High Note

Digital Marketing Plan

December 10, 2014

15.570: Digital Marketing & Social Media Analytics

Danielle Russell, Kimberlee Sheldon, Michelle Villagra

Executive Summary

At this time, High Note appears to be at a crucial point in its growth phase. The company needs to both increase the number of paying users as well as increase the total users on the platform. With a limited budget, High Note will need to carefully target key initiatives. Our team analyzed High Note’s data and based off of the results believe High Note should focus efforts in 3 main areas:

1. Recognize the differences between early adopters and later users and what attracts them
2. Prioritize marketing campaigns according to high impact activities on the platform
3. Attempt to determine correlation versus causation in future marketing campaigns by utilizing control groups

External Research and Hypothesis Testing

For a company such as High Note, striking the appropriate balance between subscribers and free users is significant. If the company intends to rely on subscription revenue to remain financially solvent, a rate of 2% may be too low. However, too high a rate, say 50% may also be inappropriate. This may indicate that the freemium offering is not overly compelling thus restricting the company’s ability to generate traffic which can later be converted into premium users. Thus, High Note should target a moderate conversion rate coupled with high volume of traffic (Spotify has 25%). High Note will then need to move consumers along the ladder of participation to encourage them to upgrade to premium services as well as attract new users. We researched a number of companies who have successfully leveraged freemium business models to better understand critical differentiators driving their success (Table 1, Appendix). One of the main differences identified was these company’s ability to clearly articulate the benefits of their premium service.

Based on our research and the standard features in traditional freemium business models our hypothesis (before analyzing High Note’s customer data) is that premium users and freemium users will have distinctly different profiles on High Note. In particular, we believe premium users will be more active on the platform (i.e. more songs listened to, loved tracks, friends, shouts, etc.) and that creation of playlists is crucial in driving premium users.

Methodology

To test our hypothesis we built individual profiles of freemium users versus premium users (Table 1, Appendix). We then bucketed the data according to different sets of activities - content consumption, community, content organization,community participation and demographics.  

1. Customer Data Analysis:

1. Compared freemium and premium user profiles using mean, S.D, odds ratios to understand differences in behavior between the two groups
2. Ran a correlation matrix and individual tests to understand which variables may move users up the ladder of participation from freemium to premium.

1. Logistic Regression Analysis:

1. Using the variables identified in the user profiles and correlation matrix we ran a bivariate regression and logistic regression to understand how well each variable could predict whether or not a freemium user switched to premium.

Results

Our initial analysis of the freemium and premium customer profiles revealed distinct differences between the two groups as hypothesized. On average, premium users were much more engaged with the platform listening to roughly 2x the songs, had 2.5x the number of friends, 3.3x more loved tracks, 6x more posts, 4.3x more shouts and 2.3x more playlists. (Table 2 and Table 3, Appendix). Interestingly, when the results are bucketed according to activity sets, we are able to create a hierarchy of activity sets using the odds ratio that distinguishes premium users relative to freemium users. Users who are involved in more activities involving content organization, community participation and content consumption are much more likely to be subscribers. A correlation matrix and bivariate regression were then utilized to identify additional individual variables which may be correlated to subscribers. Table 4 shows friend country count, subscriber friend count, songs listened and loved tracks that all had R2 higher than 0.11 showing they are statistically significant in subscription rates. Similarly, the logistic regression odds ratios which are greater than or equal to 1 indicate that these same four independent variables increase the likelihood that a user will subscribe to the premium service, with subscriber friend country count being the strongest indicator at 1.0846 (Table 6, Appendix).

Digital Marketing Strategies

Based off of our analysis, any marketing strategy for High Note needs to take into account 3 main issues: 1. Current % of subscribers and where High Note resides on the cycle of upgrades 2. The hierarchy of activities that encourage subscription 3. Correlated variables and whether they are due to homophily or if there are causal effects. We have organized these considerations into our framework for High Note’s digital marketing strategy of reach, convert, and engage.

1. Reach and Attract New Users

0.03% of High Note users are premium users which leads us to speculate that High Note is still in the early adopter phase of the life cycle of upgrades (Exhibit A, Appendix). Additionally, High Note should review its current pricing and product offering at the premium level. Newer users will likely have a lower willingness to pay than early adopters so High Note should ensure they are priced appropriately and that consumers are willing to pay the subscription fee at the premium level. The company may also consider decreasing the value offered at the freemium level, the freemium product may be too appealing resulting in fewer users moving to the premium level.

