FACEBOOK AUDIENCE INSIGHTS IN MENA FAN ACQUISITION

Harvey Bennett, Mark Brown, Hani Anabtawi & Holly Richardson RED BLUE BLUR IDEAS

About the Authors

Harvey Bennett is a partner at RBBI and heads the performance marketing division. His experience in the digital media industry, both in the United Kingdom and the Middle East, has helped RBBI grow into a leading digital marketing business in MENA.

Harvey enjoys pulling data apart and demonstrating relationships between seemingly unrelated events. His inspiration comes from books such as the undercover economist, freakonomics and bad science.

Mark Brown is account director at RBBI. Until June 2012, Mark was the regions official Facebook reseller and instrumental in launching the original Connect Ads sales team for Facebook.

His skill set ranges across devices and into the mobile display industry. He is a key member of the RBBI mobile team and has helped launch Addictive Mobility into the region (a socially powered mobile ad network available through RBBI).

Hani Anabtawi is a data analyst at RBBI. He joined from Rocket Internet in November 2012.

Hani prepared the meta-analysis and was responsible for the integrity of the data. Together Hani, Mark and Harvey have developed a unique understanding of how consumers interact with their brands on Facebook.

1|Page

About RBBI
RBBI is a specialist digital marketing agency. Our focus is on interpreting user behavior and translating insight into actions for brands to leverage.

RBBI understands the complexity of the modern user journey from discovery, to on site user experience through to conversion. This gives RBBI a unique ability to understand the various components of a brand's digital foot print, and positions the agency as a regional thought leader and pioneer in digital marketing.

Our core services are UX/UI, performance marketing, SEO, Mobile Advertising, and Analytics.

Throughout the last two years, RBBI have helped large brands establish and improve their digital business.

About the research

Facebook advertising is a powerful tool that brands can use to reach their audience and consumers. Unfortunately, there has been little feedback from Facebook in terms of actionable best practices. RBBI felt this was an area from which brands in MENA would benefit. It took 2 months for RBBI to aggregate Facebook insights data from over 35 advertisers across 7 verticals in the region. The resulting data was passed through a meta-analysis process to ensure its reliability. In total, more than 4m likes were analyzed spanning a 5 month period (October 2012 February 2013). RBBI will be releasing a series of 3 whitepapers associated with Facebook: One on fan acquisition, another on fan engagement and a third on sharing. These whitepapers are planned to be released throughout Q4 2013.

2|Page

About the audience

This research is aimed at brands and advertisers who use Facebook as a critical part of their online marketing strategy. Advertisers who aim to improve their existing Facebook strategy or those that are new to Facebook advertising will also benefit from the insights within this paper.

3|Page

Contents
1. Like sources 2. Paid v organic fan acquisition 3. Day of the week to acquire likes 4. Like seasonality 5. Paid likes by industry vertical 6. Fan base impact on recommended page favorability 10 7. Market behavior 8. Who is interacting with posts? 9. Average size of a page 10. Conclusion 13 14 16 17 5 7 8 9 10

4|Page

Like sources
Once a brand has set up their Facebook page and started publishing content, the next step is to grow the number of connections. A brands desire for more connections should not be focused on the prestige of having a large number associated to the page, but rather their ability to communicate to a loyal audience who have shown an interest in that brand. The mechanism of a user 'liking' a Facebook page essentially gives the brand permission to appear within the users newsfeed. Therefore, the more connections a page has, the more users they will be reaching and engaging with. In this section, we will be examining the various ways a user can become connected to a page. There are over 30 sources of likes to a page, most of which are illustrated in Figure 1 below. Each of these sources can be categorized into Organic, Viral and Paid. Lets distinguish these categories below: Organic likes - When a user visits a page directly or through a recommendation generated by Facebooks algorithm Viral likes - When a user sees an activity from the page as a result of a friend interacting with the page, such as a page like, post like, comment, share, or RSVP Paid likes - Likes resulting from paid like advertisements, such as amplified page posts and sponsored stories (desktop and mobile)

Our analysis demonstrates that over 50% of total likes are coming from paid sources (Figure 3, p. 7). The fact that so many likes are attributed to paid for advertising suggests that advertisers are finding this paid medium to be effective.
Figure 1: Categories of like sources

