Factors driving customer engagement with the global beauty brands via social media marketing

Diagram 4 - Number of Interactions Per Page

Source: Chart created using the data from Socialbakers Analytics,

URL: https://suite.socialbakers.com/274788/analytics/fb/multicompare

Instagram

Fans Overview

Table 6: The Total Number of a Page's Followers

Brand	Number of Total Followers
M.A.C	17 961 776
NYX	12 667 928
Too Faced	10 332 521

Source: Chart created using the data from Socialbakers Analytics,

URL: https://suite.socialbakers.com/274788/analytics/ig/multicompare

As well as on Facebook, on Instagram M.A.C Cosmetics stays the leader in terms of the number of total followers - 17 961 776 people, whereas this indicator is over 5 million fans lower on NYX page and over 7 million people lower than on Too Faced page. At the same time, the numbers of absolute growth of followers beg to differ. Although M.A.C gained more followers (145 582) than the other brands, Too Faced and NYX showed a relatively good increase in followers - 114 652 and 107 540 respectively.

Diagram 5 - Absolute Growth of Total Followers on Each Page

Source: Chart created using the data from Socialbakers Analytics,

URL: https://suite.socialbakers.com/274788/analytics/ig/multicompare

Content overview.

Again, as on Facebook, M.A.C stays the leader in the number of profile posts - 158. But it should be noted that unlike Facebook, in this case NYX get the second place publishing 131 posts, which means that the brand pays more attention to the Instagram platform. Too Faced doesn't go far behind, publishing 130 posts during the month.

Diagram 6 - Number of All Posts Published By Each Page

Source: Chart created using the data from Socialbakers Analytics

URL: https://suite.socialbakers.com/274788/analytics/ig/multicompare

Engagement overview. Too Faced confirmed its status as one of the major engaging make-up brand on social networks achieving 909,17 Interactions per 1k followers, whereas NYX took second place reaching the level of 469,05 and finally, the third place went to M.A.C with the 395,51 interactions.

Diagram 7 - Number of Interactions per 1000 Followers Per Page

Source: Chart created using the data from Socialbakers Analytics,

URL: https://suite.socialbakers.com/274788/analytics/ig/multicompare

In terms of total number of interactions, Too Faced outruns the other brands again with 9 240 696 Likes and 95 283 Comments, 9 335 979 total. Due to the large number of followers on Instagram M.A.C also had good with total 7 075 564 interactions, 31 722 comments and 7 043 842 likes. And finally NYX received 5 872 201 likes and 44 428 comments, making up 5 916 629 total.

Diagram 8 - Number of Interactions (Comments and Likes) Per Page

Source: Chart created using the data from Socialbakers Analytics,

URL: https://suite.socialbakers.com/274788/analytics/ig/multicompare

2.3 Analysis of international beauty brands content marketing on instagram and facebook

The previous subchapter allowed us to identify the most engaging global makeup brand on social media platforms which appears to be Too Faced Cosmetics. On Instagram the brand managed to achieve almost two times higher number of interactions per 1000 followers than NYX and more than two times higher number than M.A.C did, although the latter had 7 million followers more on its brand page and generally is considered to be more popular, with higher brand awareness.

Even more significant results Too Faced demonstrated on its Facebook page, where its number of interactions per 1000 fans reached 190, which is almost 12 times higher than NYX indicator and more than 17 times higher in comparison with M.A.C number.

Regarding the M.A.C Cosmetics case, the results indicate that a bigger amount of followers do not entail a higher online engagement rate. This finding allows us to believe that Too Faced brand represents a most successful social media strategy in its online communities, so we find it logical to make a deeper research on Too Faced content marketing within Facebook and Instagram media platforms and compare the results with the ones of the other two brands.

Research Methodology

When analyzing the case studies, Too Faced official pages on Facebook and Instagram will be monitored and used as a benchmark when comparing its results with other brands. Due to limitations in the volume of the current master's thesis, the content of the posts on Too Faced Facebook page will be compared only with the NYX Facebook page, whilst on Instagram platform Too Faced will be analyzed together with M.A.C Cosmetics page.

Five major engagement factors will be tested in order to disclose the correlation with online customer engagement: which post media types better appeal to the audience and encourages more consumers' replies, what kind of post content motivates users to interact, on which day a post should be published to engage more customers, if the posts should be interactive to strengthen CE and if post should include mentions of influencers. All of the mentioned determinants are viewed by brands as the key impactful components for customer engagement and are bore in mind when creating an engaging content marketing strategy.

Data Collection Method

The research will be conducted in the form of desk research, which involves the analysis of secondary data that has been taken from such open sources as the brands pages on Instagram, Facebook and SocialBakers Analytics. This method was selected owing to the fact that it is not that time consuming, but convenient and cost effective.

Data Analysis Method

The collected data on the brands Instagram and Facebook pages will be analyzed in terms of post media type, its content type, posting day, interactivity level and influencer integration the results of which will be presented in the forms of diagrams for a greater visibility.

On the basis of the literature review and previous studies on the topic mentioned in the first chapter of this paper, the following hypotheses regarding the enhancement of social media engagement with global make-up brands were formulated:

Post Content Type

- H1: Instagram posts including faces have higher influence on customer engagement in comparison with posts without faces.

