Meaning of TRP and GRP in advertising

[Hari Prasadh]

Hi,

a) First of all a basic question. I want to know as to what is the expanded form of TRP. I used to think it was television rating points. But I learned from my boss recently that it is Targeted rating points. He seems to be correct as I have this variable TRP for a Radio advertising data. I wanted to confirm this?

b) Second I would like to know as to what exactly does the phrase targeted rating point mean. How is it different from Global rating point. How are these actually measured? Is there a formula representation for the same.

c) Is there a upper limit for TRP, GRP etc (I was thinking these 2 in terms of percentages so thought that upper limit might be 100 but it doesnt seem to be so)

d) Also is it usual practice to add GRP's (sum up) to get consolidated GRP value?

e) When I want to use TV advertising as a variable in sales/volume modelling is it more correct to use GRP's rather than TRP's. Is there any conversion rate/rule between these 2?

Please guide me.

regards,
Hari
India

[Sherman Whipple]

Hari,

In a broadcast media Gross Rating Points (GRP's) are used in media buying and media effectivess analysis. They represent the Reach or audience share of your spots or placements (a function of what stations you buy and when the ads run) times the Frequency or number of spots (advertisements) run. Each GRP represents 1%.

Over a period of time and with multiple impressions, the GRP can be 200, 500 or more. For example a GRP of 100 could mean that you bought a hundred spots with a 1% reach or that you bought 2 spots with a 50% reach.

Targeted Rating Points are a refinement of GRP's to express the reach time frequency of only your most likely prospects. For example, if you buy 150 GRP's for a television spot, but you know that only half of that audience is actually your market, then you would state your TRP as 75 to calculate your net effective buy.

As for modeling, you will probably want to use the net effective buy figure or TRP, particularly if you are including costs to calculate return and sales.

[Hari Prasadh]

Sherman,

Thanks a lot for your explanation. I am able to understand the points I had posted.

4 further queries come to my mind:-

I. >They represent the Reach or audience share of your spots or placements >function of what stations you buy and when the ads run)

In the above statement I want to know what exactly does REACH mean? Does it mean the ratio of people who were watching that particular AD on TV (or listening radio) to all the people who were watching TV? I want to be sure about the denominator in the ratio (like can the denominator be -> all the people who have TV's at home irrespective of whether they were watching TV or not at that instant?)

II. Based on my query in I, I want to know as to whether both numerator and denominator talk about "number of people who were watching TV" or is it about "Number of Households who were watching TV". I wanted to be sure of this because I can have all the "households" watching my ad, but from each household if only one person is watching then my ad might not be as effective (more so if the person who is watching is not a decision maker). Also, even if it is "people" how does agencies accurately determine this figure, because the TV metres which measure/track programmes cant record how many people are watching at a time.

III. >For example a GRP of 100 could mean that you bought a hundred spots >with a 1% reach or that you bought 2 spots with a 50% reach.

Since GRP's/TRP's are to do with effectiveness, ideally wouldnt the effectiveness be more in case the figure of 50% in 2 different spots is got from the 2 halves of my population. That is I would prefer 2 different people to view the same ad at different points of time rather than the same person/household viewing the same ad at 2 different points of time. Is this point critical? If so, does the TV meters etc (are able) differentiate this?

IV. >Over a period of time and with multiple impressions, the GRP can be 200, >500 or more.

If I have GRP's for each week (for a single AD) and the GRP's have figures of 75, 200 etc. does that mean that for week 1 the particulr AD had a TOTAL REACH of 75 for first week and the same AD had a total reach of 200 next week? My query is to clarify as to whether "Over a period of time" in my case applies only to the week in question or not as I have "multiple impressions" across weeks

V. How does it work in radio case. I know Arbitron is there, but does Arbitron plug in meters in to people radio and again do they do it in car radios as well as a considerable population listens to radio while driving


regards,
Hari
India

[Sherman Whipple]

I. >They represent the Reach or audience share of your spots or placements >function of what stations you buy and when the ads run)

In the above statement I want to know what exactly does REACH mean? Does it mean the ratio of people who were watching that particular AD on TV (or listening radio) to all the people who were watching TV? I want to be sure about the denominator in the ratio (like can the denominator be -> all the people who have TV's at home irrespective of whether they were watching TV or not at that instant?)

Usually one uses the stations weighted average shares over a broadcast week, as the reach figure. This number is available from the various audit services such as Arbitron or the stations themselves. You don't have to compute it.

II. Based on my query in I, I want to know as to whether both numerator and denominator talk about "number of people who were watching TV" or is it about "Number of Households who were watching TV". I wanted to be sure of this because I can have all the "households" watching my ad, but from each household if only one person is watching then my ad might not be as effective (more so if the person who is watching is not a decision maker). Also, even if it is "people" how does agencies accurately determine this figure, because the TV metres which measure/track programmes cant record how many people are watching at a time.

Actually, that is a good question. Audience compositon is, however, predicated on the day of the week and time of day. Logs record this and the patterns are relatively stable. Weekday/daytime for example used to have an audience factor of 1, but it changes with every timeslot, and even by type of content. This information is also made available through audit bureaus and the stations themselves.

