The Value Report

The Value Report

Product Placement Evaluation Methodologies Explained and Compared

Executive Summary

On 28th February “paid for” product placement was allowed in UK television. NMG’s Value Report compares our  evaluation system, Tracker™ 2011 with published references describing the ITV1/Repucom model.

Key issues identified are:

  • NMG’s Tracker™ 2011, launched in 1987, is a bespoke product placement evaluation system, similar to those later adopted in the USA, by ITVx in 2001, Nielsen’s “PlaceViews” in 2003 and TNS Media Intelligence – “Branded Entertainment Reporting Service” in 2005.. These are established industry standard systems.
  • ITV have appointed Repucom to undertake “paid for” product placement tracking and evaluation. Repucom’s “Human Visual” system, (computerised aided brand recognition), is a high volume sports/event brand tracking tool based on SpikeNet technology.
  • Computer aided systems only identify brands according to a series of pre-agreed criteria, they do not calculate saliency.
  • Saliency scoring requires human intervention.
  • ITV’s pricing model can escalate spot rate card by up to 200% as a result of human intervention.
  • Tracker™ 2011 permits “like for like” value comparison between the emerging “paid for” and the well established free prop supply product placement market.
  • Tracker™ 2011 tracks over 6000 brand appearances a year and its bespoke Oracle database allows instant calculation of Brand/Category Share of Placement Voice, trend analysis, historic comparisons, data mining. This wider collection of brand data, over a long time period, permits interpretation of product placement value/worth against a range of relevant benchmarks – sponsorship, PR activity, serendipity, for example, as well as by reference to spot rates.

Background: SpikeNet and Magellan

There are two principal computerised logo spotting solutions: SpikeNet used by Repucom and OmniPerception’s , Magellan™ licensed in 10 countries.  In the UK, Margaux Matrix, the Formula 1 tracking agency, has been actively working with this technology since 2004.

http://www.repucom.net/technology.htm#Technology

http://www.omniperception.com/products/magellantm_brand_exposure_analysis

On 5th November 2010 ITV announced the adoption of Repucom’s “Human Visual” system:

Messer [Head of ITV Commercial Research] told Research today that it is still not clear what the UK’s market for product placement – or the research that lies behind it – will look like. “Nobody in the UK has been measuring this market or got a tried and tested methodology,” she said. “We don’t really know yet how product placement is going to behave and what the advertisers will want to know.””

Comparison: ITV/Repucom “Human Visual”  with NMG’s Tracker™ 2011 “Human Only” Systems.

ITV/Repucom “Human Visual” system

Computer based sports systems work by reference to logo “recognition”, often of one second duration, whereas product placement systems look at the total duration of the branded shot, or sequence and call this a “visibility”.

The principle strength of computer aided systems come from being able to ”see” and process  1000 to 200,000+  logo appearances within, say, a fast changing sports event. Two quick examples demonstrate:

Example One: A major international rugby match in a sponsored series with perhaps 10 sponsors. Their branding will appear on:

  • 30 players shirts/socks/headbands in a variety of places and the players themselves will be running, scrumaging, turning, falling, sometimes covered in mud.
  • Substitutes, team managers, physios
  • The pitch itself.
  • Goal posts
  • Perimeter signage
  • In the crowd on flags and banners
  • Hospitality vehicles/team coach
  • Interview rooms used for pre and post match interviews
  • Practice sessions….

From the single 90 minute match literally hours of action coverage will be generated – practice, interviews, the event and replays, cut down repeats, news coverage, chat show coverage, highlights…

Example Two: Formula One racing has 218 brands involved in 19 races generating about 60 hours of television coverage. Say, each brand got a hash grand total average of 15 min coverage which equals 900 recognitions per brand, overall 196,200 recognitions to be processed, evaluated and reported.