- H2: Facebook posts including faces have higher influence on customer engagement in comparison with posts without faces.

Post Media Type

- H3: Posts with images reach the highest rate of consumer engagement both on Facebook and Instagram.

- H4: Posts with videos accumulate a higher level of customer engagement than link posts (applicable only for Facebook pages).

Posting day

- H5: Social media posts created on workdays generate a higher level of customer engagement than those published on weekends.

Interactivity Level

- H6: Posts with a high level of interactivity lead to a higher customer engagement in comparison with low interactivity posts.

Influencer Integration

- H7: Posts mentioning influencers drive a higher consumer engagement than those which do not mention them.

Data presentation and Analysis

Facebook Platform

Too Faced

By 31 March 2018 Too Faced had 2,9 million fans on Facebook. During the month the brand published 73 posts on its Facebook official page, among them 30 posts with the highest number of interactions were scrutinized during the research.

First, we have looked on the distribution of shares among each post content type: according to Diagram 9, product-based posts are absolute leaders with a 97% share, posts including brand products and faces gained as little as 3% (published only once), whereas posts featuring just faces have not been noticed at all.

As a result, the highest level of engagement was fixed among product-based posts - avr. 3.31, total 95.92, while posts containing faces and products reached only 1,17 rate total. Nevertheless, the hypothesis H2 cannot be tested due to solid disparity in the post content type distribution.

Diagram 9 - Post Content Type Distribution on Too Faced Facebook Page

Source: composed by the author on the basis of analysis of collected data

From Diagram 10, we see that the most widely used post media type by Facebook is images (97%), where all of them had a product-based nature. The main distinctive feature of all Too Faced photos is their stylish appeal in nude and pastel colours, which immediately attract attention of the page's fans.

According to the content, the majority of posts demonstrate photos promoting existing or new brand products. The photo that could drive customers to engage with the post in most effective way was one connected with the launch of new “It Just Comes Naturally” Collection. The post received 93 255 reactions, 1 604 shares and 1117 comments from the followers, who expressed their feelings about this news. Only in this video brand products were shown together with faces. The Number of Interaction per 1000 fans reached the highest level among all thirty posts - 32,63 (Table 7).

Within the analyzed posts a video was contained in only one publication (3%). It reach a moderate level of Interaction per 1000 fans (1,17) and included Too Faced Co-Founder and Chief Creative Officer, Jerrod Blandino, who shared his inspiration behind their new make-up collection. This post received was shared by 180 users and received over 60 thousands views. Links were not used by the Too Faced moderator at all as they cannot stimulate followers to interact with the brand.

Overall, the received results have partially confirmed hypothesis H3 in virtue of imparity of post media type distribution, although images have generated the highest level of interactions among followers - avr. 3.31, total 95.92. The videos strongly lag in quantity, as their engagement indicator is significantly lower than images have - only 1,17 total. As for links, in terms of engagement they had no value for the page moderators, as they were not represented at all, so hypothesis H4 cannot be tested.

Diagram 10 - Post Media Type Distribution on Too Faced Facebook Page

Source: composed by the author on the basis of analysis of collected data

Regarding Diagram 11, the majority of content has been provided on Workday (19 posts - 63%), when consumers are most likely to be active on social media. On the weekend the page generates only 37% of the content.

On workdays, 18 photos and 1 video were published, where 95% of content included brand products and 5% included brand product and faces. Total number of interactions achieved on workdays is 251 106, which consists of reactions, comments and shares, while average Number of Interactions per 1000 Fans was 4,48.

On weekends, the absolute majority of post media types refer to images - 11 posts, thus including only product-base publications. Total number of interactions reached 35 176 (-85, 9% vs. Workday), average Number of Interactions per 1000 Fans - 1,08. These results confirm our suggestion that posts published on workdays generate a higher level of customer engagement, thus hypothesis H5 has been approved.

Diagram 11 - Posting Day Distribution on Too Faced Facebook Page

Source: composed by the author on the basis of analysis of collected data

According to Diagram 12, 100% of the analyzed posts had a high level of interactivity, achieving 97,09 in number of interactions per 1000 fans in total, with avr. - 3.24. The posts high interactivity is expressed in the form of Call to Actions in the end of each publications such as “shop now” or “check it out” offering the fans to move to the brand website. The result cannot fully approve hypothesis H6, as no posts with low interactivity were found and compared, but the fact that Too Faced Facebook page, which gained the highest consumer engagement level among the three beauty brands, seems to consider post interactivity level as one of the key factors for social media successful performance, we believe that our suggestion can be still partially approved.

Diagram 12 - Interactivity Level Distribution on Too Faced Facebook Page

Source: composed by the author on the basis of analysis of collected data

With regards to Diagram 13, the majority of the analyzed content lacked mentions of influencers (97%), as only one post (3%) integrated an influencer.

That post included Too Faced Co-Founder Jerrod Blandino, who shared his ideas on applying makeup from Too Faced new launched collection in a video and achieved a 1,17 level of Interaction per 1000 fans which is 81,5% lower than posts without influencers' presence. Nonetheless, hypothesis H7 cannot be tested due to heavy imparity in influencer integration distribution.

Diagram 13 - Influencer Integration Distribution on Too Faced Facebook Page

Source: composed by the author on the basis of analysis of collected data

NYX Professional Makeup

By 31 March 2018, NYX had 2 020 062 fans on Facebook. During the month the brand published 47 posts on its Facebook page, among them 30 posts with the highest number of interactions were scrutinized during the research.

