What a Nielsen Rating Point Means

There is a simple definition of a Nielsen rating:  The % of persons or households watching a Television program. So, a 5 household rating would indicate that 5% of U.S. TV Households watched that program, right?

As with most things, it’s a bit more complicated.

The size of television audiences change constantly as viewers tune in and out of different programs. Think about a 3 hour baseball game.  Some die-hard fans may watch the entire game, but many viewers will watch a few innings – including a good number who will tune in to check the score when other programming goes to commercial breaks.

To calculate the rating for the complete telecast of the game Nielsen measures every minute of the program and then produces an average.  This is an important concept:  the Nielsen rating is the average minute audience for a program.  A 5 household rating for a TV program indicates that 5% of U.S. Households were watching a program during an average minute.  It is NOT an estimate of the total number of persons who watched a program.

Why the focus on average minute audiences?  Average minute audience estimates help advertisers understand the size and dimensions of the audiences that viewed their commercials.

Let’s look at the ratings for a one hour program, and this time we will use ratings for Males and Females 18-49 instead of Households.

Date	Time	Net	Program	Nielsen Rating MF 18-49	Nielsen Program Rating
2/13/2016	8:00PM	ESPN	30 for 30:  Broke	2.3	2.6
2/13/2016	8:15PM	ESPN	30 for 30:  Broke	2.4	2.6
2/13/2016	8:30PM	ESPN	30 for 30:  Broke	2.7	2.6
2/13/2016	8:45PM	ESPN	30 for 30:  Broke	2.9	2.6

The average minute audience (the rating) for the entire program is 2.6, but notice the variation from quarter hour to quarter hour.  The audience for the final quarter hour is over 25% larger than the audience for the first quarter hour (2.9 vs. 2.3).

If you are a marketer with ads in the first quarter hour of this program you would not pay for the full program rating, you would pay for audience that was actually delivered for your ad – the 2.3 rating.  The 2.3 rating is the average minute audience estimate for that quarter hour.  This is important because deals between marketers are networks are based on delivering a specified number of rating points.  As a marketer you are basically buying ratings points.

To make things more complicated, Nielsen also factors in the audiences that watch a program following its initial live airing with DVRs and on-demand viewing.  These ratings are labeled as:

·     

           Live+SD (Live plus Same Day.  Live viewing + any viewing until 3AM following the initial telecast.)

·       Live+3 (Live plus 3 days. Live viewing plus any viewing in the following 3 days.)

·       Live+7 (Live plus 7 days.  Live viewing plus any viewing in the following 7 days.)

To read up on how Nielsen estimates media audiences, you can follow the link below, and future posts will cover how media audiences are estimated for marketing purposes, including how Gross Rating Points are calculated and what they mean for media planning.  Stay tuned.

http://www.nielsen.com/us/en/solutions/measurement/television.html

Tableau for Data Visualization and Harry Potter Worldwide

Tableau is a powerful tool for creating data visualizations and having fun with your data.  Tableau Public is free to users, and a great way to learn about data visualization and telling stories with your data.  Keep in mind though, everything you publish in Tableau will be shared with the community of other Tableau users.  That said, you can check out what others are doing with Tableau and get ideas for visualizations because you are part of a sharing community.

Tableau is particularly strong when it comes to creating maps.  Let's look at a simple example using the Box Office earned by the Harry Potter Movies in various countries throughout the world.  I start with a data sheet (in Excel) that looks like this:

Movie Title	Release Date	Country	Box Office $
Harry Potter and the Sorcerer's Stone	11/16/2001	U.S.	$317,575,550
Harry Potter and the Sorcerer's Stone	11/16/2001	Japan	$152,993,493
Harry Potter and the Sorcerer's Stone	11/16/2001	Germany	$67,802,864

Now imagine that these data go on for about 200 rows, with data for all the Harry Potter movies in about 27 countries.  I can import these data into Tableau and create a data visualization with allows users to use a dynamic world map to quickly see how the Harry Potter movies peformed in individual countries.  The visualization also allows for users to analyze the performance for individual Harry Potter movies worldwide.  Tableau does a fantastic job of converting information like a country name or a zip code into a data element that fits into a map.

