Measuring consumers’ engagement with brand-related social-media content: Development and validation of a scale that identifies levels of social-media engagement with brands

Journal of Advertising Research, 2016

Bruno Schivinski, George Christodoulides, Dariusz Dabrowski

Journal of Advertising Research, volume 56, issue 1, pages 64–80, March 2016

  • Scale development
  • 17 items, 3 dimensions
  • N = 2,252 + 416
  • Free to use with citation
Download the accepted manuscript (PDF)Published versionUse the scale

Free full text: the authors’ accepted manuscript, identical in content to the published article. Please cite the published version.

Keywords
  • consumer engagement
  • brand-related social-media content
  • consumers' online brand-related activities (COBRAs)
  • CEBSC scale
  • consumption, contribution and creation
  • scale development
  • scale validation
  • social media engagement scale
  • customer engagement measurement
  • brand engagement on social media
  • user-generated content (UGC)
  • electronic word-of-mouth (eWOM)
  • liking, sharing and commenting
  • online consumer behaviour
  • social media metrics
  • hierarchy of effects
  • uses and gratifications
  • brand equity
  • brand attitude
  • nomological validity
  • confirmatory factor analysis
  • structural equation modelling
  • mediation analysis
  • netnography
  • online focus groups
  • survey questionnaire
  • Likert scale
  • Poland
  • Facebook
  • social media marketing
  • advertising effectiveness
  • brand management
  • questionnaire items

A 17-item scale measures how consumers engage with brand-related content on social media at three levels: consuming it, contributing to it and creating it. Developed with three qualitative studies and validated with 2,668 consumers, the scale also shows that engagement builds step by step, from consumption to contribution to creation.

Key findings

  1. Three distinct levels of engagement. Confirmatory factor analysis supported a three-factor, 17-item structure of consumption, contribution and creation of brand-related social-media content (robust ML: χ²(115) = 557.47, CFI = 0.95, TLI = 0.94, RMSEA = 0.05, SRMR = 0.06; validation sample n = 1,126) (Schivinski, Christodoulides & Dabrowski, 2016).
  2. A reliable and valid measure. Composite reliabilities were 0.88 (consumption), 0.92 (contribution) and 0.93 (creation), with average variance extracted above the shared variance between dimensions, supporting convergent and discriminant validity (Schivinski, Christodoulides & Dabrowski, 2016).
  3. Engagement builds step by step. Consumption predicted contribution (β = 0.61) and contribution predicted creation (β = 0.81); the direct effect of consumption on creation became non-significant once contribution was included (indirect β = 0.49), indicating full mediation (Schivinski, Christodoulides & Dabrowski, 2016).
  4. Related to brand equity and brand attitude. In a new sample of 416 consumers, all three dimensions correlated positively with brand equity (r = 0.23 to 0.40) and brand attitude (r = 0.14 to 0.37) (Schivinski, Christodoulides & Dabrowski, 2016).
  5. Most consumers consume more than they create. Average engagement was highest for consumption (M = 3.68), then contribution (M = 2.65) and creation (M = 2.02), on a 0–7 scale (Schivinski, Christodoulides & Dabrowski, 2016).

Use the scale

Consumer’s Engagement with Brand-Related Social-Media Content (CEBSC) scale

Seventeen items in three dimensions (consumption 5, contribution 6, creation 6). Respondents name a brand they follow on social media and report how often they do each activity for that brand. Takes about three minutes. Free to use in research and practice; please cite the article.

Items
17 (consumption 5, contribution 6, creation 6)
Response format
7-point scale from 1 = not very often to 7 = very often, plus “not at all” coded 0
Scoring
Mean of the items in each dimension (range 0–7); higher scores mean higher engagement
Reliability
Cronbach’s α = 0.88 to 0.95; composite reliability = 0.88 to 0.95 across studies
Validated with
2,252 consumers and 299 brands (Study 4); 416 consumers (Study 5); Poland

Ready-to-use files

The questionnaire

Introduction suggested wording

This survey is about the content related to brands that you come across on social media, such as Facebook, Instagram, YouTube, TikTok, X, blogs and online forums. People engage with this content in different ways: some read or watch it (consuming), some like, comment on or share it (contributing), and some post their own content about a brand (creating). There are no right or wrong answers.