Like Sources
35.0% 30.0% 25.0% 20.0% 15.0% 10.0% 5.0% 0.0%

mobile page

mobile page browser

parent child migration

page suggestion

others

social graph

like story

vertex page

page invite

fan context story

page suggestion

post checkin

favorites

api

sponsored story

external connect

recommended pages

5|Page

story with page

page browser

page profile

photo snowlift

video flyout

invitations

ads

hovercard

medley

mobile

search

timeline

ticker

In attempt to avoid paying for ads, brands often try to create organic likes through offering contests and other utilities to users (such as games and offers) to encourage users to like the page. They rely on the virality of the contest to carry the message to their target audience. However, it is critical to understand that a contest that is not supported by advertising will not have the reach it needs in order to generate substantial volumes of likes. This is due to Facebooks newsfeed algorithm (often called Edgerank by social marketers) that typically restricts the reach of a page to between 10-20% of its fans. Despite containing more than 100,000 ranking factors to determine whether or not a user sees a post, the initial Edgerank framework of affinity, weight and time decay still play a major factor according to Facebook. These three factors contribute heavily to what content fans actually see in their news feed. Peculiarly, during our study we noticed a decline in the number of likes coming from mobile devices (Figure 2 below), despite Facebook reporting an increase in its mobile revenue1
Figure 2: Mobile like seasonality

Mobile Like Seasonality

300000 250000 200000 150000 100000 50000 0 October November December January February

The decline in likes suggests that device targeting is not yet applied as best practice in the region as it is in other parts of the world. As a result, the competition within the auction is lower and the estimated cost per fan is also lower. This discrepancy points to an opportunity for brands in the region to exploit device targeting.

http://www.forbes.com/sites/tomiogeron/2012/10/23/facebook-q3-earnings-revenue-just-above-street/

6|Page

Paid v organic fan acquisition

Our data indicates that 54% of all likes came from paid advertising. This points to the effectiveness of advertising for generating new fans for brands.
Figure 3: Organic v paid share of Facebook likes

Organic v paid share of facebook likes

46%

54%

Paid

Organic

We believe there are two fundamental reasons for the success of like advertising: reach and targeting. Reach. Upon releasing the reach generator product, Facebook revealed that on average only 16% of a pages fans (or persons Facebook friends) would see that post organically. If a page relies just on organic reach to generate likes, it restricts the number of friends of fans who will see that their friend had liked a page. The low reach results in less likes coming from viral sources, such as the ticker and newsfeed. This all goes back to the key message that if a brand wishes to extend the reach of their pages content beyond their existing fan base, they can amplify the voice of their fans actions very effectively through paid media. In 2011, Facebook launched voice of friend like ads (at the time they were simply called Sponsored Stories). These Sponsored Stories enable brands to extend the duration and reach of a viral like by pinning these stories to the right hand side of their target audiences Facebook browser. Users began to see what pages their friends were liking, and this led to the introduction of social context within advertising. As the popularity and success of this format grew, these messages were introduced into newsfeed and eventually mobile. Having these messages in the newsfeed and mobile is much more effective than being located on the right hand side, as users generally have advertising blindness with content in this location.

7|Page

Targeting. Facebook targeting methods are increasingly sophisticated and Facebook is dedicated to continuously enhancing them. 2013 has seen Facebook establish relationships with data partners that provide advertisers with enhanced targeting options. For example, in some markets Datalogix and Bluekai can provide purchase-based targeting with data gathered from ecommerce. Epsilion offers enhanced lifestyle targeting. Acxiom can provide advertisers with clusters based upon target audience income. By being able to segment the audience into specific common characteristics, brands can target specifically which segments are relevant for them to target and position content appropriately to these audiences. This results in campaigns that are much more effective at capturing the type of audience that is relevant to your brand than organic activities. Our research confirms that Facebook ads (such as the ad unit designed to get more page likes) are significantly more effective than running incentives and contests. If you must run a contest, we recommend supplementing it with an equal investment in page like ads for exponential effectiveness.

Day of Week Seasonality

We have often asked Facebook for input on the best day of the week to post. In most cases, brands are operating under the assumption that there is a specific day of the week when Facebook usage increases. That may be true. However, it does not mean that this specific day is the most effective for you to capture likes. For example, financial services are less successful on the weekend than a fun-loving FMCG, and there are only a finite amount of eyeballs (and time) for advertisers to vie for. Our approach and recommendation is to analyze your brands Facebook insights data. In the meantime, we have provided a normalized view of like acquisition by day of week, based on our data.
Figure 4: Normalised like distribution

In Figure 4 above, we identified the number of likes generated per day, and used the number of posts for each day to ascertain that the natural days users tend to like a page are Friday and Saturday. The least effective day is Wednesday. For those readers based outside the UAE, please note that the weekend here is on Friday and Saturday, and the working week commences on Sunday. 8|Page

