Internet Radio Ad Load Report
Transcription
Internet Radio Ad Load Report
Internet Radio Ad Load Report Quarter 1 2015 May 2015 © XAPPmedia 2015 All Rights Reserved Table of Contents Overview 3-4 What’s New in this Report What You Won’t Find in this Report 3 Summary Findings 3 What the Data Tells Us 4 Ad Load Declines 3 5-6 Quarter-to-Quarter Ad Load Trend Monthly Ad Load Trends 5 Publisher Ad Load Practices 6 Ad Units Per Hour Decline 5 6-7 Quarter-to-Quarter Trend Monthly Ad Unit Trend 7 Percent of 30-second Ads 7 Ad Sequencing 6 8-9 Time to First Ad (TTFA) by Publisher TTFA Distribution 9 Advertiser Participation 8 9-15 Unique Advertisers by Month Advertiser Diversity by Publisher 10 Total Unique Advertisers Grew Substantially 10-11 Advertiser Concentration 11-13 List of National and Local Advertisers 14-15 -2© XAPPmedia 2015 All Rights Reserved 9 Overview: Fewer Ads, More Advertisers Internet radio and streaming music services will depend on advertising for their economic future. As we mentioned in the Holiday 2014 Internet Radio Ad Load Report, analysts estimate that no more than 11% of listeners will opt for ad-free, subscriptionbased services. By contrast, 150 million U.S. consumers have chosen ad-supported listening. Internet radio now has large audience reach and listening hour growth has yielded significant advertising inventory. The ultimate size of that ad inventory is determined by the number of ads served per hour of listening. This metric is known as ad load. The ad load metric strips away the uncertainty of the sell-through rate of available inventory and presents only what is actually filled by ads. It also is important because eMarketer reports that many Internet radio services have expressed wariness about exceeding three minutes of total hourly ad time and risk losing audience in a highly competitive space. What’s New in this Report In the first Internet Radio Ad Load Report, we only considered data from four Internet radio services. Starting in January, we also collected data from a fifth Internet radio service and have included that data for the monthly and quarterly roll-ups. We also have incorporated more information on advertisers including a chart that depicts the relative density of advertiser concentration for particular publishers. What You Won’t Find in this Report The report does not name the Internet radio services directly. It instead presents the data from the five services either in summary form or as Publisher A, B, C, D and E. The intent is not to zero in on the ad serving strategies of any particular service, but instead to reveal the variety of approaches as well as industry-wide trends. Summary Findings Five months of Ad Load data has yielded some clear trends. The most notable trends are: 1. 2. 3. 4. 5. Decrease in Ad Load between Q4 2014 and Q1 2015 Decrease in Ad Units Per Hour between Q4 2014 and Q1 2015 Increase in the percentage of ads that were 30-seconds in length Increase in number of unique advertisers per month Increase in the total number of unique advertisers identified Since we had analyzed four Internet radio publishers in Q4 2014 and five in Q1 2015, we have presented data for Q1 with and without the fifth service, Publisher E. -3© XAPPmedia 2015 All Rights Reserved Metric Q4 2014 Q1 2015 Q1 2015 (4 Services) (5 Services) Average Ad Load Per Hour 2:44 2:21 2:18 Average Number of Ad Units Per Hour 6.50 5.20 5.06 66.53% 74.90% 79.31% 36-54 47-68 49-72 75 164 165 Percent of Ads that were 30-seconds Average Number of Advertisers per Month Unique Advertisers Identified Showing the results with all five publishers and breaking out the original four publishers presents a more consistent comparison between Q4 2014 and Q1 2015. If we consider just the original four publishes, the ad load decline is 13% whereas the addition of the fifth service in Q1 creates a new average that could be interpreted as a 15% fall. This difference arises because the new Service E carries a lower ad load than the previous average. The proper interpretation of the ad load decline is 13% quarter over quarter or 23 seconds lower. Since the average ad length across all publishers was 27 seconds in Q1 2015, this decline represents about one less ad per hour of listening than Q4 2014. Similarly, the number of ad units per hour is lower when we consider all five publishers. This is driven by the lower ad load of Publisher E and its heavy reliance on 30-second ads. The latter factor also drives up the percent of total ads that are 30-seconds in length. Publisher E has negligible impact on the total number of advertisers because the research team only identified ten advertisers on the service in the quarter and there was substantial overlap with the other services. However, the total number of advertisers identified grew rapidly to 165, up from 75 in Q4. You can see the major trend lines are similar regardless of whether Publisher E is considered. The impact is only in degree and it is relatively small. What Does the Data Tell Us? We cannot conclude for certain whether the ad load reduction was intentional or a driven by