Is influencer marketing worth it?

Two conditions to check before committing budget

The answer turns on two conditions a brand can check in an afternoon, before any budget is approved. The first is measurement capability, meaning a unique discount code or tracked link for every creator in the plan, plus a landing page separate from the homepage so the arriving traffic can be isolated. The 2024 Journal of Marketing study by Beichert and colleagues attributed more than 17 million euros in revenue because every one of the 2,808 sponsored posts in its data set carried its own unique code, spread across 1,698 nano, micro and macro influencers. The second condition is category conversation. Where people already post about the product category unpaid, a sponsored post joins a discussion that is running without it. In a category nobody posts about, the sponsored post is the first thing anyone has said about the product, which is a heavier job than a single placement can do.

Why the standard ROI number should not decide it

The figure quoted in most answers to this question does not survive a check. Influencer Marketing Hub's strategy guide describes the channel as often considered one of the most cost-effective digital channels, often returning $5.78 for every dollar spent. That figure links to another article on the same site, which repeats the number and names no study behind it. The related $6.50 per dollar figure traces to a survey Tomoson published on its own blog, on a page carrying no publication date, no sample size and no described method. Tomoson's own findings date it, since the survey reports blogs as the preferred platform for 37% of marketers and Facebook second at 25%, a media mix from roughly a decade ago. A self-published survey with no disclosed method and no date is not evidence of anything. Any single blended ROI multiple for a whole channel deserves the same treatment, since it averages across categories and price points that have nothing in common.

What the research actually measured

Two Journal of Marketing papers get cited interchangeably in answers to this question, and they measure different outcomes. Leung and colleagues reported in 2022 that a 1% increase in influencer marketing spend increases engagement by 0.457%. That elasticity sits below one, so each added dollar buys less engagement than the dollar before it. Leung and colleagues also calculated that the firms in their data set could increase consumer engagement by 16.6% by allocating the same budget in proportion to those elasticities and to base engagement levels. Beichert and colleagues measured revenue instead, using 2,808 influencer-specific discount codes across 1,881,533 purchases at one European direct-to-consumer fashion retailer, and found that low-followership targeting outperformed high-followership targeting by an order of magnitude across three return on investment metrics. That result comes from one fashion retailer plus field studies run with two more direct-to-consumer firms, and it should not be stretched to considered business purchases.

• Influencers with medium levels of posting activity outperformed the most frequent posters in the Leung 2022 data set.
• Targeting the followers whose interests overlapped most with the brand was not the best strategy in Leung and colleagues 2022, where the relationship formed an inverted U and the middle of the range beat the top.
• Posts announcing a new product launch reduced effectiveness in that same 2022 analysis by Leung and colleagues.
• Engagement in the five posts before the sponsored one explained up to 51% of the direct effect of follower count on all three return on investment metrics, per Beichert and colleagues in 2024.

Engagement findings do not transfer to revenue. Only the 2024 Beichert study attributed sales.

Where the channel earns its budget

What separates a paying campaign from a losing one is mostly the product, not the creative. A first order that breaks even is fine when the second and third orders arrive inside 90 days, and it is a loss when they never come.

• Repeat purchase inside a short window, so a first order that breaks even still pays across the customer.
• A benefit visible on camera in under ten seconds, which is the advantage the format holds over a static image.
• A category people already discuss unpaid, so the sponsored post joins a conversation instead of starting one.
• A price low enough to decide on a phone, or a trial that removes the price decision from the moment of contact.
• Enough separate placements that one strong performer is not the entire result. Two posts produce an anecdote, not a spread that can be read.
• Usage rights agreed before filming, so the post that outperforms can be run as paid media after the organic window closes.

Where it reliably disappoints

Six setups fail often enough that the outcome can be called before the campaign runs. The common thread is that the purchase happens outside any window the tracking can see, or the success criteria were written after the results arrived.

• One-off high-consideration purchases with a long deliberation window, no retargeting and no email capture. Someone who saw the post in March and buys in June is invisible to the report.
• A single post treated as a campaign. One placement produces no distribution of outcomes, so nothing can be learned from it in either direction.
• Traffic sent to a homepage instead of a page built for the audience arriving from that post.
• Success criteria defined after results land, which makes every outcome defensible and none of them informative.
• Defaulting to the highest follower counts available, which the Beichert 2024 data found underperformed low-followership targeting by an order of magnitude in direct-to-consumer settings.
• No usage rights, so the one post that beat the rest cannot be amplified and stops working when the term ends.

Most brands never find out whether it worked

The question stays unsettled largely because the outcome usually goes unmeasured. The Influencer Marketing Hub Benchmark Report 2026, drawn from more than 600 respondents, found that among brands increasing budgets, 89% select brand awareness as a KPI, 35% select conversions and 25% select attributable revenue or sales. The remaining 75% of those budget-increasing brands did not name attributable revenue or sales as a KPI at all. That same Influencer Marketing Hub report found that among respondents answering its budget question, 87.49% expect their influencer marketing budget to increase while 5.55% expect a decrease, and it grouped measuring ROI at 8.70% with attribution complexity at 7.14% for 15.84% of reported challenges. Rising budgets in a channel describe confidence, not measured return.

The disclosure workload that belongs in the budget

Monitoring creator disclosures is a recurring cost that rarely appears in the plan. The Federal Trade Commission's endorsement guidance tells advertisers to have reasonable programs in place to train and monitor members of their network. The FTC does not expect a brand to be aware of every single statement made by a member of that network, though it does expect a reasonable effort to know what those participants are saying. The guidance names the alternative directly: if regular monitoring is too much, the FTC says a brand should probably switch to pre-approval of posts. It also places the ultimate responsibility for clearly and conspicuously disclosing a material connection with the influencer and the brand, not the platform, which makes a creator's error the brand's exposure too. In October 2021 the FTC sent a Notice of Penalty Offenses to more than 700 companies over fake reviews and deceptive endorsements. That notice cited civil penalties of up to $43,792 per violation at the rate then in effect. Whoever reviews posts for correct disclosure needs hours allocated in the plan.

Four questions that settle it for one specific brand

Four questions decide this for one brand better than any industry average can, and all four can be answered before budget approval. A brand that answers no to the first one can still spend in this channel, though it will not be able to evaluate it afterward.

• Can a unique code or tracked link be issued for every creator in the plan? A no here means the return will be unknowable no matter what happens.
• Does the category get posted about unpaid on the platform being bought? Read 50 recent posts under the category hashtag before assuming it does.
• Will the buyer purchase again inside the window the tracking covers? A 12-month replacement cycle and a 30-day attribution window cannot be reconciled.
• Are the landing page and the retargeting audience live before the first post publishes? Both cost less to build in advance than to reconstruct from a campaign that already ran.

Frequently asked questions