HOW TO GET ON TV AND SCALE. A PLAYBOOK: Everything You Need to Know, from Running Your Very First Campaign to Scaling Nationwide

Transcription

1 HOW TO GET ON TV AND SCALE A PLAYBOOK: Everything You Need to Know, from Running Your Very First Campaign to Scaling Nationwide

2 OVERVIEW WHAT S THE BEST WAY TO GET STARTED ADVERTISING ON TV? We here at Simulmedia hear this question a lot, especially as a crop of newer, direct-to-consumer (D2C) brands seek scale beyond their initial, largely content, search and social marketing efforts. Some either have recently started running their first TV campaigns - and by TV we mean linear or commercial TV, the programming viewers receive over the air, through set-top boxes, or via apps or they are just starting to explore it. If you work in performance marketing or customer acquisition at a D2C brand, or if you re a founder who wants to help your acquisition teams make a more informed decision about TV, then this playbook is for you. It is written especially for D2Cers who either are early TV or pre-tv and therefore need to know how it works, given the cost-per-acquisition focus their teams likely take today. By the time you ve finished reading this playbook, you ll have an understanding of the following: It s also possible that this content may be useful to people who don t work for D2C brands. We re pretty sure you will find it very useful if you: WHY ADVERTISING ON TV CAN DRIVE GROWTH THE MEDIA MATH LANGUAGE AND MARKET MECHANICS OF TV THE INVENTORY AND PRICING OPTIONS TV ADVERTISERS HAVE TARGETING AND PLANNING MEASUREMENT AND ATTRIBUTION: TIPS ON MAKING COST-EFFECTIVE CREATIVE Are relatively or completely new to TV advertising Gather customer insights and measure the impact every aspect of marketing has on revenue Take an empirically-driven approach to marketing Take a test and learn approach to new channels, cycling through combinations of messages, creative executions, customer segments, and other variables and bidding alternatives to determine which ones will scale the fastest and with the greatest return. HOW TO DESIGN A TEST WHEN TO KNOW IT S TIME TO SCALE HOW TO SCALE Finally, what we have here won t tell you everything you need to know to advertise effectively on TV. If you want to go deeper or have questions, we ll be happy to help. Simply info@simulmedia.com or call us at (646)

3 FIRST THINGS: WHY ADVERTISE ON TV? THE MEDIA MATH OF TELEVISION ADVERTISING Brands are drawn to television advertising because it increases sales. Here are just some of the reasons why it is so effective: Television advertising has its own vocabulary. Knowing the language can help you assess, target, plan, execute and measure with a lot more confidence. Here, then, is a glossary of TV s most commonly used terms. TV OFFERS SCALE TV IS 100% PREMIUM AND FRAUD-FREE TV IS RESILIENT PEOPLE SPEND A LOT OF TIME WATCHING TV RATING Definition: Average audience per minute of an entire program (including commercials). Expressed as a percentage of a total target audience exposed to media, regardless of the target audience actually available to be reached. Example: Let s say there was a 3-minute program that in minute 1, 2, 3, respectively, had 1000, 1200, and 800 viewers out of a total target size of 100,000. The rating would be (1,000+1, )/3 = 1000 / 100,000 =.01 or 1% SHARE Definition: The percentage of a target audience exposed to media, relative to the target audience actually available to be reached. Example: Let s say only 80,000 target viewers of the 100,000 were actually available to be reached. The share would be calculated ( )/3 = 1,000 / 80,000 =.0125 or 1.25% INDEX Definition: The relative composition of a target audience watching a specific program or network, as compared to the average audience size in the television universe. Example: Let s say you re targeting W25-54 who have kids and are also cooking enthusiasts. That audience makes up about 7.4% of the total TV viewing universe. If 7.4% of the enthusiasts with kids, then that program would have an index of 100. If 14.8% of the viewers were within that target audience, then the index would be 200. The penetration of TV is about 119 million households, which is over 95% of the U.S. population. By comparison, according to Pew Research, only 70% of suburbanites, 67% of urban-dwellers, and 58% of those in rural areas are home broadband users. Among those who don t use broadband at home, only 20% of adults rely on smartphones for digital access. This means a significant portion of the country is shut-out of digital advertising. Unlike digital, TV inventory is finite. It s also immune to the shenanigans affecting digital advertising, including bots and other fraudulent views. These are just some of the reasons why so many advertisers consider TV to be the most premium of all media channels. In spite of the headlines about cord cutting, over 80% of all American households subscribe to cable, which is about as many as own washers and dryers. Think about how much time Adults 18+ spend on their phones, tablets, and PCs every day. Now add all those times together. People still watch more TV than that. Even Millennials (Adults 18-34) watch almost 3 hours of TV a day. That s more time than they spend eating, shopping, and using social media combined. For these reasons, television is an attractive advertising medium for all sorts of brands, including a growing list of D2Cs. According to the Video Advertising Bureau, 34 D2C brand disruptors collectively spent more than $500 million on TV in 2017, a 138 