a softer advertising market. However, given the known seasonality of the advertising market, our analysis suggests that at least part of the ad load reduction was directly attributable to normal declines in seasonal budget allocations. For example, broadcast radio revenue declined 17% between Q4 2013 and Q1 2014. Despite this phenomenon, Internet radio publishers will see good news in the finding that the total number of advertisers increased materially in the first quarter. One other factor may be contributing to the ad load decline. Internet radio experienced a rise in listening hours in the first quarter. That means there were more hours to spread the ads across which would deliver an ad load decline even if advertising purchases remained flat or slightly positive. -4© XAPPmedia 2015 All Rights Reserved Ad Load Falls in Q1 As we discussed in the introduction, Internet radio ad load fell in the first quarter from a Q4 2014 average of 2 minutes 44 seconds (2:44) to 2:18. The average over the two quarters is 2:31 or about five 30-second ads per hour. Monthly Ad Load Trend If we consider the monthly trend we see a fourth quarter spike up to 2:46 in December followed by a 17.5% fall in January to 2:17. February was almost identical at 2:15 before a small increase back to 2:21 in March. These findings would seem to be in conflict with the responses from the Internet Radio & Streaming Report from RAIN News and XAPPmedia. In that report, industry insiders projected a nearly 15% rise in ad load over 2015-16. However, those projections were for the full two-year period. Seasonality would suggest a fall off in Q1 with a rise in subsequent quarters to a peak in Q4. The industry could very well meet that ad load increase prediction. -5© XAPPmedia 2015 All Rights Reserved Publisher Ad Load Practices The data also show a variance in ad load practices among the five publishers in the first quarter. Publisher A has the lowest ad load at 1:39, which reflects a fall from around two minutes in the fourth quarter. Publisher E is the next lowest at 2:05 in the first quarter. The other three publishers (B,C,D) average an ad load about 40% higher than publishers A and E. Publisher D shows the maximum ad load for the quarter at 2:45. Ad Units Per Hour Ad units per hour is different from considering the total time advertisements are playing. It reflects how the advertising time is allocated among 15-second, 30-second and other spot durations. From an advertiser’s perspective it provides insight into the options offered by publishers and whether listeners are more likely to hear many short ads or fewer, longer-form ads. With that said, “long” in the Internet radio context seems limited to thirty seconds. Ad units per hour fell 24% between Q4 2014 and Q1 2015 from an average of 6.50 to 5.06. If we eliminate results from Publisher E, the fall was 22% but you can see the trend holds irrespective of the number of services. -6© XAPPmedia 2015 All Rights Reserved When considering the monthly trend, the number of ads per hour climbs, falls and then rises slightly. There is clearly an impact of how many ads are run by the publishers promoting their own premium subscription services. A number of services run 15-second promotions which increases total ads served per session without increasing ad load. However, you can also see a clear correlation over five months between the number of ads per session and ad load. The metrics climb and fall together even if the relative changes differ. The ad units per hour numbers are heavily influenced by the percentage of 30-second ads served. You can see from the charts below that 30-second ads in Q4 were 66.5% of all ads served and they climbed to 79.3% in Q1. Conversely, 15-second ads fell quarterto-quarter from 22.46% to 7.76%. When more 15-second ads are played, the average ads per hour increases. -7© XAPPmedia 2015 All Rights Reserved As we mentioned in the previous Ad Load Report, the reason for the 30-second ad bias is unclear. Since Internet radio streams content segments in sequence instead of following clock-driven programming, ad length variability is easily accommodated. The bias most likely reflects a convenient adherence to the traditional conventions for selling, purchasing and creating broadcast radio ads. Ad Sequencing Broadcast radio follows a standard clock, typically fitting its programming and advertisements neatly within one-hour or thirty-minute blocks. Consumers that start listening at twenty minutes past the hour enter one-third of the way through a programmed one-hour block or two-thirds through a thirty-minute block. Programming will vary from hour-to-hour, but every listener receives the same content or advertisement at the same time. By contrast, Internet radio sessions start whenever a consumer begins listening and his or her content and advertising stream is unique. This leads to variable programming for each session when it comes to audio content and advertisements. The ad sequence not only varies among sessions, but also across publishers. Time to First Ad (TTFA) The average time to first ad (TTFA) was 