First, we looked on the distribution of shares among each post content type: according to Diagram 14, product-based posts prevail with a 64% share, and are followed by posts with faces - 23%, the minority of the posts include both brand products and faces which is a 13% share.

Overall, the highest level of engagement was noticed among product-based posts - avr. 0,51 and 9,70 total, while face-based post reached 2,50 total and avr. -0,36. As for posts containing both faces and brand products, the engagement level equaled 1,59 total and avr. - 0,40.

Thus, the hypothesis H2 was disapproved, as NYX demonstrates that both brands and followers are more interested in product-based content, which gathers the highest level of interactions.

Diagram 14 - Post Content Type Distribution on NYX Facebook Page

Source: composed by the author on the basis of analysis of collected data

Unlike Too Faced Facebook page, NYX does not publish content only in the format of images, as it posts videos more willingly and even uses Links, though the share of photos is higher than other media types anyway - 64% (Diagram 15). According to the NYX Facebook page content, the majority of posts demonstrate photos promoting existing or new brand products, which is a similar strategy used by Too Faced.

As for the video posts, their share reached 10% of the analyzed content. They mostly included brand products, but one of the videos represented a brand product together with a model's face, overall generating 1.94 of total number of Interactions per 1000 fans 1,93 and avr. 0,64. Although images were more widely used by Too Faced (83%), they generated a lower number of average interactions - 0,45, total 11,3. Links to the brand website were used by the NYX Facebook page moderators only two times (7%), thus accumulating a low engagement level - 0,72 total, avr. - 0,36.

The received results cannot confirm hypothesis H3, suggestion H4 has been approved.

Diagram 15 - Post Media Type Distribution on NYX Facebook Page

Source: composed by the author on the basis of analysis of collected data

During the month, post day distribution represented on Diagram 16 had a similar structure as on Too Faced Facebook page: 23% of posts were published on weekends, whereas the majority of content was provided on workdays (77%). As a result, workday publications engaged fans better reaching a higher engagement level - 11,09 total and avr. 0,48, while those posts provided on weekends generated only 2,70 total and avr. 0,39. Thus, the hypothesis H5 has been approved.

Diagram 16 - Post Day Distribution on NYX Facebook Page

Source: composed by the author on the basis of analysis of collected data

According to Diagram 17, 93% of the analyzed posts had a high level of interactivity, achieving 12,59 in total number of interactions per 1000 fans, with avr. - 0,45. The posts high interactivity is expressed not only in the form of Call to Actions inviting users to buy the brand product, but also in the form of contests and reminders of exclusive sales causing followers to make an urgent purchase. The share of low interactivity posts is only 7%, which accumulates 1,20 of total interactions per 1000 fans, avr. - 0,60. As a result, hypothesis H6 has not been approved.

Diagram 17 - Post Interactivity Level Distribution on NYX Facebook Page

Source: composed by the author on the basis of analysis of collected data

With regards to Diagram 18, unlike on Too Faced Facebook page, the majority of the analyzed NYX content mentioned influencers (57%). In 65% of the cases the posts integrated with influencers via featuring their faces (Table 8). Overall, such posts generate 6,21 of interactions in total, avr. - 0.37, while publications without influencer integration generated 7.57, avr. - 0,58. Thus, hypothesis H7 has not been confirmed.

Diagram 18 - Influencer Integration Distribution on NYX Facebook Page

Source: composed by the author on the basis of analysis of collected data

Instagram Platform

Too Faced

By 31 March 2018 Too Faced had 10,3 million followers on Instagram. During the month the brand published 130 posts on its official page, among them 30 posts with the highest number of interactions were analyzed during the research.

First, we have looked on the distribution of shares among each post content type: according to Diagram 19, product-based posts are lead the way with a 83% share, posts including faces gained 14% , whereas posts featuring both brand products and faces accumulated only 3%.

As a result, the highest level of engagement was noticed among product-based posts - avr. 12.96, total 323.94, while posts containing faces reached 45,10 total, avr. - 11.28 and post including both brand products and a face generated a total of 9,92. Thus, hypothesis H1 has not been confirmed.

Diagram 19 - Post Content Type Distribution on Too Faced Instagram Page

Source: composed by the author on the basis of analysis of collected data

From Diagram 20, that like on Too Faced Facebook page the most widely used post media type on its Instagram page is images (97%), where 83% of them had a product-based nature, 14% were face-based posts 3% included both products and faces. It has been noted during the research that almost all of the images on Too Faced Instagram page correspond with the ones on Facebook, but the photo that drove followers to engage with the post the most was different - the post included a give-away or a contest of four Too Faced products. It received 301 235 likes and 39 329 comments from the followers, who were excited about the news and wanted to participate. The Number of Interaction per 1000 fans reached the highest level among all thirty posts - 33,15 (Table 9). A video again was contained in only one publication (3%), featuring a new glitter mask. But it reached not so high level of Interaction per 1000 fans (9,32).

Despite the fact, that images have generated the highest level of interactions among followers - avr. 12.75, total 369.64, hypothesis H3, again, can only be partially confirmed in virtue of disparity in the posts distribution.