percent year-over-year increase, and an all-time high for these brands. IMPRESSION When an advertising message is exposed to a member of a particular audience FREQUENCY The number of times an advertising message is exposed to someone in an audience REACH The number of audience members reached regardless of frequency CPM Cost per Thousand, or the cost of delivery 1,000 impressions to members of a particular audience. GRP/TRP (GROSS RATINGS POINT / TARGETED RATINGS POINT) Definition: Gross refers to total impressions, regardless of target, and Target refers to targeted impressions. The latter also can be expressed as targeted GRPs. Calculated by taking impressions divided by the a udience universe, multiplied by 100 (impressions / [universe x 100]). REACH POINT Definition: The number of different households or persons who are exposed at least once to an advertising schedule over a given period of time. Calculated as GRPs divided by frequency; usually expressed as a percentage. CPP (COST PER POINT) Definition: Cost of reaching 1% of a target audience based on GRPs, Reach, & Frequency. Calculated as Cost / number of GRPs. Includes duplication because it is based on impressions. CPRP (COST PER REACH POINT) Definition: Cost of reaching 1% of a target audience based on audience coverage. Based on reach. Doesn t include duplication. UPFRONT Definition: the practice of buying and selling national TV advertising time on an annual basis for the October September broadcast year. With about 80% of TV media transacted in the Upfronts, media sellers get more certainty about their revenue, and ad buyers can negotiate pricing advantages and impressions guarantees, in addition to taking comfort in knowing they won t be shut out of the market when they need to be on-air. SCATTER Definition: The buying and selling of TV inventory that has not been transacted in the Upfront. Scatter gives ad buyers flexibility to enter the market only when they re ready, though it can be more expensive than inventory purchased in the Upfront

4 THE MEDIA MATH OF TELEVISION ADVERTISING A PLANNING FACT OF LIFE THE MORE PEOPLE YOUR CAMPAIGN REACHES, THE MORE EXPENSIVE INCREMENTAL REACH BECOMES: (File these insights away for after your testing phase, because they apply primarily to bigger campaigns.) In planning, the goal of the campaign is to reach some percentage of the audience a certain number times, but increasing reach can quickly escalate costs because as you reach more of an audience, it becomes harder to reach previously unexposed people with every new spot. This is especially true if marketers continue to buy the same networks/dayparts because they ll inherently reach some of the same people more than once. As a result, they need to buy more ads, which directly impacts the cost of the campaign. HOW TO MITIGATE THESE RISKS: While no one is immune to this cost phenomenon, it is possible to blunt its impact. GENERAL FRAMEWORK FOR EXPLAINING THE COST OF REACH WHILE THESE ARE ESTIMATES, THE RELATIONSHIP IS WHAT IS IMPORTANT TO UNDERSTAND Test and learn. We ll cover testing later in this document, but you ll want to know what combination of creative and targeting produces the best outcomes before you ramp up your campaigns. Buy ads across across a wider variety of networks and counter-intuitive dayparts, such as opposite gender-skewing networks. This can enable you to reach more new people with each spot and mitigate some of that cost escalation. Additionally, some of the lower-rated networks and non-prime dayparts are often cheaper because they often don t index as high for the target. The first $100K will go far, achieving an almost 1:1 balance of reach and frequency, but as you spend more money, this balance tilts more toward frequency as it becomes harder to achieve incremental reach. For an advertiser to reach the first 10% of their target audience, each point of reach will cost anywhere between $10-20K. Once you reach 30% of your target audience, each point of reach between 30-40% will cost anywhere between $50-75K. And once you ve reached 50% of your target audience, each point of reach between 50-60% will cost about $ K. Favor reach over frequency. There is sound and proven research that supports the notion that when it comes to conversion (sales or tune-in), reaching even one new person is better than reaching the same person twice. If you can buy spots that are the most likely to reach new potential customers even if that reach doesn t come at the lowest possible CPM your CPM will be higher, but you re also likely to see even better campaign results. Conversely, relying solely on the cheapest CPMs beyond your initial tests to find signal can be counter productive. It can seem that by buying a lot of impressions and/or GRPs you ll inherently deliver more audience. That s not necessarily the case. It s all about balancing the GRPs against reach, specifically buying the GRPs that reach your audience. And with people watching about the same number of networks as they have for several years, even as their options have expanded dramatically, the best way to reach your audience in today s fragmented world is to accumulate GRPs from an expanded inventory pool and network set. To get from 1-10 points of reach is X, and to get from points of reach is 15-30X. From there, the rate increases rapidly. Again, this is because once you ve begun to reach an audience, reaching new people with each spot becomes increasingly difficult which in turn leads to marketers needing to buy more ads to hit their reach goals.