13:41 minutes from the start of the listening session in Q1. This is up from eleven minutes in the previous Ad Load Report. Without Publisher E the TTFA would have been right at 13 minutes, which still reflects a rise from the previous report. The data also showed a rise in the variance as well as the lowest and highest TTFA. The variance between the lowest and highest increased from four minutes to over six minutes. Publishers appear to be lengthening the up-front listening session before introducing ads that deliver monetization. -8© XAPPmedia 2015 All Rights Reserved The distribution of TTFA over the period follows a skewed bell curve that is fairly symmetric between 0 and 24 minutes, but includes a tail of ads served after 25 minutes. The longest TTFA recorded was 44 minutes and the shortest was zero minutes for a pre-roll video ad. Advertiser Participation The data show a steady increase in advertiser participation each month through the five months of recorded listening sessions. There was a slight fall in total advertisers identified in January despite the introduction of a new service, Publisher E. We attribute that to the natural seasonal decline of advertising between Q4 and Q1. Even though we added Publisher E, it had little impact on the total number of advertisers identified in the period. Publisher E added two unique advertisers to the data set in January and four in March. That is driven by the fact that Publisher E has a small total number of advertisers and many are already represented on other services. -9© XAPPmedia 2015 All Rights Reserved Advertiser Diversity by Publisher In the first quarter, there was a continuation of the notable difference in the advertiser diversity among the various Internet radio publishers. Publisher A led the pack with 57 unique advertisers identified, followed by Publisher B at 47. Similar to Q4, Publishers C and D were considerably more concentrated in their advertiser base with 18 and 22 respectively. Publisher E was significantly behind all other publishers with only ten advertisers identified. We do not believe these numbers represent all advertisers on any of the publishers. However, the data does provide a relative comparison between publishers and industry trend lines over time. Total Unique Advertisers Grew Substantially The research identified 75 unique advertisers during November and December 2014 and 118 in the first quarter of 2015. Overall, 165 unique advertisers were identified during the five-month period. The number of listening sessions grew in the first quarter and that likely contributed to the growth. However, there is little doubt that more advertisers are testing the waters with Internet radio and many others are incorporating it as a core element of their marketing reach. Of the 165 unique advertisers, 137 or 83% appeared in only one publisher app. The remainder appeared on two, three or four apps and no advertiser appeared on every audio publisher. The two advertisers that appeared in four publishers were GEICO and NAPA. Nearly 50% of the advertisers identified in only one app appeared on Publisher A while only 1% appeared on Publisher E. - 10 © XAPPmedia 2015 All Rights Reserved The percent of local advertisers identified increased slightly to 15%. Five of the top 10 broadcast radio advertisers reported by the RAB for Q4 2014 were spotted on Internet radio in Q1 2015. Advertiser Concentration We considered advertiser concentration in two ways. First, we looked at the number of advertisers that were unique to a single platform. Not surprisingly, Publisher A led the pack with 49% of the advertisers that showed up on only a single platform over the entire period. It also had the highest total number of advertisers leading to the lowest advertiser concentration. Publisher E was on the other end of the spectrum. Publisher E had only one advertiser that appeared solely on its app and relied on just 10 total advertisers during the first quarter. Our second consideration was the percentage of ad spots that each advertiser commanded on individual publishers. The chart on the next page depicts two data elements. The size of the colored section represents the ad load of the publisher. Publishers A and E had the lowest ad load and you can see on the chart that they also cover smaller total area than the other publishers with higher ad loads. The boxes within each colored section represent the total percentage of the publisher’s ad inventory commanded by each advertiser. Publisher A had the most unique advertisers and the least advertiser spot concentration. The two advertisers that commanded the most spots on Publisher A were SunTrust and the publisher’s promotional ads for its own service at 3.68%. By contrast, Publisher E which had the fewest total advertisers, relied on GEICO for 29% of its ads in the first quarter. Publishers B and D filled significant portions of their ad inventory by promoting their own services. These “house” ads tended to fall into one of two categories. The first category was promotion for the publisher’s subscription service offering. These offerings