Like Seasonality
Generally, there is an absence of always on fan acquisition strategies in the region. However, there are specific industries that invest throughout the year. FMCG, Apparel and Automotive are the verticals most likely to invest in an always on strategy. Despite the tendency of specific verticals to invest continuously throughout the year, there is a very clear trend that occurs through the Christmas period. During our study, we saw a steady increase in the number of likes leading towards Christmas, followed by a huge drop in the week between Christmas day and New Years day (see Figure 5 below). It could be argued that the reason for this dip is that less people are on Facebook or less likely to like a page during this period. We would argue that less paid campaigns are being run. This is valuable to advertisers because less competition in the auction will reduce the CPC price.
Figure 5: Total likes

Total Likes
350000 300000 250000 200000 150000 100000 50000 0

On another note, we were interested to understand the impact that the new mobile ad formats had on like acquisition. We already understood that there was a decline in total mobile likes from October to February, but how did that stack up as a change in distribution of device likes?
Figure 6: Mobile likes v Desktop likes

100% 80% 60% 40% 20% 0%

Week 39 Week 40 Week 41 Week 42 Week 43 Week 44 Week 45 Week 46 Week 47 Week 48 Week 49 Week 50 Week 51 Week 52 Week 1 Week 2 Week 3 Week 4 Week 5 Week 6 Week 7 Week 8 Week 9
Desktop Likes Mobile Likes

Week 39 Week 40 Week 41 Week 42 Week 43 Week 44 Week 45 Week 46 Week 47 Week 48 Week 49 Week 50 Week 51 Week 52 Week 1 Week 2 Week 3 Week 4 Week 5 Week 6 Week 7 Week 8 Week 9

Mobile likes v Desktop Likes

9|Page

Facebook reported growth in ad revenue through mobile at a global level. Regionally, our data showed a decrease in the share of mobile likes. This offers an opportunity for advertisers who are sophisticated enough to target by device by entering into a less busy auction place.

Paid Likes by Industry Vertical

Naturally, some industries attach greater importance to acquiring likes than others. The industries most commonly paying for likes are travel, financial services and automotive.
Figure 7: Paid v. organic likes by industry

Paid v Organic Likes By Industry

Financial Services Automotive Apparel Education Travel, Leisure & Events ecommerce FMCG 0% 10% 20% 30% 40% 50% Paid Likes 60% 70% 80% 90% 100%

Organic Likes

By understanding the share of likes coming from paid/organic, we can estimate the relative competition in CPC by industry. For example, knowing that travel, leisure & events is competitive, we can deduce that the cost-per-like will be expensive. This is useful for forecasting and planning your Facebook campaigns.

Fan base Impact on Organic Page Recommendations

Historically, Facebook made it relatively easy for a page to spread organically. Stories like un milon de voces contra las FARC, whereby a simple page created in someones bedroom had spread over night without paid advertising are becoming increasingly less common. This is because Facebook draws its primary revenue source from advertising, so they therefore make it difficult for pages to spread virally without paid media to springboard it. 10 | P a g e

We examined the share of likes being generated through Facebooks Recommended Pages unit and extracted the pages based on size (see Figure 7 below). Recommended Pages give users a list of pages they might be interested in. Perhaps unsurprisingly, we discovered that the larger the page, the larger the share of a Recommended Pages likes.
Figure 8: Does organic growth favour larger pages?

Does Organic Growth Favour Larger Pages?

Size of Page (Likes)
250000+ 100000-249999 50000-99000 10000-49999 0-9999 0.0% 0.5% 1.0% 1.5% 2.0% 2.5%

Likes generated through 'Recommended Pages'

Once again, the key takeaway here is that unless your page has a large number of likes, the likelihood of your page to grow organically is limited. Facebook went public before the time period of which the sample was taken for this analysis. However, we found that the share of these organic Recommended Likes had declined while the share of paid likes had increased (see Figure 8 below). Interestingly, the drop in Recommended Pages occurs shortly before Facebook posted an increase in revenue in their Q4 2012 earnings conference call.2

http://investor.fb.com/results.cfm January 2013

11 | P a g e

Figure 9: Paid v. recommended like share over time

Paid v Recommended like share over time

0.6 0.5 0.4 0.3 0.2 0.1 0

10/29/12

1/7/13

10/1/12

10/8/12

10/15/12

10/22/12

11/5/12

11/12/12

11/19/12

11/26/12

12/3/12

12/10/12

12/17/12

12/24/12

12/31/12

1/14/13

1/21/13

1/28/13

2/4/13

2/11/13

2/18/13

Recommended Likes

Paid Likes

Linear (Recommended Likes)

Linear (Paid Likes)

