GroupM study reveals hidden effects of TV
- TV advertising generates the most short to medium term sales and dominates long term response
- TV advertising drives the highest volume of cost efficient response
- TV advertising is responsible for 44% of media-driven Facebook interactions for brands
- Research findings question conventions about planning direct response ad campaigns
London, 5 November 2015: New research from GroupM, commissioned by Thinkbox, has revealed different media channels’ effects on driving response and TV advertising’s evolving and often hidden role.
Types of advertising response measured in the study – called ‘TV Response: new rules, new roles’ – were via e-commerce, bricks and mortar and telephone. The study identified the additional consumer response which was directly attributable to media investment over and above the ongoing base level of response which is the product of many historical factors, including previous media investment.
GroupM analysed the response effects of a range of channels, including Brand Response TV*, Direct Response TV**, Radio, Print, Outdoor, Direct Mail, Affiliates, Online Display and Paid Online Search. It employed a combination of analytics methodologies, including econometric modelling and GroupM’s Spotlift tool, which analysed 1.38 million TV spots, spanning 43 advertisers and 9 sectors. Each methodology was tailored to the time period during which the effects of communications were measured: immediate, short to medium term and long term (see notes to editors).
Key findings include:
TV creates the most short to medium term sales
GroupM found that media account for on average 39% of sales in the short to medium term (within 3 months of a campaign finishing); 33% of these media-driven sales are driven by TV advertising, more than any other communication channel. By comparison, paid-for online search created 22%, online display 12%, affiliates 10%, print 8%, direct mail 8%, radio 3% and outdoor 1%.
The study also showed that TV has far reaching, but often hidden, effects across the communications system.
TV drives a response through several channels directly, generating 31% of all media-driven sales delivered via telephone, 35% of all media-driven sales via bricks & mortar and 32% of media-driven sales through web traffic driven direct-to-site (including non-paid-for-search).
This study also found that TV is responsible for driving an indirect response through online channels, generating 32% of media-driven sales via paid-for online search; 30% of media-driven sales via online display and 20% of media-driven sales via affiliate marketing.
In addition to this GroupM found that TV is responsible for driving 44% of all media driven interactions for brands on Facebook (e.g. likes and comments). This effect of TV on Facebook was two-fold. Firstly, exposure to TV advertising prompts consumers to directly engage with Facebook. Secondly, as seen above, TV drives significant volumes of sales and, after purchase, consumers go on to engage with Facebook.
TV advertising drives the highest volume of cost efficient response
The study found that, because of its reach and scale, TV advertising keeps generating a cost efficient level of response at higher levels of spend than other media. Over the short to medium term the point of diminishing returns - the point at which efficiency of ad spend decreases significantly - is much lower for TV. This means there is more headroom to spend efficiently on TV than any other channel, which enables brands to drive a high volume of response and profit.
GroupM found that spend on TV advertising can be 2.7 times higher than online channels before there is a significant decrease in efficiency, 2.5 times higher than print and 2.8 times higher than radio.
TV advertising dominates longer term response
The study found that half of all media-driven response comes in the long term (3-24 months post-campaign).
GroupM also found that Brand Response TV advertising is the most efficient driver of long term response, responsible for 52% of the impact media has in the long-term. This backs up findings from previous studies, including the IPA’s ‘Advertising Effectiveness: the long and short of it’ (2013) which found that TV advertising is vital for long term brand health.
TV advertising was 40% more efficient at driving long term response per pound than the next best forms of communication, which were outdoor and print. Online display was the least efficient generator of long term response with TV 180% more efficient.
Direct Response TV should be planned to maximise coverage above frequency
Modelling the level of response delivered per exposure for 6 direct response-focussed advertisers, GroupM found that approximately 80% of total response was generated after a viewer had seen an ad for the first or second time.
The implications of this are that direct response campaigns should maximise reach and minimise the number of times the same person is exposed to an ad more than twice. This goes against how the majority of direct response TV airtime is conventionally bought, which has been mostly daytime-only campaigns that tend to perform better for frequency than reach.
Matt Hill, Research and Planning Director at Thinkbox: “This research gives us something we’ve never had before: a comprehensive understanding of the wide-ranging effects advertising has. What’s clear from GroupM’s work is that TV advertising triggers response like nothing else; it delivers both long term brand building and short to medium term sales. It has also revealed the hidden effects TV advertising has on other media and how they work together. I hope this research can be put to use helping make advertising work even better.”
David Beale, Head of Response at MediaCom: “This work provides us with the most complete view of the role of TV to date. In order to fully capitalise on the learnings and create the maximum impact on our clients’ businesses, we need to change the way we plan and buy TV, a challenge for media agencies and media owners alike.”
- Simon Tunstill | Head of Communications, Thinkbox | [email protected] | 020 7630 2326
Notes to editors & research methodology
*Brand Response (BRTV): the more brand orientated TV more commonly aired during peak airtime. The ads are less focused on driving immediate response and as such there are fewer calls to action in the voice overs and on screen
**Direct Response TV (DRTV): the more immediate response orientated TV most commonly aired during the day time periods. The ads tend to be more rational, strongly encouraging the consumer to enquire further with voiceover and on-screen calls to action via web addresses and phone numbers.
The descriptions above are directional, but there are no hard and fast rules and in reality there can be a grey area between classifications. This is unimportant for this study. The aim was to find a way of generally placing ads into one of two camps for the purpose of making general conclusions.
Three different analytics methodologies were used in the study, each tailored to the time period over which the effects of communications was measured.
Immediate response: this is response within 8 minutes of a spot airing, as measured through GroupM’s Spotlift tool. SpotLift correlates minute-level Spot and Visit data to identify and isolate TV-driven uplift. A ‘base’ of web visits is calculated based on sampled data and modelled visit patterns. The incremental uplift of a spot is measured against this base. The results aggregated in the study cover 1.38m TV spots from brands across GroupM, spanning 43 advertisers and 9 sectors.
Short to medium term response: this is typically response 0-3 months post-campaign and is measured through econometric modelling, which isolates the impact of communications over time from other drivers of sales such as brand equity, distribution, pricing & promotions and weather. The detailed results from 15 brands are aggregated to form an overview of TV and other channels’ performance over this time period. The models span a minimum of three years, but the results quoted are from the last 12 months. For 8 of the brands, the modelling frameworks were expanded to capture TV’s effect across the ‘system’. This means that as well as modelling KPIs such as ‘sales’ or ‘orders’, other intermediate KPIs were also modelled including: brand metrics, social media interactions, plus ecommerce sales were modelled by the last click before sale for search, affiliates, display and direct to site.
Longer term response: this is response from around 3 to 24 months post-campaign and involves a two stage analytics approach. Firstly, the analysis identified the brand health metric most correlated with sales in the longer term. This varies from brand to brand; it could be a measure such as consideration, brand preference or ‘this brand is for me’. This measure is then included in the econometric models which quantify the increase in sales generated from a movement in the brand health measure – e.g. if brand consideration increases, what impact does this have on sales? Secondly, once the appropriate brand health measure has been identified, a model is developed to quantify the extent to which it is driven by comms as opposed to other factors such as customer experience and PR. The analyses of 6 brands were combined in the research to provide average learnings.