Part 1. Your brand suggested wording

Please type the name of a brand that you actively follow on social media.

You will answer the following questions about this brand. Wherever you see [BRAND], think of the brand you named.

Part 2. Consumption Passive engagement: reading, watching and following brand-related content.

Think about [BRAND]. How often do you do each of the following? Select “Not at all” if you never do it. suggested wording

Item 0Not at all 1Not very often 2 3 4 5 6 7Very often
CONS1I read posts related to [BRAND] on social media.
CONS2I read fan page(s) related to [BRAND] on social networking sites.
CONS3I watch pictures/graphics related to [BRAND].
CONS4I follow blogs related to [BRAND].
CONS5I follow [BRAND] on social networking sites.

Part 3. Contribution Interaction with content created by the brand or by others: liking, sharing and commenting.

Think about [BRAND]. How often do you do each of the following? Select “Not at all” if you never do it. suggested wording

Item 0Not at all 1Not very often 2 3 4 5 6 7Very often
CONT1I comment on videos related to [BRAND].
CONT2I comment on posts related to [BRAND].
CONT3I comment on pictures/graphics related to [BRAND].
CONT4I share [BRAND] related posts.
CONT5I “Like” pictures/graphics related to [BRAND].
CONT6I “Like” posts related to [BRAND].

Part 4. Creation Producing and publishing one’s own brand-related content: posts, pictures, videos and reviews.

Think about [BRAND]. How often do you do each of the following? Select “Not at all” if you never do it. suggested wording

Item 0Not at all 1Not very often 2 3 4 5 6 7Very often
CREA1I initiate posts related to [BRAND] on blogs.
CREA2I initiate posts related to [BRAND] on social networking sites.
CREA3I post pictures/graphics related to [BRAND].
CREA4I post videos that show [BRAND].
CREA5I write posts related to [BRAND] on forums.
CREA6I write reviews related to [BRAND].

The item wording, response format, coding and administration procedure come from the article. The welcome text, brand question and block instruction are suggested wording based on the procedure the article describes; adapt them to your study.

How to administer

  • Ask respondents to name a brand they actively follow on social media, and insert that brand name into every item (for example with piped text in Qualtrics).
  • Present each dimension in its own block, on a separate page.
  • Randomise the order of the three blocks, and the order of items within each block, to avoid order effects (as in the original study).
  • Each respondent answers for one brand only.
  • The items work for any platform. To study a single platform, name it in the instructions rather than changing the item wording.
  • For use in another language, translate and back-translate the items, and check measurement invariance before comparing groups or countries.

How to score

  • Code responses 0 (not at all) and 1 to 7 (not very often to very often).
  • Compute one score per dimension as the mean of its items: consumption = mean(CONS1–CONS5); contribution = mean(CONT1–CONT6); creation = mean(CREA1–CREA6).
  • Higher scores mean higher engagement. In the original data, average engagement was highest for consumption (M = 3.68), then contribution (M = 2.65), then creation (M = 2.02).
  • Any cut-off for low or high engagement is arbitrary because the scale is metric; the original study used the bottom and top 10% as thresholds.
  • For latent-variable models, specify three correlated factors, or the hierarchical model consumption → contribution → creation (Figure 2).

Table B. Mean Scores for the Three Dimensions of the CEBSC Scale by Industry (benchmarks)

Industry n Consumption M SD Contribution M SD Creation M SD
Alcoholic beverages 128 3.78 1.50 3.10 1.53 2.22 1.28
Nonalcoholic beverages 225 2.99 1.48 2.44 1.35 1.82 1.16
Amusement and recreation 196 3.76 1.58 2.96 1.51 2.06 1.38
Apparel and accessories 205 4.49 1.51 2.86 1.70 2.10 1.54
Automotive 106 4.70 1.60 2.83 1.74 2.78 2.41
Food and kindred products 152 4.33 1.29 3.07 1.31 2.13 1.33
Hi-tech 609 3.43 1.49 2.50 1.43 1.98 1.30
Mobile operators 324 3.24 1.38 2.26 1.26 1.71 1.22
Perfumes and cosmetics 151 3.91 1.43 2.57 1.56 1.85 1.11
Services 156 4.69 1.63 3.36 1.78 2.59 1.66

N = 2,252 (full data set). Use these as reference values when comparing your own sample, keeping in mind the Polish 2014 context.