• Objective: To attract new freemium users, by generating traffic to the site.
• Digital Strategy: Utilize search engine marketing (SEM) and bid for keywords, such as music streaming, free music, popular artist names, etc. to capture customers who are actively looking for a service that High Note provides. Improve upon search engine optimization (SEO) both on-page and off-page (inbound links and affiliate marketing) to ensure organic searches for keywords like those mentioned previously result in a high a page rank for High Notes across these keywords. Additionally, it is important for both SEM and SEO to work together and improve High Note’s quality score on Google Adwords. We recommend using HubSpot’s Website Grader tool to analyze High Note’s inbound marketing. Another tactic to increase reach is to utilize social media and the blogosphere. Specifically, the Facebook Social Context Ads will allow High Note to promote its free music streaming service to friends of current premium users or target specific demographics such as age, gender, interests, etc. In addition, third party data from companies like Axiom can also help High Note target specific customer segments and deliver targeted messages.
• KPIs and Measurement: When employing SEM, SEO, and social media marketing to increase website traffic the key performance indicators will first be website traffic from specific referral sites (Facebook, blogs, twitter, etc.) and campaigns (Google adwords), and percent conversion to freemium user (i.e. created a profile). It is also important to track retention for freemium users and measure how often these users return to their profile and play music. Website analytic programs both free (Google Analytics) and paid (Omniture or Webtrends) will effectively measure these campaigns and allow High Note to create goals that measure whether a new website visitor became a freemium user.

2. Convert Freemium Users to Premium

Based on the odds ratios, consumers who were involved in content organization such as loved tracks, posts, playlists were more likely to be subscribers. Therefore, to move freemium users to premium users High Note should focus on increasing engagement on the platform and identifying the hierarchy of activities that encourage subscription.

• Objective: To move more users along the ladder of engagement and increase value proposition for freemium users to convert.
• Digital Strategy: Here our main target is current freemium users. We know that in comparison to premium subscribers these users are less engaged on the platform. Therefore, we must increase engagement via incentives and interaction with premium users. By embedding a social component into the platform High Note could encourage more interaction via shared playlists or songs, top songs and playlists, etc. Secondly, High Note could analyze current freemium users who are displaying content organization behavior similar to that of a premium user and target them to receive a special one month free trial of the premium service. Lastly, High Note is currently not utilizing video to their advantage. By also allowing users to listen to songs and watch music videos, High Note could increase engagement and time spent on the site.
• KPIs and Measurement: Understand premium user session data. In other words, how do they navigate the site what songs do they listen to, how many songs per session, how often do they access the site, etc. Session data along with demographic data will provide more insight on how High Note should influence a user’s experience with the site. For example, if most premium users return to the site 4x a week then High Note should employ email marketing perhaps promoting songs they may like to encourage freemium users to logon to the site more often.

3. Engage Premium Users

Currently, it is unclear whether the variables identified as correlated with subscription are a result of homophily or causal influence. However, it is apparent that premium users can influence peers by acting as advocates for the service.

• Objective: Leverage premium users as platform influencers to encourage freemium users to move up the ladder of engagement.
• Digital Strategy: By understanding how premium users access the site (desktop or mobile), their session data, their demographics and we will have a 360 degree view of these high worth users. It is important not only to leverage their influence and brand loyalty but also to reward them. One tactic is to give premium users an incentive for every friend invite they send or playlists they share by giving $5 off their next monthly subscription. Another tactic is to create a social reward system for user generated content (i.e. song or playlist reviews). Similar to how Instagram rewards pictures with the most likes. User generated content with the most likes/comments can be features on the homepage of other prominent high traffic page of the website. Lastly, it is important to have an omni-channel digital strategy and incorporate mobile marketing. For example, implementing geo-targeting to promote bars, clubs, or music stores that play or sell a users loved tracks.
• KPIs and Measurement: In particular, if High Note decides to target users with behavior similar to subscribers using either content organization or the correlated variables, High Note should divide the group identified and keep one group as a control and use the other as a test. High Note could then monitor whether the advertising campaign is impactful or whether adoption is occurring due to other confounding variables.

