?

The middle tier, tracked by 44-72%, includes business outcome metrics. Sales attribution, CAC, and ROAS are meaningful: they attempt to connect influencer spend to commercial results. But even here, 66% track ROAS while only 9% track EMV per working spend. ROAS tells you how much revenue a campaign generated per dollar of paid investment. EMV per working spend tells you how much earned media value each dollar generated, a measure of the amplification efficiency that is specific to influencer marketing and comparable across campaigns, markets, and time periods.

The bottom of the table is where the industry's measurement gap becomes most visible. EMV per working spend, tracked by 9%, is the metric BCG positions as one of two constant hero KPIs that should be tracked across every campaign regardless of objective.

BCG's framework recommends two constants: EMV per working spend (efficiency) and positive sentiment (quality). The industry tracks the second at 56%, which is respectable though not universal. It tracks the first at 9%. The quality indicator exists. The efficiency indicator is nearly absent.

The industry tracks how many people clapped. It does not track whether the performance was worth the ticket price.

The organisations spending the most know the least about their efficiency

If the 9% figure is concerning in aggregate, the cross-tabulation makes it alarming. Among organisations spending $1M or more annually on influencer marketing, zero percent track EMV per working spend. Not one.

This is worth restating because of what it implies. The industry's highest-spend programmes, the organisations deploying $1M, $5M, or more on influencer marketing each year, operate without the efficiency metric that would tell them whether that spend is generating proportional value. They track engagement (91%), impressions (84%), and in many cases ROAS and sales attribution. But they do not track the ratio of earned media value to working spend, which is the metric that makes influencer marketing efficiency comparable across campaigns and channels.

The marketers who do track EMV per working spend, all three of them, operate at significantly lower spend levels or declined to disclose their budgets. Two use social listening for influencer selection. Two report detailed tribe understanding. The profile of the EMV/spend tracker is a smaller, more intelligence-driven operation, not a large enterprise programme.

This creates a structural problem. The organisations with the largest budgets and the most at stake are the ones with the least visibility into whether their spend is efficient. And because they do not measure efficiency, they cannot compare the performance of campaigns with different targeting approaches, which means they have no data-driven way to evaluate whether precision targeting would improve their results. The measurement gap perpetuates the precision gap.

This gap has a compounding irony. When asked what is driving their expected spend increases, 62% of marketers cite 'proven ROI and strong campaign performance.' Another 45% cite shifting budget from other channels, and 45% cite growing consumer engagement with creator content. The budgets are growing because the industry believes the channel works. But the measurement infrastructure to prove it works, the efficiency tracking, the attribution methodology, the standardised frameworks, has not been built. The industry is scaling its investment on conviction, not evidence.

What the $5M+ tier actually tracks

The $5M+ spend tier offers the sharpest picture of this imbalance. These are the industry's most resourced and most operationally mature programmes:

The $5M+ tier outperforms the average on ROAS (75% vs 66%), MMM (38% vs 22%), and real-time monitoring (71% vs 56%). These are operationally sophisticated programmes. But they underperform on EMV per working spend (0% vs 9%), tribe resonance evaluation (13% vs 29%), and search volume uplift (25% vs 44%). They have built the execution and reporting infrastructure. They have not built the intelligence and efficiency infrastructure.

Most marketers who are confident in their ROI cannot actually prove it

24% of marketers say they are 'very confident' they can demonstrate influencer marketing ROI to the C-suite. Another 62% are 'somewhat confident but attribution remains imperfect.' The remaining 14% are not confident or not confident at all.

The 'very confident' group is worth examining closely, because what they are confident about and what they can actually demonstrate are not the same thing.

When cross-referenced against their actual measurement infrastructure:

Every very-confident practitioner tracks engagement and sentiment. These are genuine feedback signals: engagement tells you the content was seen and acted upon, sentiment tells you it was received positively. This is not nothing. It provides real evidence that campaigns are generating activity and that the activity is broadly positive.

But only 29% use marketing mix modelling, which is the methodology that places influencer marketing in the context of all channels and isolates its incremental contribution. Only 14% track EMV per working spend. And none, zero percent, have all three of the measurement capabilities BCG recommends as a complete framework.