However, when comparing the relationship between Paid Like ads and Recommended Page likes, we found a negative correlation (see Figure 9 below). In other words, the number of Recommended Page likes is lower when there is a heavy investment in paid advertising. Facebook appears to be subsidizing like acquisition when advertising budgets are low. These subsidized likes represent a greater proportion of total new likes when a page has reached a substantial size ( >99,999 likes).
Figure 10: Likes from ads v recommended page likes

Likes From Ads v Recommended Page Likes

0.6 0.5

R = 0.1198

Likes From Ads

0.4 0.3 0.2 0.1 0 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2

Recommended Pages Likes

We can deduce that, for a brand to generate any meaningful volume of organic visibility through facebook, they must grow to a minimum size of ~100k likes. However, when a brand is advertising it is less likely to generate organic likes.

12 | P a g e

2/25/13

Market Behavior
With over 1 billion users of Facebook worldwide, reaching your target audience can be challenging. Your brand may desire to acquire likes from particular markets or geographic locations. Our analyses shows the majority of brands in the region are competing for likes in Egypt, Jordan, Saudi and UK.
Figure 11: Lifetime likes by market

Lifetime likes by market

30% 26% 25% 6% 20% 8% 15% 10% 5% 0%
AE AT BD BH CA CH CZ DK EG ET FR GR HR IL IQ IT JP KW LK MA MX NL PE PK PS QA RS SA SE SK TH TR UA UY YE AZ GE NG AF NP

1% 1%

14% 3%

1%

1% 1% 13% 1% 2%

4% 1% 1% 3%

5% 0%

2%

1% 1% 1%

1%

1%

Be sure you are reaching the correct markets when you run your advertising. With the largest population in MENA, it is no surprise to see Egypt so far ahead of all other markets. However, be careful not to fall into a trap and accidently reach an audience that will never help you achieve your business goals.

13 | P a g e

Who likes your pages?

Reaching the right audience is difficult. Knowing that your communications are being delivered to the people most likely to respond to your brand in a meaningful business manner is almost impossible, even with digital and all its bells and whistles targeting methods. We reviewed the demographic data of our 40 brands and found that nearly 61% of the audience doing the liking was female (Figure 11 below).
Figure 12: Likes by gender

Likes by gender
F M U

0.22%

38.97% 60.81%

This means one of two things. Either women in MENA are more inspired by brands than men, or content being posted by brands is resonating louder with their female audience. The key takeaway here is that it is important to understand the audience with which you are communicating, and the audience most likely to amplify your message by liking you. Understanding that the males and females behave differently on Facebook is as important as communicating across generations. Figure 13 below identifies the age groups that are doing the liking. If your brand is seeking to conquer Facebook, our data suggests going directly for an audience aged 18 24. However, depending on the vertical you occupy this may change. For example, travel and leisure get more likes from men than women, and the most popular age bracket is 25 34. Although financial services are reaching an audience very similar to travel and leisure, this vertical typically concentrates its fanning on UAE.

14 | P a g e

On the other hand, sportswear is most popular with much younger age groups. We initially felt this was likely because of the popularity of the brands within this vertical. However, the FMCG vertical includes some of the most recognizable and longest living brands in history, yet it generally attracts likes from an older audience.
Figure 13: Likes by age brackets broken down by vertical

likes by age brackets broken down by vertical

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% FMCG ecommerce 13-17 Travel & Leisure 18-24 25-34 Education 35-44 Sportswear 45-54 55-64 Automotive 65+ Financial Services

15 | P a g e

Average size of a Page

As a benchmark, we found that over a 5 month period the average page size was 103,000 likes. Weve broken these figures out by vertical below:

FMCG: Ecommerce: Travel & Leisure: Education: Sportswear: Automotive: Financial Services:

315,800 47,800 39,400 8,000 141,600 161,000 33,180

16 | P a g e

Conclusions
1. Paid advertising is by far the most successful method to acquire fans 2. Mobile is a growing part of the fan acquisition arsenal 3. There is an opportunity in the region to capitalize on mobile as a source of cheap likes 4. More likes come from paid advertising than organic 5. Post on the weekend if your objective is to acquire viral likes triggered by your existing fans 6. Facebook is generating more and more likes. However, there seems to be a break in liking immediately after Christmas. This could be an opportunity for brands to take advantage of lower CPCs 7. Travel & Leisure is generating the most paid likes across the region 8. Pages with more than 99,999 likes will benefit more from Facebooks organic like units (Page Recommendations) 9. Egypt is the source of most page likes in the region 10. Women are more likely to like brand pages more than men 11. 18 - 24 year olds are most likely to like a brand, but this varies by industry 12. Average page size varies by industry

17 | P a g e