Appendix

Table 1: Freemium Business Models

Table 2: Summary Statistics for Freemium and Premium Users

Table 3: Correlation Matrix

Table 4: Bivariate Regression R2 For Adopter and Independent Variables

Table 5: Linear Regression

Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.2455 on 107207 degrees of freedom

(1 observation deleted due to missingness)

Multiple R-squared:  0.03947,        Adjusted R-squared:  0.03943

F-statistic:  1101 on 4 and 107207 DF,  p-value: < 2.2e-16

Table 6: Logistic Regression on Independent Variables and Odds Ratios

Deviance Residuals:

Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Odds Ratios

Exhibit A: The Life Cycle of Upgrades

Exhibit B: Ladder of Participation

Works Used

Bapna, Ravi and Akhmed Umyarov, “Do Your Online Friends Make You Pay? A Randomized Field Experiment in an Online Music Social Network,”  

Bersin, John, “Facebook Vs. LinkedIn - What’s the difference?” Forbes, June 21, 2012, http://www.forbes.com/sites/joshbersin/2012/05/21/facebook-vs-linkedin-whats-the-difference/, accessed December 2014,

Bostrom, Raoul et. al, “Playing the Social Tune,” Business Today India, Nov 24, 2013. http://businesstoday.intoday.in/story/london-case-study-how-spotify-evolved-its-use-of-marketing-tools/1/200269.html. Accessed December 2014

Conner, Cheryl, “New Research: LinkedIn User Trends (And Top 10 Surprises),” Forbes, May 4, 2014, http://www.forbes.com/sites/cherylsnappconner/2014/05/04/new-research-2014-linkedin-user-trends-and-10-top-surprises/, accessed December 2014

Forbes, “What’s Driving Pandora’s Total User Count?” Forbes, Sept 4, 2014, http://www.forbes.com/sites/greatspeculations/2014/09/04/whats-driving-pandoras-total-active-user-count/, accessed Dec 2014

KISSmetrics.com, “The 7 Ways DropBox Hacked Growth to Become a $4 Billion Company,” KISSMetrics.com, https://blog.kissmetrics.com/dropbox-hacked-growth/, accessed December 2014

Kosoff, Maya, “Spotify Just Made it Easier Than Ever to See What All Your Friends Are Listening Too,” Dec 11, 2014, http://www.businessinsider.com/spotify-introduces-new-top-tracks-in-your-network-feature-2014-12. accessed Dec 2014

Kumar, Vineet, “Making ‘Freemium’ Work,” Harvard Business Review, May 2014.

Martin, Erik, “The Freemium Frenzy,” EContent, Nov 2012, Vol. 35 Issue 9 pg, 20-25

Oestreicher-Singer, Gal and Lior Zalmanson, “Content or Community? A Digital Business Strategy for Content Providers in the Social Age,” MIS Quarterly 2013, Vol. 37, pg 591-616

Owsinski, Bobby, “Pandora Teases Streaming Music Subscribers with a New Ad Model,” Forbes, Oct 24, 2014, http://www.forbes.com/sites/bobbyowsinski/2014/10/24/pandora-teases-streaming-music-subscribers-with-a-new-ad-model/, accessed December 2014

Pagliery, Jose, “Pandora Raises Subscriptions Fees,” CNN Money, Mar 12, 2014, http://money.cnn.com/2014/03/19/technology/innovation/pandora-fees/, accessed December 2014

Peoples, Glenn, “Subscription Model Picks up Steam,” Billboard, Oct 10, 2012, Vol. 124 Issue 38, pg 8.

Seabrook, John, “Revenue Streams: Is Spotify the music industry’s friend or foe?” The New Yorker, Nov 24, 2014, http://www.newyorker.com/magazine/2014/11/24/revenue-streams, accessed Dec 2014

Taylor, Marcus, “Digital marketing and the music industry: six key trends for 2014,” Econsultancy, Jan 2, 2014, https://econsultancy.com/blog/64051-digital-marketing-and-the-music-industry-six-key-trends-for-2014. Accessed December 2014

Yarow, Jay, “LinkedIn is Turning Itself Into a Very Valueable Media Company,” Business Insider, Feb 26, 2013, http://www.businessinsider.com/linkedin-is-turning-itself-into-a-very-valuable-media-company-2013-2, accessed December 2014