Diagram 20 - Post Media Type Distribution on Too Faced Instagram Page

Source: composed by the author on the basis of analysis of collected data

Regarding Diagram 21, it can be seen that the majority of content has been provided on Workdays (80%), so the brand uses the same tactics on its Facebook and Instagram pages. On weekends the page generated only 20% of the content.

As a result, workday publications engaged followers in a better way reaching a higher engagement level - 313,12 total and avr. 13.05, while those posts provided on weekends generated only 65.84 total and avr. 10,97. Thus, the hypothesis H5 has been also approved on Too Faced Instagram page.

Diagram 21 - Post Day Distribution on Too Faced Instagram Page

Source: composed by the author on the basis of analysis of collected data

According to Diagram 22, distribution of posts with different interactivity levels divided into halves. 50% of the analyzed posts with a high level of interactivity achieved 205,20 in number of interactions per 1000 fans in total, with avr. - 13.68, while the other 50% with a low interactivity level earned a lower engagement: 173,76 total, avr. - 11.58. The posts high interactivity is expressed not only in the form of Call to Actions inviting followers to buy a brand products, but also in the form of online contests and giveaways, which not noticed on Too Faced Facebook page. Overall, the result of the analysis can approve hypothesis H6.

Diagram 22 - Post Interactivity Level Distribution on Instagram Page

Source: composed by the author on the basis of analysis of collected data

With regards to Diagram 23, the majority of the analyzed content tended not to mention influencers (73%), whereas posts which integrated an influencer got a 27% share.

Posts mentioning influencers achieved a 94,14 level of Interaction per 1000 fans, avr. - 11.77, whereas posts not mentioning them gained 284,82 total, avr. - 12,95. Thus, hypothesis H7 has not been approved.

Diagram 23 - Influencer Integration Distribution on Instagram Page

Source: composed by the author on the basis of analysis of collected data

M.A.C Cosmetics

By 31 March 2018, M.A.C Cosmetics had 17,9 million followers on Instagram. During the month the brand published 158 posts on its official page, among them 30 posts with the highest number of interactions were analyzed during the research.

First, we looked on the distribution of shares among each post content type: according to Diagram 24, product-based posts prevail with a 57% share, and are followed by posts with faces - 23%, the minority of the posts include both brand products and faces which is a 20% share.

Overall, the highest level of engagement was noticed again among product-based posts - avr. 4,87 and 82,81 total, while face-based post reached 38.86 total and avr. - 5,55. As for posts containing both faces and brand products, the engagement level equaled 36,42 total and avr. - 6,07.

Thus, the hypothesis H1 can be approved, as according to our analysis, the engagement level of posts with faces is higher than those containing only brand products, but lower than of posts including both products and faces.

Diagram 24 - Post Content Type Distribution on M.A.C Instagram Page

Source: composed by the author on the basis of analysis of collected data

In comparison with other beauty brands under analysis, M.A.C represents only posts in the format of images. The majority of the published images are photos of close-up faces and brand products. All the scrutinized posts managed to accumulate a quite high number of interactions per 1000 followers - 158,09 total, avr. - 5,26.

The received results do not allow to test hypothesis H3 due to the fact that no video posts were published, thus could not be compared and properly analyzed.

Diagram 25 - Post Media Type Distribution on M.A.C Instagram Page

Source: composed by the author on the basis of analysis of collected data

During the month, post day distribution represented on Diagram 26 differed from that of the other brands: 53% of posts were published on weekends, whereas only 47% were provided on workdays. But after analyzing the main indicators, we can see that workday publications still engage users in a better way reaching a higher engagement level - 81,23 total and avr. 5,80, while those posts provided on weekends generated only 76,86 total and avr. 4,80. Thus, the hypothesis H5 has been again approved.

Diagram 26 - Post Day Distribution on M.A.C Instagram Page

Source: composed by the author on the basis of analysis of collected data

According to Diagram 27, only 40% of the analyzed posts had a high level of interactivity, achieving 60,09 in total number of interactions per 1000 fans, with avr. - 5. The posts high interactivity is mostly expressed in form of Call to Actions inviting users to try the brand product or asking open-ended questions about them. While the share of low interactivity posts is 60%, it accumulates 98 of total interactions per 1000 fans, avr. - 5,4. As a result, hypothesis H6 has not been approved.

Diagram 27 - Post Interactivity Level Distribution on M.A.C Instagram Page

Source: composed by the author on the basis of analysis of collected data

With regards to Diagram 28, the majority of the analyzed M.A.C page content mentioned influencers (83%). The posts include not only the photos of the influencers and their makeup made by M.A.C representatives, but also pictures of M.A.C products made by the influencers. Overall, such posts generate 128,64 of interactions in total, avr. - 5,14, while publications without influencer integration generated 29,45, avr. - 5,89. Thus, hypothesis H7 has not been confirmed.