Their confidence is real. It is built on engagement and sentiment data, which are meaningful signals. What it is not built on is evidence that the spend was efficient relative to alternatives, or that influencer marketing generated incremental business impact that would not have occurred through other channels. That is the evidence a CFO requires, and it is the evidence the current measurement architecture does not produce.

Of the marketers most confident in their ROI measurement, none have the full measurement infrastructure BCG recommends. Their confidence is in their data collection, not their business impact measurement.

Whether you have a measurement framework determines what you can see

44% of marketers say they have a fully standardised measurement framework applied across all campaigns. 28% are partially standardised. 16% say each market or agency uses its own approach. 9% have no formal measurement framework at all. 3% rely entirely on their agency's reporting.

The difference a standardised framework makes is significant. When cross-referenced against KPI tracking:

EMV per working spend, BCG's recommended hero efficiency metric, is tracked by 21% of organisations with a fully standardised framework. Among those with partial or no standardisation, it drops to zero. Marketing mix modelling follows the same pattern: 43% among the standardised group, near-zero elsewhere.

This suggests that measurement maturity is not evenly distributed. A relatively small group of organisations, those with fully standardised frameworks, has built the infrastructure for efficiency measurement and sophisticated attribution. The rest of the industry, 56%, operates with partial, fragmented, or no measurement standardisation, and consequently has near-zero capability to track efficiency or use MMM.

The implication is important for the precision gap thesis. Without a standardised framework that includes efficiency metrics, an organisation cannot evaluate whether better targeting produces better results. Each campaign is measured in isolation, with different KPIs in different markets, making cross-campaign learning impossible. The measurement infrastructure does not just fail to compensate for imprecise targeting. It prevents the organisation from detecting whether targeting quality matters at all.

The last chance to catch a targeting error, and the industry misses it

Content review is the final checkpoint before influencer content goes live. It is the last opportunity to evaluate whether the content will resonate with the audience it is supposed to reach. The priorities marketers apply at this stage reveal whether the entire workflow, from brief to publication, is oriented toward the brand or toward the consumer.

The top four criteria are brand-centric: does the content match the brief, look professional, sound authentic, and deliver the brand message? These are quality checks. They confirm the brand got what it paid for. They do not ask whether the audience will care.

Only 29% evaluate content for resonance with the target consumer tribe. Only 10% evaluate for keyword or search optimisation. The two criteria most closely connected to whether the content will reach and resonate with the right consumers are the two least likely to be checked.

BCG's framework distinguishes between must-haves (brief compliance, product accuracy, disclosure) and nice-to-haves (trend alignment, emotional hooks, tribe resonance). The survey shows the industry has operationalised the must-haves. 81% check brief alignment. 52% check brand message accuracy. 32% check compliance. These are the table-stakes quality controls.

But BCG's performance differentiators, the nice-to-haves, are where most of the industry drops off. Only 29% check for tribe resonance. Only 13% check for cultural trend alignment. These are the criteria that distinguish content that merely complies with the brief from content that genuinely connects with the audience it is designed for.

Who checks for tribe resonance, and what else they do differently

The marketers who evaluate content for tribe resonance show a markedly different profile from those who do not:

Those who check for tribe resonance are 4 times more likely to use social listening for selection (56% vs 14%) and 2.4 times more likely to report detailed tribe understanding (78% vs 33%). They are the same marketers who invest in consumer intelligence upstream. The review criterion is a reflection of the entire workflow: if the brief is built on tribe intelligence, the review checks for tribe resonance. If the brief is built on brand requirements, the review checks for brand compliance.

There is one counter-intuitive data point: 56% of tribe resonance checkers still allocate $0 to consumer research, compared to 45% of non-checkers. This suggests that some marketers are evaluating for tribe resonance intuitively, based on experience or judgment, rather than against formally researched tribe profiles. They are doing the right check with the wrong benchmark. The intention is correct. The intelligence infrastructure behind it is not.