5 UNDERSTANDING THE VARIETY OF TV ADVERTISING OPTIONS TARGETING AND PLANNING While TV advertising is considered 100% premium and fraud-free, there are many of different flavors of TV advertising. Here are just some of the most popular options. We ve created this table to help you understand your options and select the best one for your needs. TV advertisers today have many more options for targeting their campaigns than they did even 10 years ago. Nielsen, long the gold standard for defining TV audiences, now has company from providers that include Oracle, Experian, and even advertisers own customer data. As a result, it s easier than ever to reach your strategic marketing target on TV. Note that the more targeting filters you apply, the more expensive the campaign, so targeting should be the last attribute you adjust as you seek to scale. NATIONAL OR LOCAL ADVANCED TV OR TRADITIONAL DIRECT RESPONSE OR GENERAL MARKET AD LENGTH, E.G. :60, :30, :15: Do you want your ad to air in all markets in the same program, or do you want to cherry pick certain locations? With advanced TV you can use technology to help you more quickly and easily build, execute and measure your plans. As you scale, it also can help you target more granularly. Traditional TV ad buys are easier to make, if only because more people can help you do it, and they may be cheaper but not as measurable or able to target your strategic audience, as opposed to an age-gender demo. As the name suggests, direct response TV generally requires an offer and call-to-action. It can be less expensive; that s in part because broadcasters won t guarantee that your ads will air. If they owe advertisers make-goods, they may take it from the direct response pool. General market is guaranteed to air. That can make it more expensive, but it gives advertisers certainty. The longer the message, the more you can explain your brand. Longer ad lengths also tend to be more expensive. NATIONAL BROADCAST NATIONAL CABLE LOCAL BROADCAST LOCAL CABLE SATELLITE (E.G. DISH, DIRECTV, DIRECTV NOW) DIRECT RESPONSE ADVANCED TV: ADDRESSABLE ADVANCED TV: OVER THE TOP (OTT, E.G. HULU) ADVANCED TV: DATA OPTIMIZED LINEAR TV SCALE, US ONLY million HHs 98 million HHs million HHs 98 million HHs 32.6 million HHs million HHs 64 million HHs 59 million HHs; 14.1 million OTT-only HHs million HHs CPMS RELATIVE TO NATIONAL BROADCAST Varies Varies HERE S AN OVERVIEW OF SOME OF THE MOST POPULAR TARGETING OPTIONS. IF YOUR BUSINESS ALREADY HAS RELATIONSHIPS WITH ANY OF THESE PROVIDERS, YOU MAY BE ABLE TO RE-USE THE SAME TARGETS YOU VE ALREADY CREATED. NIELSEN provides demographic data that enables the creation of audiences based on age, gender, presence of children, household income, pet ownership, and more. ORACLE DATA CLOUD (ODC) is an aggregator and analyzer of household-level retail and CPG purchase data sourced from a network of merchant contributors. Visa has partnered with ODC to distribute Visa Audiences, an audience segmentation based on card transaction behaviors. MRI is a panel of over 26,000 US adults that is run by research agency GfK and fused to Nielsen. This panel works through a bi-annual interview survey, i.e. reported data, not observed, that captures attributes related to measure media use, product consumption, lifestyles and attitudes. EXPERIAN has demographic data on nearly all people and households in the US, with attributes similar to that of the Nielsen panel. Experian also offers Mosaic segments, which define households based on lifestyle segmentation. IRI PROSCORES provides segmentation based on all 135 million U.S. households purchase propensities (likelihood to purchase) across a broad set of consumer packaged goods (CPG) categories, subcategories, and brands. CRM DATA can be used to create a custom TV advertising audience, not dissimilar from lookalike audience creation on Facebook.