At a glance

Design
Five studies: online focus groups, online depth interviews, netnography, scale development (Study 4) and validation (Study 5)
Qualitative studies
25 focus-group participants, 32 interviewees and a netnography produced a pool of 35 items
Scale development
2,578 Polish consumers recruited online; 2,252 valid responses about 299 brands, split into calibration and validation samples of 1,126
Validation
New sample of 416 Polish consumers; brand equity (Yoo & Donthu, 2001) and brand attitude as nomological variables
Analysis
Exploratory and confirmatory factor analysis, structural equation modelling and bootstrapped mediation (Mplus 7.2, robust maximum likelihood)

Figure 1. Confirmatory factor analysis: three-factor CEBSC framework

e11CONS1: I read posts related to Brand X on social mediae21CONS2: I read fanpage(s) related to Brand X on social network sitese31CONS3: I watch pictures/graphics related to Brand Xe41CONS4: I follow blogs related to Brand Xe51CONS5: I follow Brand X on social network sitesCONSUMPTION0.82*0.840.650.630.86e61CONTR1: I comment on videos related to Brand Xe71CONTR2: I comment on posts related to Brand Xe81CONTR3: I comment on pictures/graphics related to Brand Xe91CONTR4: I share Brand X related postse101CONTR5: I “Like” pictures/graphics related to Brand Xe111CONTR6: I “Like” posts related to Brand XCONTRIBUTION0.84*0.890.860.880.630.66e121CREA1: I initiate posts related to Brand Xe131CREA2: I initiate posts related to Brand X on social network sitese141CREA3: I post pictures/graphics related to Brand Xe151CREA4: I write reviews related to Brand Xe161CREA5: I write posts related to Brand X on forumse171CREA6: I post videos that show Brand XCREATION0.90*0.900.810.850.800.680.650.770.51
Standardised loadings and factor correlations, validation sample (n = 1,126). χ²(115) = 557.47, CFI = 0.95, TLI = 0.94, RMSEA = 0.05, SRMR = 0.06; robust maximum likelihood; all coefficients significant (p < 0.001); * path fixed to 1 for identification. Item labels as printed in the article’s figure; the final item wording is in the questionnaire above.

Figure 2. Hierarchical relationship of the three dimensions

CONSUMPTIONCONTRIBUTIONCREATION0.61 (0.02)0.81 (0.02)
Standardised coefficients (standard errors), validation sample (n = 1,126); χ²(115) = 557.47, CFI = 0.95, TLI = 0.94, RMSEA = 0.05, SRMR = 0.06; all coefficients significant (p < 0.001).

Tables

Table 1. Reliability and Validity of the Consumer’s Engagement With Brand-Related Social-Media Content (CEBSC) Scale

α CR AVE MSV CONT CONS CREA
CONT 0.92 0.92 0.65 0.59 0.80
CONS 0.88 0.88 0.54 0.42 0.65 0.77
CREA 0.93 0.93 0.68 0.59 0.77 0.51 0.83

The square root of the average variance extracted values appears in italics. CR = composite reliability; AVE = average variance extracted; MSV = maximum shared squared variance; CONT = contribution; CONS = consumption; CREA = creation.

Table 2. Mediation Analysis

Hypothesis Direct β without mediator Direct β with mediator Indirect β Mediation type observed
Partial mediation Consumption→Contribution→Creation 0.51*** 0.02 (n.s.) 0.49*** Full mediation

Bootstrapping sample = 5,000. *** p > 0.001; n.s. = nonsignificant.