typically offer ad-free listening and other features such as saving music locally to a device and the ability to play a song at any time. The second category reflected promotions features of the ad-supported service. The intent typically was to inform the listener of app features, new content on the service or new curated stations. - 11 © XAPPmedia 2015 All Rights Reserved - 12 © XAPPmedia 2015 All Rights Reserved Considering the Impact of “House” Ads It is interesting to contrast Publisher B from Publisher A by stripping out the “house” promotional ads and looking at the total ad load represented by outside advertisers. When you do this, you find that Publisher B has an 18% higher ad load than Publisher A instead of 48%. The range would be even tighter except Publisher B’s house ads average closer to 15-seconds than 30-seconds for the other publishers. By contrast, Publishers C and D are both running 2:07 of ad inventory with external advertisers and Publisher E is at 1:59. Overall, there is a clear trend. When you remove “house” ads, publishers B, C, D and E all fell within a tight range of average Ad Load at two minutes plus or minus seven seconds. We can conclude from this analysis that Internet radio is a dynamic advertising market with publishers following a variety of strategies to generate revenue and promote their premium services. We can also see that advertiser concentration risk varies greatly across the publishers. Conclusion The combination of advertising budget seasonality and a significant increase in Internet radio listening hours led to a Q1 Ad Load decline. This decline reflects nearly one less advertisement per listening session. At the same time, the number of total advertisers on Internet radio climbed, creating the opportunity for lower revenue concentration risk for audio publishers. In all, this was a vote of confidence in Internet radio’s reach and effectiveness as an advertising platform. To improve listening monetization, Internet radio publishers may have an opportunity in Q2 to increase ad load, but they would do so in a highly competitive environment and risk losing audience. Growing audience and increasing ad rate CPM will likely be the top priority for the balance of 2015. To learn more about Internet radio an audio streaming services, you can also download our other reports on Ad Load and Industry trends. - 13 © XAPPmedia 2015 All Rights Reserved List of National Advertisers *Blue indicates new advertiser in Q1 2015 5 Hour Energy ABC Accountemps Ace Hardware Advance Auto Parts AHCA American Optometric Association Amtrak Angry Orchard Applebee’s Arm & Hammer Audible Autotrader Autozone BENQ TreVolo Bing Biore BMW Burger King California Tortilla CBS All Access Comcast Xfinity Constant Contact Cox Curlys Dentyne Ice Diesel Drake’s esurance ExxonMobile Fordham University Free Slots Game Geico Gold’s Gym Graze Guitar Center H&R Block Home Depot Honda Hyundai IHOP JC Penney John Frieda JOIN.ME Kohl’s Land Rover USA Lifebeat.org Lowe’s Lyft Macy’s MasterCard McDonald’s Medexpress Mercedes-Benz Mervis Diamond Importers MetroPCS NAPA Net10 Wireless Nissan O’Reilly Auto Parts Palm Beach Tan Paramount PNC Bank Progressive Insurance RetailMeNot Ribbow Media Rite Aid Sam Adams Sam’s Club Sleep Number Bed Sony Music Special Olympics Squarespace Stamps.com State Farm Taco Bell Target TaxAct Trader Joe’s Truth.org Turbo Tax U of Texas Anderson Cancer Center Universal Music Universal Pictures UrgentRx Utz Verizon Walgreens Walmart Warner Music Wendy’s Wix Yahoo! Zengo - 14 © XAPPmedia 2015 All Rights Reserved List of Local Advertisers *Blue indicates new advertiser in Q1 2015 ABC News 7 Anne Arundel Medical Center Arlington Honda Brunswick Bowling Don Beyer Kia El Centro Restaurant Fairfax Hyundai Fairmont State Fiat of Tyson’s George Washington University Hollywood Casino Hyundai of Chantilly INOVA Health John C. Flood of Virginia Liberty, Whitetail Ski Resorts NASM Navy Federal Credit Union NBC News 4 Northwest Federal Credit Union NOVA/Northern VA Community College Pohanka Honda Pohanka Hyundai SecureTransit.org Ted Britt Chevrolet Ted Britt Ford The Chicago School of Professional Psychology, Washington, D.C. The Eye Center Washington Capitals Washington Gas - 15 © XAPPmedia 2015 All Rights Reserved WASHINGTON DC 20006 About XAPPmedia XAPPmedia™ is the leader in interactive audio advertising and is the first company to Give Consumers a Voice™ allowing instant connections with brands through mobile audio apps. XAPP Ads™ present branded content followed by an opportunity for consumers to interact with ads by voice to receive more content, be connected directly with offers or get back to more listening. Unlike other mobile ad formats, XAPP Ads are effective even when consumers are ultramobile, which means they are listening but cannot interact with a mobile screen visually or by touch. The immediate voice conversion opportunity brings more value to advertisers and increases ad unit yield for audio app publishers. XAPPmedia was founded in 2012 and is headquartered in Washington, DC. - 16 © XAPPmedia 2015 All Rights Reserved