(In 15 races in 2009 the Virgin Brawn F1 team alone cumulated 9 hours of international media coverage, or in “recognitions” 32,400.)

http://www.autoevolution.com/news/virgin-group-benefitted-from-60m-brand-exposure-with-brawn-12216.html

Good timely analysis can identify the need, for example, to move perimeter signage or enhance logo standout to increase the sponsor’s visibility. The computer’s speed in processing mass data means this vital information can be acted upon before the next match or race takes place. “Human only” systems cannot compete.

However, even at the pre-analysis stage subjective human input is needed to programme the computer’s  recognition software.  For example, the software must be told what the minimum duration of shot to be ignored is, less than one second?  Will 50% of the logo being seen count as a “recognition”, or should it be 100% only? If the camera pans backwards across the logo does this count as “recognition”? Is an upside down logo worth the same as a right way up? Does a muddy logo count? What happens when other sponsors’ logos are in shot, what is the logo clutter mark down?

Get this bit wrong and you can dramatically increase the number of meaningless visibilities that the computer scores.

Further, derived from its sports evaluation heritage Repucom’s “Human Visual” system divides the TV screen into 9 parts, the centre one having maximum value, the bottom right least value.  Sports cameramen are trained to keep the action centre screen and this is where the viewers’ eyes will concentrate.

However, “hero” prop placement doesn’t work that way. Directors enjoy using the high definition, 16:9 TV screen like a cinema screen; the whole screen is fertile ground for brands, as demonstrated by this NMG free prop supply placement with the BlackBerry bottom right next to the diary in ITV’s Secret Diary of a Call Girl:

“Computer aided” is not infallible. Computers have trouble telling if the logo is on, for example, a player’s shirt central to the action, or on a perimeter sign, passive to the action. However, in high volume, sports events if the computer misses a few brand appearances, or calculates them wrongly, if 3019 brand appearances should have been 3038, and a few shirts are mistaken for perimeter signs and vice versa, it really does not matter.

NMG’s Tracker™ 2011 – “Human Only”

NMG’s product placement education was in Hollywood in the 80s.The advice given to the author was simple and remains the basis of NMG’s evaluation philosophy today:

“Visibility analysis should be based on what an average person viewing the programme, in average viewing circumstances, would see. If it is on TV, replicate home viewing, if it in the theatre [cinema] go the theatre [cinema] and watch it with pop corn on your knee and a notepad on the other!”

NMG’s Tracker™ 2011 benefits from a water tight total “human loop”. NMG’s placement team attend client meetings and discuss and understand client brand briefs.

The same team works directly with productions, reading scripts, going on set, understanding the characters and story lines and spotting where the brand’s and production’s  requirements overlap and then supplying the product.

NMG’s production team also runs the visibility reporting, they know the brands and they know the productions. A continuous “human loop”. In this way NMG simply does not have issues with contextual assessments.

NMG’s Tracker™ 2011 has 5 levels of saliency, the “1 star” has a zero rate card value for those fleeting appearances. For example, with the Dolce Gusto/This Morning  study 31 of the 110 appearances measured were NMG 1 Star – no value accrued. (ITV/Repucom would attribute 15% of spot value).

What NMG calls saliency, ITV calls contextuality. ITV proposes to apply contextuality scores from 15% rate card to 200% rate card. How and, more importantly, who, is going to apply up to 185 different contextuality levels?

People versus computers?

In comparing reliability, on one side computers will never get tired, humans do, which is why NMG’s visibility reporting is rotated amongst the trained staff. However, computers are also not infallible. You only have to stand in line at an airport whilst the passport machine’s biometric facial recognition programme fails to recognise you.

Typically you are directed to a passport official and in a fraction of the second the passport officer’s trained brain’s cognitive abilities are demonstrated. He/she glances at you, recognises that since the passport photo you have got a tan/gained weight/had a haircut/grown a beard or all of these and waves you through…and the passport officer repeats this process with great accuracy thousands of times each day.

Product placement in UK TV produces relatively low levels of visibility. Empirically brand appearances per hour range from less than 5 up to 15 to 25 on a production such as Channel 4’s Shameless. This next case study demonstrates product placement brand appearance levels.