III. >For example a GRP of 100 could mean that you bought a hundred spots >with a 1% reach or that you bought 2 spots with a 50% reach.

Since GRP's/TRP's are to do with effectiveness, ideally wouldnt the effectiveness be more in case the figure of 50% in 2 different spots is got from the 2 halves of my population. That is I would prefer 2 different people to view the same ad at different points of time rather than the same person/household viewing the same ad at 2 different points of time. Is this point critical? If so, does the TV meters etc (are able) differentiate this?

No there is no differentiation. GRP is a quantitative, not qualitative or subjective measure. There are other methods of derriving this information, some of which I cover in a paper which you can download here:

http://www.whipplesargent.com/ad_res.pdf

IV. >Over a period of time and with multiple impressions, the GRP can be 200, >500 or more.

If I have GRP's for each week (for a single AD) and the GRP's have figures of 75, 200 etc. does that mean that for week 1 the particulr AD had a TOTAL REACH of 75 for first week and the same AD had a total reach of 200 next week? My query is to clarify as to whether "Over a period of time" in my case applies only to the week in question or not as I have "multiple impressions" across weeks

No. The reach value cannot exceed 100 for any given buy. This is why use the average audience share over the interval to determine one reach figure and then multiply this by the frequency to arrive at our GRP's for that buy and then add the GRPs for each by to assess our totals.

V. How does it work in radio case. I know Arbitron is there, but does Arbitron plug in meters in to people radio and again do they do it in car radios as well as a considerable population listens to radio while driving

In the past, Arbitron worked with logs and surveys and not meters like Neilsen. I presume they still do.

Sherman

[Hari Prasadh]

Sherman,

Thanks for your patience in explaining my questions. I would post again in case I have more queries/clarifications on the same

Regards,
Hari
India

[jreish]

Arbitron continues to use a paper diary method, which consists of first contacting people by phone (one contact per hh) and asking if their entire household would like to participate. Arbitron then mails as many diaries as the phone respondent says there are people in the household along with a small incentive ($2-$5 cash I believe, paying more for harder-to-reach demo cells). Respondents fill the diary in by hand for a week listing the start and stop times for each station they listened to, then mail them back in.

This methodology has been tweaked over the years but basically is the same as when they started decades ago. It is highly criticized as being impossible to accurately complete (how many different times does the average commuter switch radio stations during a rush hour drive, and how are they expected to capture that while driving?) and therefore relies too much on recall and top-of-mind awareness and not enough on actual listening. In addition, response rates have been dropping for years.

Arbitron has been developing and experimenting for more than 10 years on what the call the Portable People Meter (PPM). It is an electronic device that requires radio stations to embed a unique signal in their broadcast that is audio but is out of the range of human hearing, and the device "listens" to this signal and electronically records a log. The respondent would be asked to carry/wear this device everywhere for the survey period and load it into a docking station to upload the data. Besides not relying on recall, this device is also supposed to be able to capture exposure to stations that a respondent does not actively choose to listen to -- sitting in a waiting room, shopping in a store, etc. From the advertiser's perspective, exposure to a commercial should count in all of these scenarios.

There have been many tweaks to this device as it has been developed over the years to address specific concerns:
- What if the person is listening with headphones? The respondent is supposed to hook the PPM inline between the radio and the headphones.
- What if the volume on the radio is too low for the person to hear it? Arbitron has engineered it so that, supposedly, it will only record the signal if it is at a volume considered "listenable".
- What if the person has the device in a briefcase or in their pocket? I don't know the details but they have tried to work something out on that.
- What if the person turns on a radio and leaves the PPM sitting on a table while they leave the room? It has a built-in motion sensor and will stop logging if there is no motion for a certain period of time.

And so on.

This or a similar device appears to be the wave of the future as major advertising agencies are signing on. There have been major field tests of the device most recently in Houston and previously in Philadelphia. If/when it is finally ready for mass roll-out, the PPM potentially could be used to also capture television exposure, internet streaming media, movie theatre ads, anything that has an audio channel. Potentially even some media that do not have audio could have a speaker installed just to broadcast the PPM signal, such as billboards in an airport.

For more info go to http://www.arbitron.com/portable_people_meters/home.htm.

[Hari Prasadh]

Joel,

Thanks for detailed information on Arbitron recording method.

Regards,
Hari
India

[Sameer]

Dear Sherman,

Thanks for explaining GRPs TRPs for Mr. Hari & for the people who are reading this conversation. I would like to know few things regarding the same discussion:
GRPs is the product of Ratings & Frequency
GRP= sum of Ratings points x Frequency

Please explain; Where does Reach (Unduplicated viewership) work in the media calculations?

Many Thanks,

Sameer

[Sherman Whipple]

Hi Sameer,

Actually, GRP is the reach (or percentage of total audience at that time) of the particular buy times the frequency of spots bought. While the reach number is typically provided by a ratings service, it is not the rating.

The reach for a particular spot would, by definition be unduplicated, but this is not germaine to the discussion of GRP's since with frequency, there could be a lot of duplication, in fact repeat exposure is quite desirable. You may be confusing TURF which is another approach to evaluationg a media campaign.

Sherman