Without precise measurement, the system cannot learn

The measurement data, read as a complete picture, reveals why the precision gap is self-perpetuating.

A feedback loop that closes the gap would require three things: a precise input (consumer intelligence informing selection), a precise output measure (efficiency tracking showing whether the targeting worked), and a mechanism to feed the output back into the input (using measurement data to refine the next campaign's targeting).

The industry has none of the three at scale. 48% invest nothing in consumer intelligence. 9% track efficiency. And because targeting is measured through activity metrics (engagement, reach, clicks) rather than efficiency metrics (EMV/spend, MMM), the system cannot distinguish between a campaign that generated engagement because it reached the right people and a campaign that generated engagement because the influencer has a large, broadly responsive audience. Both show high engagement. Only one represents precision.

This is why the precision gap is structural, not incidental. An industry that does not measure efficiency cannot learn that better targeting produces better efficiency. And an industry that cannot learn from targeting cannot improve its targeting. The gap does not self-correct. It requires a deliberate intervention: changing both the input (consumer intelligence) and the measurement (efficiency tracking) simultaneously.

The measurement architecture does not just fail to compensate for imprecise targeting. It prevents the industry from detecting whether targeting quality matters at all.

Closing the precision gap is a journey that needs to start now

The data in this report is not a verdict on influencer marketing's potential. It is a diagnosis of where the industry is leaving value on the table. Budgets are growing — 74% of brands are increasing spend — but the infrastructure that would make that spend precise has not kept pace. The good news: the gap is not caused by a lack of intent or belief. Not a single survey respondent said they had tried precision targeting and found it did not work. Nobody disputed the value. The barrier is structural, and structural barriers can be fixed.

Precision is not a single switch to flip. It is a capability stack that compounds over time. Brands that start building now will be two to three years ahead of those who wait. Brands that continue optimising within the current system — better algorithms, more influencers, faster execution — will generate more of the same results, more efficiently. Precision influence starts one layer upstream.

Start with who you are trying to reach — and actually do the research, do the research, do the research!

The most consequential finding in this survey is also the most deceptively simple: 53% claim their first step when planning a campaign, is to define the target audience. But 48% allocate zero budget to consumer research. Starting with the audience means nothing if the audience definition comes from existing CRM segments or media plan tags rather than dedicated investigation.

A consumer tribe profile is not just your typical audience definition. It is a map of what a community discusses, what it values, who it trusts, and what content formats resonate. Building that understanding requires deep social analytical skill, proper social listening tooling and capabilities as well as creativity and strategic thinking. The first step, and the highest-leverage one, is closing the gap between the definition brands claim to have and the one they have actually built.

Select influencers for resonance, not reach

Once the tribe is understood, influencer selection changes fundamentally. It’s not just about thresholds of historical performance or brand fit, they need to be part of the target consumer tribes as trusted voices. 72% of practitioners say they use consumer tribe alignment as a selection input. Only 25% use social listening data on what those consumers actually discuss. The gap between these two numbers is the precision gap in operational terms: a 47-percentage-point distance between claiming tribe alignment and having the data to make it real.

Selecting for resonance means asking a different question. Not "does this influencer reach our demographic?" but "does this influencer's content enter the conversations our target tribe is already having?" “does this influencer’s content tap into our target consumers’ culture?” That question requires social listening. It is not available in a standalone engagement rate.

Brief for the consumer, not just the brand

Most influencer briefs are written from the inside out: what the brand needs the content to say, show, and comply with. That orientation produces content that satisfies the client but may never enter the consumer's world.

The brief is where tribe intelligence either gets used or gets wasted. If the research identified what the target community talks about, what emotional hooks drive their decisions, and what content formats feel native versus intrusive — that intelligence belongs in the brief, not in a strategy deck that nobody reads during execution. The performance difference between campaigns that merely run and campaigns that resonate almost always lives here: in whether the briefing translated consumer insight into creative direction, or stopped at brand requirements.

