What Makes Influencer Marketing Work?

ContentPan, M.; Blut, M.; Ghiassaleh, A.; Lee, Z. W. Y. · 2025Journal of the Academy of Marketing Science
Topicsinfluencer marketing·social media·meta-analysis·persuasion knowledge·source credibility·purchase behavior·sponsorship disclosure

Imagine your team ranks creators for next quarter's influencer budget by follower count. A credible creator is one followers see as authentic, knowledgeable and appealing. Which creator qualities should lead selection for your campaign's goal?

The strongest influencer factor differs for attitude, engagement, intention to buy and purchases.

Data chart

The strongest influencer factor changes with the goal

How closely each factor moves with attitude, engagement, intention to buy, or purchases, averaged across studies and corrected for measurement error, on a scale from -1 (opposite directions) through 0 (no relationship) to 1 (move together).

No relationshipAttitude, follower count.15Attitude, community identity.53Engagement, follower count.07 (n.s.)Engagement, community identity.52Intention to buy, follower count-.28 (n.s.)Intention to buy, enjoyable content.65Purchases, follower count.21Purchases, direct communication.51
  • The leading factor changes with the campaign's goal.
  • Follower count has no reliable link with engagement or intention.

Community identity leads for attitude and engagement, enjoyable content for intention to buy, and direct communication for purchases.

Table 4, pp. 66 to 67. Values are average relationships, not causal gains. Follower-count estimates use 14, 18, 7, and 3 results, respectively; leading-factor estimates use 34, 27, 9, and 13. The .07 and -.28 follower-count values are not reliably different from zero; the negative estimate does not establish harm.

Data chart

Disclosure has a more positive attitude link for products expressing identity

How closely labeling a post as sponsored moves with consumers' attitude in each product group, averaged across studies and corrected for measurement error, on a scale from -1 through 0 (no relationship) to 1.

.12Products expressing identity-.08Products serving practical needs
  • Disclosure's attitude link differs by product type.

Disclosure has a more positive attitude link for products that express identity than for products serving practical needs.

Table 6, p. 70. The .12 estimate pools 12 results and the -.08 estimate pools 21. The difference between groups is reliable; these are associations, not a before-and-after experiment. Legal requirements and possible trust benefits are outside this comparison.

Action guide

  1. Prioritize direct communication and community building alongside brand fit.Require interactive content and personal replies in creator plans.
  2. Fund useful and enjoyable content for intention to buy.Include demonstrations, reviews, humor and storytelling.
  3. Test platforms built around content or information seeking first.They may suit influencer campaigns better; some estimates rest on one result. Choose interactive creators there who fit the brand and have broad reach.
  4. Match creators and creative to your product type.Use information and interaction for products judged through use, and demonstrations for practical products. For products that express identity, disclose sponsorship and choose creators with large followings and brand fit. Treat comparisons with products checkable before purchase cautiously.
  5. Track engagement and intention to buy together, alongside purchases.Both are closely linked with purchases.

Evidence

  • Among the factors, the strongest links are followers' sense of community with attitude (.53) and engagement (.52), enjoyable (.65) and useful content (.55) with intention to buy, and direct communication with purchases (.51). Links are study averages, corrected for measurement error, on a -1 to 1 scale. Credibility helps explain attitude and engagement.
  • Follower count has small average links with attitude (.15) and purchases (.21), and no reliable link with engagement or intention. Its attitude and engagement links are stronger on platforms built around content or information seeking; some estimates use one result. Its attitude link is .18 for products expressing identity and -.12 for practical products, from five and four results.
  • For products judged through use, brand fit links more strongly with attitude, and interaction with attitude and intention, than for products checkable before purchase. Many comparisons use one result. Brand fit links more strongly with attitude for products expressing identity, and useful content with attitude and engagement for practical products.
  • Disclosure has no reliable average link with attitude, engagement or intention. Its attitude link is .12 for products expressing identity and -.08 for practical products.
  • Engagement's average link with purchases exceeds intention's (.75 versus .68), but intention has the larger link when a model includes both.

Key takeaway

Match creators and content to the campaign's goal, platform and product.

Limitations

Social media studies in 27 countries give the most evidence on attitude, engagement and intention, less on purchases and least on sales. Sales may also depend on price, product quality, inflation and unemployment.

Source

Pan, M., Blut, M., Ghiassaleh, A., & Lee, Z. W. Y. (2025). Influencer marketing effectiveness: A meta-analytic review. Journal of the Academy of Marketing Science, 53(1), 52–78. https://doi.org/10.1007/s11747-024-01052-7

https://doi.org/10.1007/s11747-024-01052-7

Read the paper ↗