Diagram 28 - Influencer Integration Distribution on M.A.C Instagram Page

Source: composed by the author on the basis of analysis of collected data

Table 7: Top-5 Most Popular Posts of Too Faced on Facebook

#	Number of Interactions per 1000 Followers	Total Number of Interactions	Reactions	Comments	Shares	Number of Fans	Content Type	Post Media Type	Post Day	Interactivity	Influencer Integration
1	32,63	95 976	93 255	1 117	1 604	2 941 672	Product	Image	Workday	High	No
2	20,22	60 087	58 428	831	828	2 972 222	Product	Image	Workday	High	No
3	11,97	35 010	33 730	466	814	2 925 119	Product	Image	Workday	High	No
4	2,86	8 422	7 552	656	214	2 941 672	Product	Image	Workday	High	No
5	2,77	8 228	7 416	568	244	2 973 226	Product	Image	Weekend	High	No

Table 8: Top-5 Most Popular Posts of NYX on Facebook

#	Number of Interactions per 1000 Followers	Total Number of Interactions	Reactions	Comments	Shares	Number of Fans	Content Type	Post Media Type	Post Day	Interactivity	Influencer Integration
1	1,37	2 751	2 444	216	91	2 004 643	Product	Image	Workday	High	No
2	0,86	1 739	1 614	60	65	2 011 392	Product	Image	Workday	Low	No
3	0,74	1 489	1 215	217	57	2 006 173	Product	Video	Workday	High	No
4	0,70	1 402	1 244	75	83	2 014 091	Face	Image	Workday	High	Yes
5	0,66	1 324	1 145	119	60	2 013 479	Product	Image	Workday	High	No



Table 9: Top-5 Most Popular Posts of Too Faced on Instagram

#	Number of Interactions per 1000 Followers	Total Number of Interactions	Likes	Comments	Number of Followers	Content Type	Post Media Type	Post Day	Interactivity	Influencer Integration
1	33,15	340 564	301 235	39 329	10 277 082	Product	Image	Workday	High	No
2	26,83	275 390	271 940	3 450	10 265 321	Product	Image	Workday	High	No
3	15,84	162 504	162 183	321	10 261 455	Face	Image	Workday	High	No
4	15,30	157 810	156 683	1 127	10 317 052	Product	Image	Workday	Low	Yes
5	14,95	154 079	152 807	1 272	10 310 438	Product	Image	Workday	High	No

Table 10: Top-5 Most Popular Posts of M.A.C Cosmetics on Instagram

#	Number of Interactions per 1000 Followers	Total Number of Interactions	Likes	Comments	Number of Followers	Content Type	Post Media Type	Post Day	Interactivity	Influencer Integration
1	9,84	175 847	174 305	1542	17867829	Product & Face	Image	Workday	Low	Yes
2	9,27	166 395	165 430	965	17961776	Product	Image	Weekend	Low	No
3	8,35	149 875	149 064	811	17953425	Product & Face	Image	Workday	Low	Yes
4	8,27	147 702	147 073	629	17863516	Product	Image	Weekend	High	No
5	8,12	145 850	145 052	798	17951267	Face	Image	Workday	High	Yes

3. Findings and recommendations

Hypotheses testing: cross case analysis

All the hypotheses provided by scholars and scrutinized in the first and second chapters of the current research will be summarized and review below.

Post Content Type

- Hypothesis H1 has been not confirmed by Too Faced brand, but has been confirmed in case of M.A.C Cosmetics.

- Hypothesis H2 could not be properly tested on the example of Too Faced brand, although it was not confirmed by NYX.

Post Media Type

- Hypothesis H3 has been partially confirmed by Too Faced, while it has not been confirmed on the instance of NYX and could not be tested in the case of M.A.C Cosmetics.

- Hypothesis H4 has been confirmed by the example of NYX brand, although it could not be tested in the case of Too Faced.

Posting day

- Hypothesis H5 has been fully confirmed by all of the three brands.

Interactivity Level

- Hypothesis H6 appeared to be the most contradictory one: It has not been confirmed neither by NYX, nor by M.A.C, but it has been confirmed in case of Too Faced with its Instagram page and partially confirmed by Too Faced Facebook page.

Influencer Integration

- Hypothesis H7 has not been confirmed by all of the three brands cases.

The summary of the above-mentioned hypotheses is presented in the Table 11.



Table 11: The results of testing hypotheses (Composed by the author on the basis of tested hypotheses)

Factor	Hypothesis	Brand	Result
Post Content Type	Instagram posts including faces have higher influence on customer engagement in comparison with posts without faces.	Too Faced	Not confirmed
		M.A.C	Confirmed
	Facebook posts including faces have higher influence on customer engagement in comparison with posts without faces.	Too Faced	Could not be tested
		NYX	Not confirmed
Post Media Type	Posts with images reach the highest rate of consumer engagement both on Facebook and Instagram.	Too Faced	Partially Confirmed
		NYX	Not confirmed
		M.A.C	Could not be tested
	Posts with videos accumulate a higher level of customer engagement than link posts.	Too Faced	Could not be tested
		NYX	Confirmed
Posting day	Social media posts created on workdays generate a higher level of customer engagement than those published on weekends.	Too Faced	Confirmed
		NYX	Confirmed
		M.A.C	Confirmed
Interactivity Level	Posts with a high level of interactivity lead to a higher customer engagement in comparison with low interactivity posts.	Too Faced	- Partially Confirmed on Facebook - Confirmed on Instagram
		NYX	Not confirmed
		M.A.C	Not confirmed
Influencer Integration	Posts mentioning influencers drive a higher consumer engagement than those which do not mention them.	Too Faced	- Could not be tested on Facebook - Not Confirmed on Instagram