Below is a snapshot of the Harry Potter visualization, but you are much better served following the link below and actually using the full dynamic features of Tableau.

Link for Harry Potter Worldwide Box Office Data Visualization

Tableau Public provides a series of training videos that demonstrate how to get started developing Data Visualizations, so visit the following link to begin your data journey:

Tableau Public

ZIP Codes for Marketing Analytics

ZIP codes can be a great source for understanding the populations of narrow geographic areas, and there are several free online tools that will quickly convert ZIP codes into demographic, psychographic and consumer profiles.

Some background on ZIPs. The U.S. government assigns approximately 43,000 five-digit ZIP codes, and each ZIP code services 40-150 local post offices.  Each digit in the 5-digit zip code narrows the region to which it is assigned, and there’s a great online site that demonstrates this (Zip Decode).  This site will allow you to geographically drill-down to a specific 5-digit ZIP Code one digit at a time.

One of the easiest ZIP tools is Nielsen’s Tapestry segmentation, which divides consumers into 67 segments based on their socioeconomic and demographic composition.  Use the link below and enter a ZIP Code.

Nielsen Tapestry

What you’ll get is a basic profile of the most common consumer segments in that specific ZIP Code area.  For example, the ZIP Code 63005 encompasses Chesterfield, MO, which is 58% Top Tier – a segment characterized by:  married couples with older children…[that can] indulge…in personal services at upscale salons, spas and fitness centers…Evenings and weekends are filled with opera, classical musical concerts..

You get the idea.  But that’s a fairly rich description for just a five digit ZIP code, and you can use their dynamic map to look at surrounding ZIP codes and also identify their most prominent Tapestry segment.  For example neighboring 63017 is 34% Exurbanites who are likely to be:  Empty nesters…sociable and hard working...[who] go online for everything, while quality instead of price governs shopping choices.

For even greater detail on ZIP code clusters, you can explore Claritas ZIP Code look-up, which provides information from their Lifestyle Segmentation System.  Their portfolio of products includes PRIZM, PRIZM Premier, and P$YCLE, but for simplicity sake let’s look at PRIZM.                 

Online they will provide the top 5 segments for any ZIP code.  Let’s look at 07063, which covers Plainfield, NJ:

Among the top segments are Brite Lites, Lil City, which is described as a group of “well-off, middle-aged couples settled in the nation’s satellite cities…who typically have no kids, and college educations.”  Unfortunately, Claritas does not provide information on the representation of each of the segments in the ZIP code, so we don’t know to the size of the Brite Lites, Lil City in this ZIP code. (That’s for paying clients).

Claritas does provide some basic demographic information on each ZIP code, including Household Size and Composition, Income, Age, Race and Ethnicity, which are valuable measures for understanding the composition of a market.

Finally, there is the Big Daddy of data collection:  The Census Bureau.  The Census is conducted every ten year by the Federal Government (it’s required by the Constitution), and is meant to provide accurate data on the population of the U.S.  We can use their ZIP Code Lookup function to dive in on any ZIP Code area (link below).  This source provides a series of tables on demographic information for the ZIP Code area.

Census Bureau ZIP Code Look-Up

Why would we use ZIP codes in analytics?  Any business that is geographically based can benefit from using these resources to understand their local market.  If you are considering opening up a additional branch of a retail store in a specific market, ZIPs can help you profile that that new customer base and better understand how to serve them.  Businesses with a broader service area can also use ZIPs to learn about their existing customers:  what ZIPs represent the highest concentration of customers?  What do those ZIPs tell us about our customers?  What are some ZIPs with low concentrations of customers?  Why are we failing in those areas?

So go online and play with some ZIP codes.

