Blair Robertson, Chief Analytics Officer at TV Squared, asks why, despite impressive growth, marketers still find themselves having to justify TV ad spend?
TV is fighting fit, with forecasts predicting a on ad spend – from $71.29bn in 2016 to an impressive $77.93bn predicted for 2020 – and keeping US viewers watching for over every day. So why is it that even though TV has proved its resilience and continues to experience growth in ad spend – it often has to justify that reports of its death have been greatly exaggerated?
The answer is down to one factor: measurement. While the reach and scale of TV is still great, it has long struggled to match the precision of digital media attribution. Rather than the instant, detailed data marketers have come to expect from web-based campaigns, insights from TV campaigns are usually generated using a mixture of ratings and broad audience segments, weeks after a campaign has aired.
For marketers, this has historically made TV ad spend hard to justify. And when a 30-second slot during a premium programme like the Super Bowl can cost upwards of , it’s easy to see why many are choosing formats with better tracking capacity.
Yet thanks to recent developments in analytics, TV is starting to catch up with its digital counterparts. Using new tools, it is now possible to produce accurate and actionable insights into the impact of ads on the box.
So what can data teach us about buying TV inventory? Simply stated not all inventory is created equal, some drives high levels of audience engagement and some not so much.
Deploying measurement technology that can accurately determine the characteristics of each category enables smarter buying. This insight is available the same day ads are broadcast, enabling marketers to optimise campaigns in-flight, as well as adjust strategy afterwards.
Every day TVSquared tracks the sales impact of TV advertising on over 250,000,000 web and app users across 400 brands in 50 countries; results are different for every brand but there are some interesting observations that hold at the category level. For example, research into responses following TV ads from diet brands shows viewers are more likely to engage with diet-focused ads on a Sunday afternoon than a Friday evening — a finding that may seem obvious but one that in many cases lay undetected.
In a similar vein, ads served by young female fashion brands on Tuesdays were found to perform better than on any other day, driven by the desire to have new clothes ready for the weekend.
Unfortunately, most observations aren’t so obvious or consistent and that’s where the power of empirical ‘test and repeat’ comes in; looking cross network, rotator, creative, or region means that TVSquared will typically provide 5–10 targeted recommendations for improvement every week. Identifying and changing the best and worst 10% of the buy on a regular basis can easily add up to 20–80% increase in sales volumes or reduction in CPA over the course of a few months.
Attribution is now possible for TV ad buying and brands are starting to benefit from this by creating a bespoke targeted approach for their campaigns from this key data.
Brands can now unlock the best times to reach and connect with their audiences to optimise their TV ad campaigns, increasing engagement with key audiences.
Digital media has been billed as the agile, quantifiable, and inevitable successor of TV, but it also attracts its own measurement criticisms of late with ad fraud, trust, and transparency issues. If the industry sets its sights on new, scientific tools that precisely track the effectiveness of TV, it may at last remove the question mark over its advertising viability and serve measurable, impactful campaigns for generations to come.
About Blair Robertson, Chief Analytics Officer, TV Squared
Blair is a seasoned data scientist and visualization guru with a background in Artificial Intelligence. He has spent the past 13 years helping global companies improve decision-making and product development through the application of analytics. As TVSquared’s chief analytics officer, Blair is an expert in a range of statistical techniques such as data modeling, signal processing and machine learning, which he and his team deploy to continually improve the company’s technology platform.
Previous engagements have included time as principal analyst at Sumerian, where he led analytic engagements with dozens of the worlds largest companies. Blair has also acted as a private consultant for a wide-variety of data-driven companies in financial services, utilities and hi-tech, and as a financial technology consultant with Accenture PLC.