6 TARGETING AND PLANNING METRICS MAXIMIZE IMPRESSIONS MAXIMIZE REACH The TV media planning process is unique to each marketer and their individual brand goals. Some brands just want to test and learn as they dip their toes in the water for the first time. Some may want to focus on cost efficiency while others may want to sponsor big events. Some may look for mass reach, while others take a much narrower approach. Here s an overview of how it typically works, noting that when we refer to agency, we mean whomever you choose to manage your media buying, regardless if it s an internal or external resource. START WITH A BRIEF These briefs are written to help guide your agency. They should capture your campaign goals, target audience(s), timing, budget, key issues, and any other ideas that could be relevant to their brand strategy. COMPETITIVE ASSESSMENT Upon review of the brief, your agency may conduct an analysis of how the competition advertises on TV. The results may inform your media plan development. STRATEGIC ROADMAP CREATION This essentially serves as the framework for the TV media plan. It often will include media objectives and strategies, plan building blocks, audience targets, seasonality, and geography. Quantitative communication goals are also added, such as GRP levels, reach and frequency, flighting, continuity, etc. A recommended daypart mix, e.g. primetime and daytime, is established. This mix is often based on the desire to maximize reach for a given budget. CLIENT REVIEW Many agencies then hold brainstorming sessions with the marketer and/or activation team. These sessions allow for the media plan to be thought about in the context of other potential media activations. IMPRESSIONS UNIQUE REACH COST PER THOUSAND IMPRESSIONS (CPM) COST PER THOUSAND REACH (CPR) AVERAGE FREQUENCY GROSS RATING POINTS (GRPS) COST PER POINT (CPP) HOW ADVANCED TV PLANNING DIFFERS FROM TRADITIONAL PLANNING MORE AUDIENCE, LESS INDEX An audience-based approach seeks to maximize targeted audience delivery (impressions or reach) with respect to cost, while an index-based approach looks at target audience concentration within a particular program and its contextual relevance to a brand without respect to the cost of advertising on that program. MORE ADS, NETWORKS PER BUY Maximizing the reach of your target audiences in a hyper-fragmented market the average cable subscriber now receives about 200 channels often means having to buy more than 1,000 spots on as many as 50 networks. Traditional buyers, and sometimes even the sellers, may not be willing to buy this way. MORE BUYING PRECISION More advanced approaches will buy day-specific inventory because the audience often varies by day or time of week. Traditional planning doesn t tend to differentiate too much by day, often using weekly rotators, which will air an ad within a designated daypart only. For example, if a late night rotator is included in a media plan, the only thing a marketer knows is that their ad will air sometime between midnight and 6am. CHOICE OF CAMPAIGN OPTIMIZATION Advanced TV often gives advertisers two different optimization alternatives. They can maximize impressions, or they can optimize for unique reach for a given target audience. These two optimization types are fundamentally opposite, thus giving the user dramatically different results based on the method utilized. An impression-maximizing media plan will drive as many impressions as possible against the target audience, thus lowering the relative cost of impressions but also limiting the unique reach of that audience. In contrast, a reach maximizing media plan will attempt to find as much unique reach as possible in the target audience, thus lowering the overall impression load while also limiting the average frequency. See the nearby table that shows the tradeoffs between the two media plan methods.

