Formal Professionalization of early-stage Social Media “Influencers”—attitudinal Drivers and Their Relation to Personality Traits

ABSTRACT

By seeking relationships with advertisers, actively posting social media users become “influencers” in the sense of influencer marketing. As such, they combine content based on personal stories or specific capabilities with third parties’ commercial interests. Unlike research on well-known social media celebrities, in this study we focus on nascent media professionals in the early stages of developing from co-consumers to having their first financial success. We conducted a survey of 3500 such “nano” to “micro” influencers from a leading agency’s German database to identify the principal components of attitudes that extant research proposes as drivers of formal professionalization. To achieve this, we operationalized and adapted theoretical concepts to influencers as a new kind of content creator. Characteristically, professionalization paralleled by reach is driven by a combination of established scales on entrepreneurial orientation, manifest in “proactiveness” and “risk-taking,” as well as a brand orientation, resulting from “market mavenism” and “brand engagement in self-concept.” However, a structural equation model not only suggests that there are additional drivers but also links the identified drivers to personality traits based on the five-factor OCEAN model. Hence, becoming an influencer is to some extent catalyzed by socialization; it requires the ability and willingness to entice an increasingly large followership with unique content, which represents both a challenge and an inspiration for incumbent media organizations.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. Influencer agencies like Territory, mavrk, linkilike, buzzador, StarNGage, tubevertise, or mediakix give a tentative boundary to “micro influencer” of between 1,000 and 50,000 followers (and “nano” or “everyday” influencers less than that). Influencers who have already achieved a certain level of stardom within a category have about 1 million followers. Likewise, social media celebrities have about 1 million followers on at least one platform (see also Kolo, 2019).

2. Possible contractual relationships were covered by several categories (“How do you monetize your blog? Check all that apply”): “Selling physical/digital products/experiences,” “Pay per post/video,” “Getting free samples of products in exchange for a review on your blog,” “Text Link Ads/In-text Ads (you get paid by placing text-based ads within the text of your post),” “Product reviews (you get paid with a commission for every sale that resulted from your review),” “Pay per click (you get paid for every click a visitor makes on certain ads),” “Donations,” “Other.”

3. For the time spent on posting (“On average, how much time do you devote to develop your blog posting?”) the following categories were eligible: “Less than 1 hour,” “1–2 hours per day,” “3–4 hours per day,” “More than 4 hours per day.”

4. Followership was measured in categories: “Less than 250,” “250–1000,” “1000–5000,” “5000–10,000,” “10,000–25,000,” “25,000–50,000,” “Over 50,000” (“How many followers/subscribers do you have on the following platforms?” (Facebook, Instagram, YouTube, Blog). We also documented from the replies how many platforms an influencer was using.

5. For revenues we asked: “Approximately how much do you earn per month with your activities as an influencer (including the market value of the free product samples you get for your activities)?” with the possible categories: “Nothing,” “Less than €500,” “€500–999,” “€1000–2499,” “€2500–4999,” “€5000–9999,” “€10,000–24,999,” “€25,000–49,999,” and “More than €49,999.”

6. Although, according to the agency, at least some respondents would not explicitly call themselves an “influencer.”

7. A Spearman correlation between followership and share of influencers spending more than one hour for their posts (see Figure 3a) produced a coefficient of .985 (p < 0.001).

8. A Spearman correlation between followership and share of influencers with contractual relationship beyond barter (see Figure 3b) produced a coefficient of 0.974 (p < 0.001).

9. A Spearman correlation with the followership and share of influencer operating on more than one platform (see Figure 3c) produced a coefficient of 0.941 (p < 0.001).

10. Cronbach’s alpha factor brand orientation = 0.700; Cronbach’s alpha entrepreneurship orientation 2 = 0.730. For the underlying established scales, we obtained a Cronbach’s alpha as follows: brand engagement 0.924, market mavenism 0.935, risk taking 0.830, proactiveness 0.650.

11. A Spearman correlation with audience size produced the following correlation coefficients: brand engagement 0.207 (p < 0.001), market mavenism 0.261 (p < 0.001), risk-taking 0.138 (p < 0.001), proactiveness 0.070 (p < 0.001). For the correlation with formal professionalization, we obtained brand engagement 0.304 (p < 0.001), market mavenism 0.324 (p < 0.001), risk-taking 0.188 (p < 0.001), proactiveness 0.159 (p < 0.001).

12. Quality criteria for PCA resulted as follows: composite reliability = 0.854, average variance extracted = 0.745.

13. All factors of the OCEAN model could be confirmed with at least acceptable (Peterson, 1994) Cronbach alpha values.

14. Data met the assumption of independent errors according to Durbin Watson for formal professionalism (1.742), and model significance was given by stepwise inclusion of the two independent variables with F(2, 2708) = 242.5; p < 0.001.

15. Data met the assumption of independent errors according to Durbin Watson for brand orientation (1.661), and model significance was given by stepwise inclusion of the five personality factors with F(5, 2457) = 124.6; p < 0.001. Durbin Watson was given also for entrepreneurial orientation (1.736), and model significance was given by stepwise inclusion of all personality factors without conscientiousness F(4, 2457) = 40.4; p < 0.001.

16. We also tested for direct influence of personality traits on formal professionalization. However, the total explained variance was very low (6.7%), hence we discarded this assumption. If it exists at all, any direct impact appears negligible.