How to Use Google Trends (It's Free)

Google Trends is a incredibly easy to use tool to use, and it allows users to quickly "trend data" on a variety of topics.  These data are based on the volume of searches conducted on Google, and serve as a good indicator for buzz or interest in a topic, product or service.  To access and begin playing with Google search data go to:

https://www.google.com/trends/

Let's do a basic analysis in a few steps.

1) At the top of the page you will see an area to "Explore Topics." Since movies are usually a good topic for trending -- they need to create buzz quickly since they don't stick around very long in theaters, let's look at a current movie and some upcoming movies to see how much buzz they are creating (don't worry, we'll explain the data once we have something to look at).  We're going to use Hidden Figures (#1 at Box Office yesterday) as our Topic.

Google defaults to a 5-year trend so we will change that to look at only data for the past 30 days:

Also, see the dropdown menu which indicates Worldwide, I'm going to change that to United States to keep my data more focused (no image since I hope you're getting the hang of this by now).  Here's the topline results (again, we'll discuss interpretation and more detailed analysis later).

2)  Now that we've got some data to look at let's make our analysis richer by adding some competitors.   Note the box on the upper right part of the page labeled "+ Compare."  Let's use that to make some comparisons to other films: La La Land and Fences (both of which had earlier release dates).  Below you'll see those results for those three films:  Hidden Figures (Blue), La La Land (Red) and Fences (Yellow).

Now that we've got some substantial data to work with, let's take a moment to understand what the data really mean with some interpretation.  The data are indexed to 100, meaning that 100 represents the peak volume of interest (searches) for the entire data set.  On January 9 La La Land produced a 100 index value, which was the day following its seven-win victory at the Golden Globes. Fences peaked with an index value of 78 on December 26, which is the day following its release in over 2,000 theaters nationwide. Hidden Figures peaked on January 7 with a value of 86, again consistent with its release to over 2,000 theaters (from the 25 it was playing in earlier).

Let's take a moment to interpret those values again.  La La Land had a 100 index value on January 9 and a 43 on January 8, indicating that its search volume more than doubled (43 to 100) with its Golden Globe wins.  We can do the same analysis with any of the data for the films or the data between the films. So, if La La Land had a 100 on January 9 and Hidden Figures produced a value of 86 two days earlier we can see that La La Land's peak was 14% higher than that for Hidden Figures (100 vs. 86).

Here's how Google explains the index values:  

Numbers represent search interest relative to the highest point on the chart for the given region and time. A value of 100 is the peak popularity for the term.

Cool?

3)  Data analysis always benefits from looking at the dimensions of the data, and Google Trends provides some great features that make this easy.  Let's focus on Hidden Figures, and scroll down the screen from the trends chart we've just created.  Here we'll find a map that looks like this:

If we scroll over the map it will identify the index value for each state, and we can see that Maryland produced the most interest with a value of 100 (remember it's an index, so that's the peak value), followed by the District of Columbia (98) and Virginia (87).  If you click on the Subregion menu and then the List View icon that is directly next to it on the right, you'll get a list of the states and their index values.

You can play around with this and drill down to Metro areas, but at this point you should understand that Google Trends allows for drill-downs by region.  If we had done our analysis on a Worldwide basis we could have compared Google search volume for different countries for each of our topics.

4)  Finally, let's briefly look at Related Queries.  These are other search terms that were searched by the same persons who searched our Topic:  Hidden Figures.  This is helpful information because it can help us design a search engine strategy or fathom what it is potential customers are looking for when they search our Topic.  In this case it's mostly related search terms like Hidden Figures movie, but we also see that a competitive film, Fences, was also searched by many of the same persons who were searching Hidden Figures. Also, searches for "hidden fences," which is a mash-up of two movie titles appeared to generate some search volume (following some confusion at the Golden Globes where presenters muffed the title).

There you have it.  There's more you can do with Google Trends by playing and experimenting, but this walk through should be enough to get you started.