7 MEASUREMENT AND ATTRIBUTION Measuring the impact of TV advertising is essential, but it can be a challenge in many of the same ways that affect digital advertising. For example, the longer your attribution window, the more challenging the task of linking advertising exposure to outcomes, especially when you consider the impact TV can have on the performance of other channels, not just sales. The good news is that marketers now have more ways of measuring TV advertising s effectiveness. This includes matching purchase data from either 1st or 3rd party data sets with viewing data that s sourced from set top boxes or automatic content recognition (ACR). In addition, advertisers can contract with third parties such as Marketshare Partners, Nielsen Catalina, TVSquared, and IRI to get independent measurement of TV advertising s impact. If you have all four of these inputs, the fastest way to get a read on the impact of your TV advertising on sales is to build creative that contains unique calls-to-action (CTAs) relative to your website, and then compare website visits to your campaign schedule. When ad viewers visit your site, you ll also be able to cookie them, making more detailed measurement possible. Once you get post-log reports, which show when and where your TV spots ran, you ll then be able to see the relationship between ads airing and web traffic. Regardless of how you tackle attribution and measurement, you ll need to account for some core inputs first: MAKING EFFECTIVE CREATIVE: YOU HAVE MORE OPTIONS THAN YOU MAY THINK. CREATIVE EXCELLENCE Brands that haven t advertised on TV before understandably can get nervous about the creative. There s a lot riding on getting it right, given the premium content that s on-air and the need for brands to show their best selves. But that pressure does not have to mean higher production costs and longer lead times. SPOT LENGTH CONCEPTUAL MODEL OF YOUR SALES FUNNEL A BASELINE FOR THE MOST UPPER FUNNEL METRICS Nothing impacts campaign performance more than the creative you run. That s why so many TV marketers entrust creative development to experts at agencies and test their creative ideas at the storyboard stage to determine which ones are most likely to resonate with their target audience. If you re concerned about creative costs for your first test campaign, here are four cost-saving suggestions: Avoid retainers: Some creative solutions may request or even require a retainer attached to a larger commitment, but not all do. Select one with more flexible terms and that s willing to test and learn with you. Explore animation: If you can avoid live action shooting, you may be able to save significantly. We ve seen high quality animated spots that cost in the low-to-mid five-figures to produce. Shoot overseas: If you must use live action, consider shooting it outside the U.S. Though shooting in faraway places can incur travel costs, it can cost less in the end as a result of lower labor and production costs. Produce :15s: Shorter ads mean less production and editing time, which is why they can cost less. Smart marketers also consider the maturity of their brand when they design creative. If their brand is new to the market, they ll spend more time explaining the brand and its value, and they ll embrace longer ad formats, e.g. :60s and :30s, rather than :15s. Start with the top-most level. Are you tracking website sessions or other high volume first things engagement metrics? To ensure you have a confident read on the impact of your marketing, you should first execute a baseline assessment. As an input, you will specify the confidence you would like the baseline to describe your upper funnel sessions. This will give you guidance on how many events are outside your baseline. For instance, given a 95% confidence level, one can expect 5% of events will be outside of the range of confidence just by chance. Thus, you must ensure that your budget is large enough to lead to sufficient events above the confidence interval so that they register as significant in the analysis.

8 MEASUREMENT AND ATTRIBUTION Imagine a TV campaign promoting a company s Valentine s Day products. Let s say the creative for this fictitious campaign included a CTA to visit the company s website to access a special offer. This advertiser reasonably would expect to see a spike in website visits that corresponds with the airing of the TV campaign, followed by a return to the company s normal baseline. The relationship between the ad campaign and website visits might look something like this: Similarly, this campaign also might impact organic search volume. Its chart could look a lot like the one showing website visits: VALENTINE S DAY TV AD CAMPAIGN: IMPACT ON WEBSITE VISITS VALENTINE S DAY TV AD CAMPAIGN: IMPACT ON ORGANIC SEARCH VOLUME 40,000 35,000 35,000 VALENTINE S DAY CAMPAIGN BEGINS 30,000 VALENTINE S DAY CAMPAIGN BEGINS 30,000 25,000 25,000 20,000 20,000 15,000 15,000 10,000 10, /28/18 2/27/18 2/26/18 2/25/18 2/24/18 2/23/18 2/22/18 2/21/18 2/20/18 2/19/18 2/18/18 2/17/18 2/16/18 2/15/18 2/14/18 2/13/18 2/12/18 2/11/18 2/10/18 2/9/18 2/8/18 2/7/18 2/6/18 2/5/18 2/4/18 2/3/18 2/2/18 2/1/18 2/28/18 2/27/18 2/26/18 2/25/18 2/24/18 2/23/18 2/22/18 2/21/18 2/20/18 2/19/18 2/18/18 2/17/18 2/16/18 2/15/18 2/14/18 2/13/18 2/12/18 2/11/18 2/10/18 2/9/18 2/8/18 2/7/18 2/6/18 2/5/18 2/4/18 2/3/18 2/2/18 2/1/18 WEBSITE VISITS ORGANIC SEARCH VOLUME