Volvo Case Study: Midsomer Murders, S14E01. Filmed August 2010, broadcast 23 March 2011.

This classic programme has a running time of 1 hour 29 minutes. “Vizzing” – viewing the show, recording and evaluating, saving stills of DCI Barnaby’s product placed Volvo S80 took 18 minutes. (Volvo is not a NMG client).

Twelve separate shots included the Volvo S80 totalling 189 seconds of running time. As this was a “hero” prop placement the individual quality of these shots was high including 5 x 5 star, 2 x 4 star and 3 x 3 star.

The trained observer had no trouble in identifying these “grill” shots and we assume neither would a computer aided system:

However, the following stills come from two sequences which formed part of a 27 second and 31 second series of linked action shots. Presumably the computer would have only seen a dog and grass, and not only not recorded the scenes, but would fragment and reduce the overall on-screen duration thus downgrading their value?


In NMG’s experience saliency is heavily linked to actions that may take place in a different scene or even episode. No computer based scheme can make these connections.

Consider these examples:

  • The character holding a Boots electric toothbrush and is smiling with joy? Good- tick the 158% box! But next scene it turns out he has something stuck in his teeth.
  • The lead actor jumps into the Toyota, close up shot of the logo. Good – tick the 163% box! But in the next scene the car breaks down.
  • A character calls for a soft drink by name. Great – Tick the 200% box! He turns out to be a serial rapist

NMG’s Tracker™ 2011 system excels because the trained human observe outperforms a computer in this low volume, high saliency brand visibility assessment.

The President of ERMA, the US trade association for entertainment marketing agencies and studios, confirmed that to his knowledge, all product placement evaluation in the States is undertaken by trained observers and not by computerised sports evaluation techniques.

http://www.erma.org/

NMG’s bespoke Oracle database permits “you name it” data mining, instant reports, SOV analysis, year on year trends all with full audit trail, stills and video clips in support.

Finally, computer based logo recognition software cannot measure what it cannot see. A character pours a glass of Evian from the distinctive branded bottle into an unbranded glass and passes it to the co-star. Next scene she sips and says “this is so pure and refreshing”. The computer will only “see” an unbranded glass with a clear liquid in it and will exclude this valuable scene from further analysis.

NMG understands that ITV/Repucom will use human “intervention” to assess contextuality, spot and insert missing brand appearances, although it is not clear by whom and where this will happen. Given the nature of product placement visibility reporting this person, or people, will have to watch the programme in real time.

Which poses the question: if someone is going to watch the programme in real time, why bother with the computer in the first place?

NMG  Conclusions:

Clients and agencies view ITV’s spot ratecard as the beginning of a negotiation and will not want to rely on prices calculated by ITV according to its own parameters.

The use of a high volume, computer driven, sports evaluation technique clouds the evaluation process without obvious advantage to clients.

NMG’s Tracker™ 2011 provides clients with an independent, industry standard  pricing tool used in the UK today by over 90 major brands.

NMG’s methodology captures all placements in one process, not three, irrespective of whether the logo is in shot or not, provided there is a tangible value to the brand.

The system is transparent and operated by an independent company. NMG will model/track/evaluate “paid for” placements on a fixed fee basis thereby having no vested financial interest in the outcome of the evaluation.

Tracker™ 2011 is transparent, has a clear audit trail and thus makes it straightforward for clients and ITV to appoint an independent “referee” should a saliency judgement be in dispute. This will prevent time consuming disputes and help grow this new market on long term foundations.

Report author:

John Barnard, FCA,

Non Executive Chairman, NMG

John.barnard@newmediagroup.co.uk

30 March 2011

For further information, or for a fixed fee quotation, please contact:

Sarah Curran,

NMG Product Placement

Business Development Director

Sarah.curran@newmediagroup.co.uk

Tel: 01753 655866

Posted in: