From numberwang to number win

The obsession with quoting numbers with trails of zeroes behind them has long been part of the web’s preoccupation with counting things – but as TV campaigns head down the same path, how do we make sense of the data? By Sky IQ‘s Emma Holden.

In our recent AdScience study, almost half of respondents agreed that the availability of online data has created an expectation amongst clients for better measurement in TV advertising.

Brands have seen the impressive numbers coming out of their online campaigns, and are rightly asking how they can achieve the same insight from their TV campaigns in order to answer the big questions – such as “what does someone do next after viewing our ad?”

And yet, nearly half of respondents commented that their agency could be using data more effectively. The scramble to do better with data has, unsurprisingly, led to an increase in the bandying about of oversimplified numerical superlatives.

Online and social media have occasionally been accused of ‘numberwanging’ – the term coined by comedians Mitchell and Webb to describe the act of throwing large or unique numbers around while never fully explaining why they are notable or valuable.

Certainly, we often hear that a particular piece of content garnered over a million views, twenty thousand click-throughs, ten thousand ‘likes’, five thousand downloads or five hundred streams; but what do these claims really mean?

The obsession with quoting numbers with trails of zeroes behind them has long been part of the web’s preoccupation with counting things – it’s why one of the first things underneath a YouTube video is its view count. It certainly sounds good on paper, and on the face of it, the more visible a campaign is, the better.

With TV’s increasing complexity comes a greater pressure to tread the same path. For example we are exposed to overnight ratings on a daily basis and shown impressive numbers for downloads from VoD and streaming of TV content across a range of different devices and platforms. Connected TV is blurring the boundaries: how can we derive meaning from this increasing supply of data?

BARB’s viewing figures offer a single, industry-wide and highly robust measure of who’s watching what. We can readily ascertain whether a show or ad campaign has achieved the anticipated number of viewers. We currently trade around £4.6 billion per year on TV advertising based on these figures – little wonder then, that how we interpret this data and what we do next with it, is highly important in trying to establish the effectiveness of ad spend.

The biggest challenge our industry faces is in connecting insight from a wide range of datasets, and putting it in context – thereby turning a numberwang into a number win.

Taking your viewing figure data and adding a level of context is a vital first step towards a meaningful metric: 10 million simultaneous impacts equates to 20% of the UK’s Adult viewers, for example. It’s only then that you start building an audience profile. There are increasing layers of context that can help build a far more detailed picture of viewers and their value to the advertiser. What level of understanding should we be striving for?

A good place to start is to define your TV audience. We can now combine a myriad of lifestyle variables to create more detailed profiles of a target audience, described in the advertiser’s own terms and customer segments rather than standard demographics. We can target these groups on TV and digital and optimise the reach and frequency for these segments.

We can connect with brand tracking metrics such as brand recognition, campaign awareness, and persuasion (i.e. are they more likely to buy as a result?) to build a picture of ad awareness and purchase intention.

For a more direct link from advertising to purchase, we can use customer sales data to determine who saw an ad and who went on to buy a product. And by linking with other data sets such as website traffic, store footfall, average transaction value, revenue and margin we can really drill down into what response you get for each pound spent on TV.

Developments in TV have enabled greater connectivity with other datasets. We can now join the dots between an ad, its effect on other marketing media such as direct mail, the response it receives and, eventually, whether or not it leads to an increase in purchases.

The key is to combine the right datasets in the right way. There are datasets from all corners of the marketing mix that we can employ in order to create a holistic view of our audiences.

For example, by combining Nectar loyalty card data with data from Sky IQ’s viewing panel over a bank holiday weekend, home retail brand Homebase was able to identify that key customer segments were 26% more likely to visit a Homebase store if they had seen the advertising for its bank holiday event.

Clearly, the ability to plot TV ad spend against sales is immensely powerful. Sometimes, though, the desired output isn’t as tangible.

Often, when building awareness, brands simply want reassurance that viewers are doing more than just passively viewing their ad. The first step to ascertaining this is to ensure they’re reaching the right people.

That’s where the ability to answer multi-faceted questions really comes into its own. For example, being able to identify exactly what percentage of affluent under-30s homeowners watched an ad, or the realistic conversion from the trading audience (e.g. ABC1 adults) to the actual target audience segment or customer group.

This level of data offers a far greater understanding of reach, campaign quality and targeting ability than ever before – and enables informed decisions around increasing reach or frequency for a more responsive audience.

We’re now at a stage where TV metrics are able to measure audience engagement at a more granular level than ever before. Advertisers can now extract meaning from their mountains of existing metrics, by combining them with mainstream media metrics to create the best possible holistic view of their audience.

By following consumer journeys across multi-media landscapes, we can answer those all-important questions around how to target the most valuable viewers – and what those viewers did next. And that beats numberwanging any day – by around a zillion per cent.