Table 3. Reliability and Validity Scores of the CEBSC Scale
α CR AVE MSV BEQ CONS CONT CREA BA
BEQ 0.93 0.93 0.70 0.57 0.84
CONS 0.88 0.92 0.66 0.42 0.40 0.81
CONT 0.92 0.92 0.67 0.61 0.27 0.65 0.82
CREA 0.95 0.95 0.75 0.61 0.23 0.50 0.78 0.87
BA 0.94 0.94 0.84 0.57 0.76 0.37 0.20 0.14 0.92

The square root of the average variance extracted values appears in italics. CR = composite reliability; AVE = average variance extracted; MSV = maximum shared squared variance; BEQ = brand equity; CONS = consumption; CONT = contribution; CREA = creation; BA = brand attitudes.

Table 4. Correlation Analysis Outcomes across the CEBSC Scale, Brand Equity, and Brand Attitudes
Correlated factors Estimate SE t-value
Consumption/Contribution 0.65 0.02 27.75
Consumption/Creation 0.50 0.03 19.18
Contribution/Creation 0.78 0.01 53.58
Consumption/Brand equity 0.40 0.04 9.09
Contribution/Brand equity 0.27 0.04 6.27
Creation/Brand equity 0.23 0.04 5.41
Consumption/Brand attitude 0.37 0.03 9.69
Contribution/Brand attitude 0.20 0.04 5.22
Creation/Brand attitude 0.14 0.04 3.59
Brand equity/Brand attitude 0.76 0.03 25.32

χ²(288) = 600.95, CFI = 0.96, TLI = 0.95, RMSEA = 0.05, SRMR = 0.06; p < 0.001; Estimator = robust maximum-likelihood; n = 416.

Table C. Descriptive Statistics, Factor Loadings and Explained Variance for the Items of the CEBSC Scale
Item Calibration λ R² M (SD) Validation λ R² M (SD) Full λ R² M (SD) Study 5 λ R² M (SD)
Consumption
Cons1 I read posts related to Brand X on social media. 0.83 0.68 3.79 (1.99) 0.82 0.68 3.89 (1.94) 0.83 0.68 3.84 (1.97) 0.87 0.75 3.72 (2.06)
Cons2 I read fan page(s) related to Brand X on social networking sites. 0.83 0.69 3.78 (2.06) 0.84 0.71 3.90 (2.05) 0.84 0.70 3.84 (2.05) 0.85 0.72 3.67 (2.14)
Cons3 I watch pictures/graphics related to Brand X. 0.64 0.41 4.22 (1.89) 0.66 0.43 4.34 (1.90) 0.66 0.44 4.28 (1.90) 0.77 0.60 3.87 (2.01)
Cons4 I follow blogs related to Brand X. 0.63 0.39 2.70 (1.88) 0.63 0.40 2.81 (1.90) 0.64 0.41 2.76 (1.90) 0.69 0.48 2.69 (1.97)
Cons5 I follow Brand X on social networking sites. 0.87 0.76 3.66 (2.04) 0.86 0.74 3.76 (1.97) 0.86 0.74 3.71 (2.01) 0.87 0.76 3.49 (2.05)
Contribution
Contr1 I comment on videos related to Brand X. 0.85 0.73 2.16 (1.63) 0.84 0.71 2.27 (1.72) 0.85 0.72 2.22 (1.68) 0.83 0.69 2.37 (1.83)
Contr2 I comment on posts related to Brand X. 0.87 0.76 2.35 (1.69) 0.90 0.80 2.43 (1.76) 0.88 0.78 2.39 (1.73) 0.90 0.81 2.51 (1.93)
Contr3 I comment on pictures/graphics related to Brand X. 0.87 0.75 2.17 (1.68) 0.86 0.74 2.26 (1.71) 0.87 0.75 2.22 (1.70) 0.86 0.74 2.42 (1.85)
Contr4 I share Brand X related posts. 0.89 0.79 2.43 (1.76) 0.88 0.78 2.52 (1.80) 0.89 0.79 2.47 (1.78) 0.90 0.80 2.59 (1.95)
Contr5 I “Like” pictures/graphics related to Brand X. 0.62 0.39 3.34 (2.00) 0.63 0.40 3.40 (2.02) 0.63 0.39 3.37 (2.01) 0.68 0.46 3.33 (2.17)
Contr6 I “Like” posts related to Brand X. 0.67 0.45 3.20 (1.98) 0.67 0.44 3.28 (1.99) 0.67 0.44 3.24 (1.98) 0.73 0.53 3.30 (2.10)
Creation
Creat1 I initiate posts related to Brand X on blogs. 0.89 0.78 1.94 (1.55) 0.90 0.78 1.95 (1.52) 0.89 0.80 1.95 (1.54) 0.91 0.82 2.21 (1.76)
Creat2 I initiate posts related to Brand X on social networking sites. 0.87 0.76 2.01 (1.58) 0.90 0.76 2.17 (1.70) 0.89 0.79 2.09 (1.64) 0.89 0.79 2.35 (1.83)
Creat3 I post pictures/graphics related to Brand X. 0.87 0.76 1.98 (1.54) 0.82 0.76 2.19 (1.67) 0.84 0.71 2.08 (1.61) 0.89 0.79 2.29 (1.80)
Creat4 I post videos that show Brand X. 0.83 0.69 1.96 (1.53) 0.85 0.69 2.11 (1.60) 0.84 0.71 2.03 (1.57) 0.86 0.73 2.27 (1.80)
Creat5 I write posts related to Brand X on forums. 0.80 0.65 1.96 (1.53) 0.80 0.65 2.11 (1.60) 0.80 0.64 2.04 (1.57) 0.85 0.72 2.29 (1.79)
Creat6 I write reviews related to Brand X. 0.75 0.57 1.91 (1.52) 0.69 0.57 2.04 (1.61) 0.72 0.52 1.97 (1.56) 0.82 0.68 2.16 (1.74)