9 MEASUREMENT AND ATTRIBUTION As you can see, each of these three measurement alternatives can help you attribute changes in customer behavior to your TV advertising. They also can be done without the help of a data scientist. On the other hand, we d expect this advertiser s Facebook cost per acquisition (CPA) to decline during the campaign. Presuming the use of a CPM or ocpm bidding option, here s a chart that illustrates the relationship between the same hypothetical Valentine s Day campaign and Facebook CPAs: This approach has its advantages and disadvantages. POSITIVES INCLUDE NEGATIVES INCLUDE VALENTINE S DAY TV AD CAMPAIGN: IMPACT ON FACEBOOK CPAS VALENTINE S DAY CAMPAIGN BEGINS Speed of execution No data science required: Given the heavier data requirements of a more science-driven approach, this may be the most effective and fastest TV attribution and measurement hack No TV data required to measure campaign effectiveness A creative tax: In order for each ad to have a CTA, you might have to devote as much as five seconds of each ad just to showing the CTA. That s five seconds in which you can t talk about your brand and can represent as much as 33% of a :15 second spot. Short-term horizon: This approach only reliably measures advertising s short-term effects. It s possible that an advertising campaign could permanently create a new baseline for website visits or search or social volume. If that were to happen, this method isn t likely to detect it /1/18 2/2/18 2/3/18 2/4/18 2/5/18 2/6/18 2/7/18 2/8/18 2/9/18 2/10/18 2/11/18 2/12/18 2/13/18 2/14/18 2/15/18 2/16/18 2/17/18 2/18/18 2/19/18 2/20/18 2/21/18 2/22/18 2/23/18 2/24/18 2/25/18 2/26/18 2/27/18 2/28/18 If you have a data science team, they should be able to help you develop a more sophisticated method of understanding the impact of TV advertising on sales and the other channels you operate. They also can help you go next-level by assessing the impact of individual creatives, dayparts, networks and even programs. The key is to acquire sufficient volume of data over time to build and test a model. Regardless of the channels you re measuring, a dearth of data makes reliable modeling hard, frustrating marketers who just want to know if what they re doing and changing actually works. FACEBOOK CPAS The good news is that TV is the least walled garden of all the major marketing channels. Marketers have come to accept that they re not going to get reliable data from Google and Facebook that allows an apples-to-apples view of their customers and their interactions with these two channels. With the right approach and enough data, though, you can learn a lot about TV s impact on your business.

10 DESIGNING A TEST HOW TO SCALE The same factors that comprise successful digital, mobile or social campaign tests also apply to TV advertising. Let s look at two of these factors: targeting and budgeting. AIM WIDE: For your first test, aspire to reach the widest possible audience with your budget. Until you get results back, you won t know which day of the week, time of day, creative unit and target audience will be most effective. In the absence of that knowledge, your best bet is to buy enough bulk so you can get meaningful signal. If your test produces positive results, then you can buy more bulk. As you acquire more impressions, you ll be able to determine which factors have the biggest impact on your campaign s performance. From there, you ll be able to spend more on the factors that you see delivering better outcomes, e.g. shows on a particular day of the week, or :15s vs. :30s. BUDGETING: As a rule of thumb, reaching 15 million people will enable you to simply determine if TV can work for your brand. Assuming you can acquire TV inventory at a rate of $1 CPM for a Persons 2+ audience, this should cost about $50,000 for your first campaign. If you see positive results, try increasing spend to $100,000. At this spend level, you may be able to figure out which days of the week, dayparts and creatives all of which are relatively inexpensive to adjust are performing best. If you see, for example, that Sundays perform best, try moving more of your budget to that day. Because audience targeting is the most expensive variable, wait to adjust it until you have a read on the less expensive variables. This testing choice on budgetary spend will set guidelines for results that can deliver a significant signal on performance metrics. The overall test campaign could expose million W18-49, given the $50,000 to $100,000 spread. If you expect upper funnel traffic to convert at 0.1% due to this exposure, you would expect between 15K and 30K incremental sessions. This variation must be larger than the baseline assessment for the measurement time period in order to provide valid signal, which highlights the potential benefit of spending more and therefore reaching more of your target in that initial test. There are statistical tests that can help you practice experimental design such as this. Look to your partners