A useful frame is the distinction between compliance and resonance. Compliance criteria — correct product usage, brand visibility, disclosure, call to action — are necessary but not differentiating. Every brand enforces them. Resonance criteria — does the content tap into a conversation the tribe is already having, does it use the emotional register the research mapped, does it feel like it belongs in that community's feed — are where precision campaigns separate from the pack. Most brands review for the first and treat the second as a bonus.

The brief is also the right place to address an emerging capability gap that the data makes impossible to ignore. Search and keyword strategy for influencer content ranks dead last as a self-assessed team capability — yet nearly a third of practitioners expect it to be a top-3 effectiveness driver within two years. The disconnect is not about awareness; it is about operationalisation. Embedding keyword guidance into briefs — for social search on TikTok and Instagram, and for AI-powered discovery reaching tens of millions of consumers — is not a technical afterthought. It is a distribution decision made at the briefing stage. The brands that treat it that way now will have a compounding structural advantage. The ones that treat it as a caption problem for the creator to solve will not.

Amplify what the tribe has already validated

Precision targeting generates a signal that most organisations fail to act on: real evidence of what resonates with the consumer community, at scale, in the wild. High-performing content is not just a good result — it is a validated creative asset that can be extended through paid amplification to reach more of the same tribe, at a fraction of the cost of producing new content from scratch.

The industry knows this intellectually. The IAB reports paid amplification of creator content growing at 48% year-on-year. In practice, most organisations cannot easily act on it because the teams that make the creative decision and the teams that control amplification budget are structurally separate. Influencer and paid media sit in different parts of the organisation, operate on different planning cycles, and often have no formal workflow for connecting one to the other.

This is a solvable organisational problem. It does not require a restructure — it requires proximity and a shared planning rhythm. When the influencer team and the paid team review content performance together, boosting decisions happen in days rather than weeks, and the feedback from amplified content informs the next briefing cycle. The flywheel closes. Without that proximity, amplification remains reactive and the creative signal from precision targeting goes largely unused.

Measure efficiency, not just activity

The industry has a measurement infrastructure problem that runs deeper than attribution. Most organisations track what happened — impressions, engagement, clicks — but not whether those results were efficient relative to what the investment could have generated. These are different questions, and conflating them produces confident-sounding reporting that cannot actually answer the question a CFO or CMO ultimately needs answered: was this a good use of the budget?

Two metrics create the foundation for answering that question consistently. EMV per working spend captures execution efficiency: how much earned media value each dollar of paid investment generated, trackable across campaigns, markets, and time periods in a way that engagement rate is not. Positive sentiment captures activation quality: whether the content is building the brand or just generating passive impressions. Together, they provide a minimum viable basis for comparing campaigns, justifying investment, and identifying where precision targeting is actually making a difference. Neither requires new technology. Both require the decision to track them.

What makes measurement a precision gap issue, not just a reporting issue, is the feedback loop it creates — or fails to. If campaigns are measured through activity metrics alone, the organisation has no data to demonstrate that better targeting produces better efficiency. And without that data, the case for investing upstream in consumer research cannot be made in commercial terms. The measurement architecture is not just the end of the campaign. It is the foundation of the next one.

Build the capability stack deliberately

Precision influence is not a project. It is a capability journey — and the data shows the industry has built exactly half of it. Every execution capability in the technology stack has near-universal adoption: influencer discovery (100%), performance reporting (84%), campaign workflow (81%). Every intelligence capability sits below 50%: social listening (47%), tribe identification (44%), real-time sentiment monitoring (44%).

The tools most teams rely on were designed to solve influencer management problems: finding creators, running campaigns, reporting results. They do so well. But they were not designed to answer the upstream question of whether those creators reach the right consumer communities, or whether the content resonates once it gets there. Consumer analytics, tribe identification, and real-time sentiment monitoring require a different category of tooling — and in most stacks, that category is either absent or bolted on as an afterthought.

The pattern that emerges from the data is instructive: adding a second tool to an all-in-one platform almost never adds intelligence capability. The second tool tends to be another execution layer — a reporting add-on, a workflow tool, a fraud detection layer. Intelligence capabilities only become reliably present when organisations build a deliberate, multi-tool stack with dedicated solutions for consumer analytics alongside their influencer management platform. The implication is that closing the intelligence gap is not a procurement decision that happens naturally as teams mature. It requires an explicit choice to treat consumer analytics as a core infrastructure investment, not an optional upgrade.