On the strength of the data mentioned in Table 11, it can be stated that hypothesis H5 is the only one which has been fully confirmed by all of the three cases: followers of the beauty brands unanimously showed more engagement on workdays, than on weekends. One more unanimous result in terms of customer engagement enhancement on social media pages was presented in influencer integration: the analysis showed that posts mentioning influencers gained a lower average number of interactions per 1000 followers, than those that did not, thus we can state that hypothesis H7 has not been confirmed. Hypothesis H4 has been proved by NYX, but could not be properly tested by Too Faced. Nevertheless, we believe that despite the fact that Too Faced does not approve this hypothesis in numbers, on practice it can be considered as approved, because Too Faced content marketing managers most likely do not use links on Facebook as they are aware that these media posts do not engage their followers. Concerning hypothesis H2, the case of NYX has fully refuted it, and the fact that the most engaging of the three brands such as Too Faced did not tend to post images with faces on their Facebook page only proves the rebuttal of the hypothesis, thus it will be considered as not confirmed. Posts depicting faces did not drive a higher level of engagement within Too Faced Instagram page, but they did in case of M.A.C Cosmetics Instagram page, where posts with faces achieved 5,55 of average number of interactions per 1000 followers (vs. avr. 4,87 of the product-based posts), thus we will consider this hypothesis partially confirmed. Overall, hypothesis H3 will also be regarded as partially confirmed, as the fact that such engaging brand as Too Faced and such popular and most-followed beauty brand as M.A.C publish videos on their pages very rarely which might imply that they do not gain a high engagement rate. Finally, we will view hypothesis H6 as partially confirmed one, as Too Faced brand, which was considered the most successful one in terms of engaging its followers, tends to always include interactive texts in their posts, thus generating a higher level of customer engagement.

As the result of this analysis, two hypotheses (№4 and №5) have been totally confirmed; three hypotheses (№1, №3, №6) have been only partially confirmed, while hypotheses №2 and №7 have not been confirmed. The summary of results can be observed in the Table 12.



Table 12: Major findings of the research (Composed by the author on the basis of tested hypotheses)

Hypothesis	Result
H1: Instagram posts including faces have higher influence on customer engagement in comparison with posts without faces.	Partially Confirmed
H2: Facebook posts including faces have higher influence on customer engagement in comparison with posts without faces.	Not confirmed
H3: Posts with images reach the highest rate of consumer engagement both on Facebook and Instagram.	Partially Confirmed
H4: Posts with videos accumulate a higher level of customer engagement than link posts.	Confirmed
H5: Social media posts created on workdays generate a higher level of customer engagement than those published on weekends.	Confirmed
H6: Posts with a high level of interactivity lead to a higher customer engagement in comparison with low interactivity posts.	Partially Confirmed
H7: Posts mentioning influencers drive a higher consumer engagement than those which do not mention them.	Not confirmed

The main goal of this thesis was to identify the key tools that enhance customer engagement with global beauty brands on their social media platforms. Based on the previous scientific literature on the topic, five factors have been identified in the first chapter of this paper (post content type, post media type, posting day, the interactivity level of a post and integration of influencers on the created publications) and divided into several sub-factors. In the second chapter, these sub-factors were analyzed more deeply within the global beauty industry.

As a result, the following points were detected:

- Only four factors out of five turned out to be credible (influencer integration appeared to be not relevant);

- Three sub-factors are considered to drive online customer engagement: posts with a high interactivity level, posts including images and posts depicting faces (the latter is only applicable for Instagram);

- One sub-factor turned out to be significantly important (posting on workdays).

Based on the received results, key factors driving customer engagement with global beauty brands on their social media pages were defined in the current chapter (Figure 2).

Figure 2 - Factors driving customer engagement with global beauty brands via social media marketing

(Created by the author based on the conducted analysis)

Based on the findings mentioned above and relevant information found during the master's thesis preparation the following recommendations for global beauty brands can be proposed in terms of enhancing customer engagement on their social media pages:

1. Focus more on Instagram, than on Facebook, as the first one generates a much higher number of interactions;

2. During a post strategy planning take into consideration that the highest level of customer engagement is achieved on workdays;

3. Publish more posts including images, but at the same time not to ignore posts with videos;

4. Not to waste time on creating link posts on Facebook pages, as they drive a very low level of customer engagement;

5. Try to make posts more interactive by asking followers various questions, using different call to actions to encourage users to buy, try or just learn more about a brand products, making contests and giveaways and also by reminding people of special offers or upcoming brand events and activities;

6. Publish more posts depicting faces on Instagram;

7. Provide followers with more information on the brand products by publishing posts depicting existing or newly launched brand products.

Limitations and propositions for further research on the topic

First of all, the current master's thesis is based on a narrow sample of cases - only three beauty brands were analyzed - and a narrow sample of the brands posts that were scrutinized, thus the received results might not be absolutely appropriate to a bigger number of similar brands. To better validate the current analysis it might be useful to conduct a quantitative research, such as a survey to understand which factors the actual followers of beauty pages consider the most important.

Secondly, there may be other factors that have a crucial impact on social media customer engagement, but are not still discovered by scientists and are not presented in the present paper.

Subsequently, the driving factors of social media customer engagement with global beauty brands may differ in terms of beauty categories: a separate analysis could be conducted on such beauty categories as fragrance or skincare.