or internal data science team for guidance. Many D2C advertisers hold all their channels to the same performance metric, mainly cost per acquisition (CPA). TV can be very competitive on this KPI. On the other hand, TV also may have a positive indirect impact on sales by improving performance of search and social. These are things you ll want to learn in your test. DID THE TEST WORK?: As we mention in the section on measurement and attribution, to assess whether your test worked, you ll want to be ready to measure impact before you begin. For example, you should anticipate driving more traffic to your website or more downloads of your app. You also should expect some change in sales. After the test has run, compare site traffic and sales to your pre-tv state, holding as many other marketing variables constant to isolate the impact of TV. Then, track changes in these metrics over varying time periods to determine when your campaign s impact has peaked. DON T GIVE UP TOO SOON: Once you re under way, don t end the test before you ve spent your full budget. Understand that TV may have a longer attribution window, even if you include a compelling call-to-action in the creative. Also, don t rely on just one test to determine how you should optimize and scale. We ll talk about scaling in a separate section, but we recommend making gradual shifts in your approach to understand the impact of campaign changes on performance. WHAT ABOUT DIRECT RESPONSE TV ADS?: Direct response TV (DR) allows advertisers to buy cheap impressions - perhaps as much as 30% less than regular ads on any given network. The downside: you may not learn much from DR campaigns because most DR inventory is preemptible. That means there s no guarantee it will air; you may budget $120,000 for a given week but only see $40,000 actually run. For this reason, it s also hard to identify variables that contributed most to performance. Again, we recommend a gradual, steady approach to scaling. Advertisers we consider to have scaled on TV spend about $1 million every month (presuming they use only :30 second spots) or $500,000 (if they use exclusively :15s). This enables them to reach about 30% of their audience every month, a number that grows to 45-50% if you spend at that level for two consecutive months. Seeing the scaling exercise as a gradual climb in which you test and learn at every stage, we suggest holding other marketing investment steady and test only one change at a time. Aim for a minimum of five of these gradual changes as you strive to find the optimal combination of targeting, creative ad mix (e.g. :30s and :15s), and media plan. HERE S HOW AN EXAMPLE OF HOW A GRADUAL SCALING MIGHT UNFOLD FOLLOWING YOUR INITIAL TEST OF $50,000: Month two: try doubling the total campaign spend to $100,000. As long as you see positive signal and improvements in your testing, keep increasing spend. $1M a month in advertising to get 30% coverage of an audience is a healthy, big number. Consider repeating this phase of the test to see if you can replicate the results. Make incremental changes in the plan, especially around the inventory you buy. If you see stronger results from a particular network, buy a bit more in a test, but don t abandon other networks altogether.

11 READY TO SCALE YOUR TV BEYOND A TEST? Here are some considerations as you plan your advertising beyond the test phase. If you want to get more advanced at a much larger scale, you ll want to explore your options of providers and tech to find the one that s best for your business. Here are some topics to explore with potential partners. We ve broken this down into three sections: audience targeting, media planning & buying, and campaign measurement. AUDIENCE TARGETING: Here we re referring to targeting strategic audiences using high quality 1st or 3rd party data sets based on observed behavior, as opposed to using traditional methods based on index and broad age/gender demographics. For example, if you re Instant Pot (a home appliance brand), would you prefer to target a traditional, W25-54 audience, or find actual W25-54 who have kids and who are also cooking enthusiasts, and then target similar people through national TV? MEDIA PLANNING AND BUYING: TV viewing audiences are more fragmented than ever. What s not often spoken about, however, is that campaigns are still planned and bought the way they were back when there were 30 networks. This has led to today s campaigns being plagued by inflated frequency and decreased reach. Technology offers a solution. With the right platform, marketers almost instantly can evaluate available inventory across all of national TV and create media plans that maximize the reach of a strategic target audience. When talking about targeting, here are a few key questions to ask: Here are a couple of questions to ask an advanced TV provider to make sure you don t have to settle for a strategy that under-serves your growth ambitions: WHAT 3RD PARTY DATA SETS DO YOU ACTIVATE? CAN I USE MY CRM DATA FOR TARGETING? WHAT PROVES THAT THE AUDIENCE I WOULD USE FOR TARGETING IS COMPRISED OF PEOPLE IN MARKET FOR MY PRODUCTS? HOW DO YOU SELECT INVENTORY? WHAT INVENTORY SOURCES ARE AVAILABLE ON YOUR PLATFORM? HOW