No single platform covers social listening, real-time monitoring, and influencer management with equal depth — the data coverage limitations, particularly on TikTok and Instagram, mean that what looks like full capability in a product demo often narrows considerably in practice. Building the intelligence layer means accepting that it will require more than one tool, and budgeting accordingly. And this is also perfect time to call out to the leading tooling partners in the field, especially the ones with AI functionality and possibility to help brands simplify their techstack, has a true opportunity ahead of them.

That said, tooling itself is not enough, building the intelligence layer requires deliberate investment across four enabling dimensions:

Methodology and playbook. Establish a Centre of Excellence to develop campaign learnings and precision influence playbooks that local teams can apply without rebuilding from scratch. Make best practice transferable.

Learning and development. Build structured training programmes tailored to team maturity and role — covering both internal teams and agency partners. 44% of survey respondents report only ad-hoc training on consumer intelligence and social analytics.

Operating model. Clarify which capabilities live in-house and which sit with agencies, with clear SLAs and defined workflows between global, regional, and local teams. The data shows that where agencies provide consumer intelligence, brand confidence is grounded. Where they do not, brand confidence persists anyway — disconnected from any evidence.

Analytics, insights, and tooling. Give teams access to the tools that support the intelligence layer, not just the execution layer. Build in-house analytical capability so insights are generated on demand rather than commissioned episodically.¨

Adopt a test-and-learn rhythm and commit to it

Precision influence matures through iteration, not perfection. The most effective teams treat every campaign as a structured learning opportunity: what targeting assumptions did we test, what did the data say, what does that change for next time? Monthly cross-functional reviews — bringing together influencer, paid media, consumer insights, and measurement teams — create the shared learning forum that makes improvement systematic rather than accidental.

Following each campaign, a structured learning session should capture three things: what performed against the tribe brief (and why), what did not (and why), and what gets codified into the playbook for the next market or activation. These insights compound. The brands that build this discipline now will have a playbook in 12 months that their competitors are still trying to assemble.

The precision gap is real, it is measurable, and it can be closed — but only for the organisations that start now. The survey shows that nobody disputes precision targeting would improve results. The barrier is not belief. It is the structural distance between the industry's ambition and its infrastructure. Closing that distance is the work. And the time to begin is before the next campaign brief is written.

Limitations

This is a preliminary study of 41 practitioners. The findings are directional, not statistically generalisable. Several limitations should be considered when interpreting the data.

The cross-tabulated analysis, which is the report's primary analytical method, involves sub-groups that are sometimes small. When we report that "50% of tribe-confident practitioners allocate nothing to research," that represents 7 of 14 individuals. The pattern is consistent across multiple verification points, which increases analytical confidence, but it should not be read with the same certainty as a finding from a 500-person sample.

The sample skews heavily toward agencies and consultants (63% combined). Brand-side representation is minimal at 2%. This means the report predominantly reflects how agencies perceive brand behaviour, not how brands describe their own practice. An agency saying "our clients do not invest in consumer research" is a different data point than a brand saying "we do not invest in consumer research." Both may be accurate. They carry different biases.

All cross-tabulated findings in this report represent correlations within this sample. They do not establish causation. When we observe that social listening users are 4.2 times more likely to have tribe identification technology, this may reflect a causal relationship (social listening enables precision) or common-cause confounding (more sophisticated teams adopt both independently). The report notes this where relevant but readers should apply this caveat throughout.

The survey instrument uses self-reported data. Practitioners may overstate their capabilities (as the tribe-confidence paradox itself suggests) or interpret question language differently. The term "consumer tribe," for example, appears to mean different things to different respondents, which is itself a finding, but it also means that responses to tribe-related questions may not be directly comparable across practitioners.

A full-scale study with 500+ respondents, balanced across agencies, brands, and platforms, is planned for later in 2026 to test whether the patterns identified here replicate at scale and across a more representative sample.