Moreover, the research on content strategy of the global beauty pages has taken place only within two social networks - Facebook and Instagram, thus the future studies may cover other social media platforms like YouTube, Twitter or Snapchat. Finally, it might be interesting to compare the content of global beauty pages on social media with local or regional pages of the same brand.



Conclusion

The current master's thesis focused on analyzing determinants for a successful content marketing strategy which could drive customer engagement with global make-up brands on their social media pages. To achieve this goal, customer engagement has been researched by discovering the concept and various definitions of the term, distinguishing online engagement as well as factors enhancing the level of this type of consumer engagement which were defined in the previous studies made on the topic. Subsequently, the focus shifted towards customer engagement on social media platforms and its peculiarities within beauty industry. The global make-up brands on Facebook and Instagram were observed more deeply and the factors driving customer engagement on their social media pages were combined on the basis of the data collected from scientific articles.

The theoretical foundation from the first chapter allowed us to formulate hypotheses regarding the determinants enhancing social media engagement with world beauty brands, which were tested in the second part of the current master's thesis. The case study research method was selected by dint of which three global makeup brands were analyzed. As a result of the conducted analysis, two hypotheses were fully confirmed, three hypotheses were partially confirmed, and the other two hypotheses were not confirmed.

The outcomes of the second chapter of the current master's thesis then guided us towards the instruments which drove a high level of consumer engagement with global beauty brands on their social media pages. These factors include posting during workdays, high level post interactivity, posts depicting faces and publications with images.

Naturally, every research paper has its limitations and in our case it is limited by the makeup beauty category, the number of beauty brands and the number of posts that have been analyzed, the amount of factors that have been explored and number of social media platforms. Nevertheless, it can be used as a motive for conducting further analyses, which is especially crucial for such undiscovered phenomenon as the customer engagement.

The aim of the current master's thesis has been fully achieved, confirming that customer engagement is a highly valuable measure of success of brands content marketing on social media platforms and that it is impossible to generate a high level of consumer engagement without understanding the triggers that make customers interact with the brand.

Finally, during the research we got deep comprehension of which types of posts appeal more to consumers within global beauty brands on its Facebook and Instagram pages. The outcomes of the study will be valuable for marketers in the developing a content strategy on social media and driving consumer engagement in the beauty industry.



References

1. Ardley, S. & Brooke, G. (2014). Social Media Analytics - https://www.ox.ac.uk/public-affairs/social-media

2. Bakhshi, S., Shamma, D. A., & Gilbert, E. (2014). Faces engage us: Photos with faces attract more likes and comments on Instagram (pp. 965-947). In Proceedings of the 32nd annual ACM conference on Human factors in computing systems: ACM.

3. Beauty Brands on Social Media - Statistics & Facts. (n.d.). Retrieved on 14.02.2018 from http://www.statista.com/topics/2381/beauty-brands-on-social -media/

4. Bowden, L.XH., (2009). The process of customer engagement: a conceptual framework, Journal of marketing theory and practice , Winter. pp.63-74.

5. Broderick, J. (2016). Engaging customers during a website visit: a model of website customer engagement , International Journal of Retail & Distribution Management, Vol. 44 Iss 8 pp. 814 - 839

6. Brodie, R.J., Hollebeek, L.D., Juric, B. & Llic, A., (2011). Customer engagement: conceptual domain, fundamental propositions, and implications for research, Journal of service research; Vol. 14 (3), pp. 252-271

7. Brodie, R.J., Ilic, A., Juric, B. and Hollebeek L. (2013). Consumer Engagement in a Virtual Brand Community: An Exploratory Analysis, Journal of Business Research 66, No. 8, pp.105-114.

8. Brown, J., Broderick, A.J. & Lee, N. (2007). Word of mouth communication within online communities: conceptualizing the online social network, Journal of interactive marketing , pp.2-20.

9. Carlson, J., Rahman, M., Voola, R., De Vries, N. (2018). Customer engagement behaviours in social media: capturing innovation opportunities , Journal of Services Marketing, Vol. 32 Issue: 1, pp.83-94,

10. Chaffey, D. (2018). Global Social Media Statistics Summary 2017. Retrieved on 14.02.2018 from http://www.smartinsights.com/

11. Chauhan, K., & Pillai, A. (2013). Role of content strategy in social media brand communities: a case of higher education institutes in India. The Journal of Product and Brand Management, 22(1), 40-51.

12. Christy, A., Tuten, T. (2015). Creative Strategies in Social Media Marketing: An Exploratory Study of Branded Social Content and Consumer Engagement, Psychology & Marketing, 32 (1), 15-27.

13. Chu, S., Kim, Y. (2015), Determinants of consumer engagement in electronic word-of-mouth (eWOM) in social networking sites. Retrieved on 14.02.2018 from https://doi.org/10.2501/IJA-30-1-047-075

14. Cvijikj, I. P., & Michahelles, F. (2013). Online engagement factors on Facebook brand pages. Social Network Analysis and Mining, 3(4), 843-861.

15. De Vries, L., Gensler, S., & Leeflang, P. S. H. (2012). Popularity of Brand Posts on Brand Fan Pages: An Investigation of the Effects of Social Media Marketing. Journal of Interactive Marketing, 26(2), 83-91.