DOES YOUR SOLUTION MAKE MY PLANNING AND BUYING MORE EFFICIENT? If you re already using 3rd party data sets from companies like IRI or Oracle Data Cloud for targeting in digital, it s important to understand if a prospective vendor can activate those same datasets on TV. By adding TV to your cross-channel targeting capabilities, you ll be able to approach your marketing in a more holistic fashion. Targeting with 1st party data may make it easier for you to reach people who are similar to your best customers. It requires that you match your data with a provider s viewing panel, so be sure to inquire about a prospective provider s data matching process. Find out which safe havens they regularly work with, if any, and how much experience they have in this area. If they ve only been down that road with a handful of clients, completing the match could become a major obstacle. A viewing target is only as good as the panel it s based on, so it s essential to know whether their panel is nationally representative and if not, where the gaps are. The sheer size of the panel only serves to ensure a higher level of statistical accuracy, but you ll want to feel comfortable with how a prospective vendor translates the people in their panel into your target audience. It s important to note that while many TV providers claim to use granular audiences for planning, when it comes to actual inventory selection, they usually revert to the industry standard and rank spots based on how they index against standard age/gender demos. To truly take advantage of the efficiencies offered by advanced TV technology, look for vendors who are able to secure the inventory recommended to reach more of your actual strategic target audience. Many advanced TV providers offer household addressable advertising, which has a footprint approximately 60% of the size of linear TV. Other advanced TV providers have relationships with only local media owners and must stitch together multiple buys to give the impression of a national footprint. Look for vendors with network-direct relationships that enable them to reach the broadest array of customers across the national TV viewing universe. Also, make sure you understand if the inventory you re buying is preemptible or guaranteed to air. Allocating 35% of your TV ad dollars to primetime inventory may seem like a good move in theory, but even reducing that to 25%, given the cost of primetime advertising, would enable you to buy hundreds of other spots across multiple dayparts to reach even more people in your audience more efficiently. Ask prospective providers for their philosophy on how your ad dollars should be allocated whether their answer is right for you will depend heavily on your brand s specific business goals.

12 READY TO SCALE YOUR TV BEYOND A TEST? MEASUREMENT: Historically, TV measurement has been based on impressions, rating points, and more recently, audience reach. It s one thing to know how many GRPs a campaign delivers or how many people you reach, but neither of those measurements provides the context needed to understand the impact of a campaign on business outcomes, nor are they enough to understand how marketers can optimize for the future. New technology can help marketers fill those gaps. In speaking with prospective providers, probe into how they use data for measurement and what sort of campaign optimizations they are able to offer as a result. Their answers should reveal if they re a good fit for your business: CAN I USE MY PROPRIETARY DATA FOR MEASUREMENT? WHAT WILL YOU BE ABLE TO TELL ME ABOUT HOW MY CAMPAIGNS PERFORM? WHAT SORT OF ATTRIBUTION OPTIONS DO YOU OFFER? DO YOU HAVE FLEXIBILITY IN TERMS OF CUSTOMIZATION? By matching exposure data with transaction data from your CRM, you ll be able to know which viewers saw an ad and what action they subsequently took. This will yield powerful insights on which customer segments were most responsive to a campaign. Additionally, if you re using 1st party data for targeting, using that same data for post-campaign outcomes measurement is invaluable for developing deeper analytics on customer behavior and media performance. The key things to listen for in this answer are whether you ll be able to understand which days, dayparts, and networks perform best in a campaign, and what kind of conversion lift rate a campaign generates among different audience segments. This knowledge not only allows for the optimization of future media buys, it can also help you refine your target audience segments to improve the conversion rate over time. There are a number of attribution models for advanced TV campaigns, from person-level matching of transaction and viewing data to the more general time-based correlation. If the prospective vendor doesn t have a model that enables you to measure your brand s KPIs or give you options in terms of attribution windows and reporting metrics, ask if they have preferred measurement partners who might. Look for best-in-class names so you can feel confident in the eventual results.

13 This concludes the playbook, though there s much more to learn about TV advertising. If you have questions, we d love to hear from you. Simply visit or give us a call at (646)