Calibration sample (n = 1,126): χ²(115) 564.31, CFI = 0.95, TLI = 0.95, RMSEA = 0.05, SRMR = 0.06; Validation sample (n = 1,126): χ²(115) 557.47, CFI = 0.95, TLI = 0.94, RMSEA = 0.05, SRMR = 0.06; Full data set (n = 2,252): χ²(115) 719.47, CFI = 0.93, TLI = 0.92, RMSEA = 0.05, SRMR = 0.05; Study 5 (n = 416): χ²(313) = 651.71, CFI = 0.95, TLI = 0.95, RMSEA = 0.05, SRMR = 0.06; p < 0.001; Estimator = robust maximum-likelihood.

Table A. Activities Pertinent to Each Dimension of the “Consumer’s Online Brand-Related Activities” Framework
Activity Sources
Consumption
To download brand-related widgets/applications d, e
To follow a brand on social networking sites a, b, c, d
To follow brand-related blogs c, d, e
To listen to brand-related audio e, *
To play brand-related games d, e
To read brand-related emails c, ***
To read brand-related fan page(s) on social networking sites a, b, c, d
To read brand-related posts on social media a, b, c
To read brand-related reviews a, b, c, d, e, ***
To read other people’s comments about a brand on social media a, b, c, d, e, ***
To send brand-related virtual card e, *
To watch brand-related ads (e.g., banners, YouTube ads) d, ***
To watch brand-related pictures/graphics a, b, c, d, e
To watch brand-related videos b, c, e, ***
Contribution
To add brand-related videos to favorites c, d, ***
To click on brand-related ads d, ***
To comment on brand-related pictures/graphics a, b, c, d, e
To comment on brand-related posts c, d, e
To comment on brand-related videos a, b, c, d, e
To engage in brand-related conversations e, *
To forward brand-related emails to my friends/family c, **
To join a brand-related profile on social networking sites e, *
To “Like” brand-related fan pages a, b, c, d, ***
To “Like” brand-related pictures/graphics a, b, c, d
To “Like” brand-related posts b, c, d
To “Like” brand-related videos a, b, c, d, ***
To participate in online contests/drawings sponsored by a brand d, **
To rate brand-related products e, *
To share brand-related pictures/graphics a, b, c, d, ***
To share brand-related posts a, b, c, d
To share brand-related videos a, b, c, d, **
To take part in brand-related online events b, d, **
Creation
To create brand-related audio e, *
To create brand-related hashtags „#” on social networking sites c, ***
To create brand-related posts e, *
To initiate brand-related posts on blogs a, b, c, d, e
To initiate brand-related posts on social networking sites a, b, c, d
To post brand-related pictures/graphics a, b, c, e
To post brand-related videos b, c, d, e
To post pictures exposing self and a brand b, c, d, ***
To write brand-related posts on forums c, d
To write brand-related reviews c, d, e

a Activity detected during Study 1 (bulletin board – consumption); b Activity detected during Study 1 (bulletin board – creation); c Activity detected during Study 2 (in-depth interviews); d Activity detected during Study 3 (netnography); e Activity previously reported in literature; * Item not identified during the qualitative procedures; ** Item removed from the analysis during the exploratory factor analysis; *** Item removed from the analysis during the confirmatory factor analysis.