16. De Vries, N. & Carlson, J. J. (2014). Examining the drivers and brand performance implications of customer engagement with brands in the social media environment.

17. Dessart, L., Veloutsou, C. (2015). Consumer engagement in online brand communities: a social media perspective, Journal of Product & Brand Management; Vol. 24 (1), pp. 28-42

18. Duffy, G. (2016). How Cosmetic Brands Drive Huge Engagement on Instagram. Retrieved on 14.02.2018 from https://www.newswhip.com/

19. Dutta, K. & Jauhari, V. (2009). Services: Marketing, Operations and Management, Oxford University Press,

20. Evans D., (2010). Social Media Marketing: the next generation of business engagement, Wiley Publishing, 1st ed.

21. Gummerus, J., Liljander V., Weman E., Pihlstrцm M. (2012). Customer engagement in a Facebook brand community , Management Research Review, Vol. 35 Issue: 9, pp.857-877,

22. Hajli, M. N. (2014) A study of the impact of social media on consumers , International Journal of Market Research 56 (3), pp.95-113

23. Hammedi, W. (2015). Online customer engagement Creating social environments through brand community constellations , Journal of Service Management, Vol. 26 Iss 5 pp. 777 - 806

24. Haven, B. (2007). Marketing's New Key Metric: Engagement. Retrieved on 14.02.2018 from http://snproject.pbworks.com/f/NewMetric_Engagement.pdf

25. Higgins, E. Tory and Scholer, Abigail A., (2009). Engaging the Consumer: The Science and Art of the Value Creation Process Columbia Business School Research Paper.

26. Hoffman, D. L., & Fodor, M. (2010). Can You Measure the ROI of Your Social Media Marketing? MIT Sloan Management Review, 52(1), 41-49.

27. Hollebeek, L. (2011). Exploring customer brand engagement: Definition and themes. ResearchGate, 19(7).

28. Hollebeek, L., Glynn, M.S., Brodie, R.J. (2014). Consumer Brand Engagement in Social Media: Conceptualization, Scale Development and Validation, INTMAR, 28 (2), pp.149-65.

29. Hu, Y., Manikonda, L., & Kambhampati, S. (2014). What We Instagram: A First Analysis of Instagram Photo Content and User Types.

30. Huang, S. (2018). How consumers become loyal fans on Facebook . Retrieved on 14.02.2018 from https://doi.org/10.1016/j.chb.2018.01.006

31. Instagram for Business. (2018). Retrieved on 14.02.2018 from https://business.instagram.com/?locale=en_GB

32. Instagram. (2018). About Us. Retrieved August 24, 2016, from Retrieved on 14.02.2018 from https://www.instagram.com/about/us/

33. Instagram: Most-followed beauty brands 2017. (2017). Retrieved on 14.02.2018 from https://www.statista.com/statistics/

34. Jayasingh, S., Venkatesh, R. (2015). Customer Engagement Factors in Facebook Brand Pages , Asian Social Science; Vol. 11, No. 26; 2015

35. Journal of Business & Economics Research, (2013). Does Social Media influence consumer buying behavior? An investigation of recommendations and purchases.

36. Kaplan, A. M. & Haenlein, M. (2010). Users of the world, unite! The challenges and opportunities of social media. Business Horizons, Vol. 53, No. 1, pp. 59- 68

37. Keller, K.L., Lehmann, D. (2003). How Do Brands Create Value? Marketing Management , No. May/June, pp. 27-31.

38. Mangold, W.G., Faulds, D.J. (2009). Social media: The new hybrid element of the promotion mix , Business Horizons Vol. 52, No.4, pp.357-365

39. Marwick, A. E. (2015). Instafame: Luxury Selfies in the Attention Economy. Public Culture, 27(1 75), 137-160.

40. Mersey R., Davis, Malthouse E. & Calder B., (2010). Engagement with Media , Journal of Media Business Studies, Vol. 7(2), pp. 39 -56

41. Milnes, H. (2016). Affordable makeup brand Nyx is forging ahead with digital stores. Retrieved on 14.02.2018 from http://digiday.com/brands/

42. Mollen, A. and Wilson, H. (2010). Engagement, telepresence and interactivity in online consumer experience: reconciling scholastic and managerial perspectives.

43. Muntinga, G., Moorman, M. & Smit, E.G. (2011). Introducting COBRAs: Exploring motivations for brandXrelated social media use International Journal of Advertising , pp.13-46.

44. NYX Gets Social Media Influencers To Develop Its New Store Concepts. (2015). Retrieved on 14.02.2018 from http://www.beautypackaging.com/contents/

45. Oliveira, M., Huertas, M., Lin, Z. (2016). Factors driving young users' engagement with Facebook: Evidence from Brazil

46. Parsons, A. & College, K., (2013). Using social media to reach consumer: a content analysis of official Facebook pages , Academy of Marketing studies Journal .

47. Patterson, P., Ting, Y., de Ruyter, K. (2006). Understanding Customer Engagement in Services, Advancing Theory, Maintaining Relevance, Proceedings of ANZMAC 2006 Conference, Brisbane, 4-6 December.

48. Peters, K., Chen, Y., Kaplan, A.M., Ognibeni, B., & Pauwels, K. (2013). Social Media Metrics: A Framework and Guidelines for Managing Social Media. J...