The question

Consumers engage with brands on social media in many ways, from reading a post to liking it to publishing their own. The Consumer’s Online Brand-Related Activities (COBRAs) framework grouped these into consumption, contribution and creation, but there was no validated instrument to measure them. We developed and validated one, and tested whether the three levels form a hierarchy in which people move from consuming to contributing to creating.

Implications

Practical implications

Managers can use the scale to audit and track social-media programmes, and each item points to a concrete activity they can encourage. Because engagement builds from consumption to contribution to creation, brands should keep producing content their audiences will consume, which in turn invites liking, commenting and sharing, and then user-generated content. The three dimensions are best used together; individual subscales suit studies focused on one activity, such as content creation.

Theoretical implications

The study turns the COBRAs framework into a validated measurement instrument and defines engagement with brand-related social-media content as a behavioural construct. It shows a hierarchy among the dimensions, with contribution fully mediating the link between consumption and creation, and supports nomological validity through positive correlations with brand equity and brand attitude. The scale can serve as a dependent variable for studying the drivers of engagement.

Limitations and future research

  • The list of brand-related activities (Appendix A) is not exhaustive; new activities will appear as platforms change, and the instrument should be updated accordingly.
  • The data were not controlled for past brand usage, which future studies could use as a moderator.
  • The research was conducted in one country (Poland); replications should test the scale’s equivalence across nations and cultures.
  • Future research could combine the scale with other behavioural variables in latent class analysis to identify engagement segments, and use multilevel models to compare industries and brands.

Questions this paper answers

How can I measure consumer engagement with brands on social media?

The 17-item CEBSC scale measures engagement with brand-related social-media content at three levels: consumption (5 items), contribution (6 items) and creation (6 items), using a 0–7 frequency scale (Schivinski, Christodoulides & Dabrowski, 2016).

What are COBRAs?

Consumers’ online brand-related activities: the ways people engage with brand-related content online, grouped into consuming, contributing and creating. Muntinga, Moorman and Smit (2011) introduced the framework, and Schivinski, Christodoulides and Dabrowski (2016) developed the scale that measures it.

Is the CEBSC scale reliable?

Yes. Composite reliabilities were 0.88 to 0.93 in the validation sample and Cronbach’s alpha ranged from 0.88 to 0.94 in a second study of 416 consumers (Schivinski, Christodoulides & Dabrowski, 2016).

Do consumers move from consuming brand content to creating it?

The evidence supports a hierarchy: consumption predicts contribution (β = 0.61), contribution predicts creation (β = 0.81), and contribution fully mediates the link between consumption and creation (Schivinski, Christodoulides & Dabrowski, 2016).

Key concepts

Consumer’s online brand-related activities (COBRAs)
A set of brand-related online activities on the part of the consumer that vary in the degree to which the consumer interacts with social media and engages in the consumption, contribution, and creation of media content (definition from the article).
Consumption
The minimum level of engagement: reading, watching and following brand-related content, whether created by firms or by users.
Contribution
Peer-to-peer and peer-to-content interactions about brands, such as liking, commenting and sharing, without creating new content.
Creation
The strongest level of engagement: producing and publishing brand-related content, which can in turn be consumed and contributed to by others.

Open science and materials

How to cite

Schivinski, B., Christodoulides, G., & Dabrowski, D. (2016). Measuring consumers’ engagement with brand-related social-media content: Development and validation of a scale that identifies levels of social-media engagement with brands. Journal of Advertising Research, 56(1), 64–80. https://doi.org/10.2501/JAR-2016-004

Related publications

All publications