GM/C = X versus CICO
The Basics of Direct Response Radio / Part V
How do you track Media Cost Per Sale?
Once you've selected a few stations that will deliver your Target Audience and Qualified Audience at reasonable CPMs, you're almost ready to place a test buy.
Of course you hope to achieve an acceptable MCPS, but how will you track it? The most common method is CICO (Cash In : Cash Out). How much did you make? How much did you spend? But MCPS is a function of three variables: creative impact, media cost, and conversion rate. GM/C=X tracks all three variables simultaneously and is, therefore, much more accurate than CICO.
Media Cost Per Sale is the single most important number in Direct Response.
MCPS must be an acceptable percentage of gross revenues or a dollar figure attached to a unit sale. All your creative, media, telemarketing and on-line conversion strategies must result in an acceptable MCPS. In most cases, your skein of suppliers won't pay any attention to the Bottom Line. That's your job.
The easiest way to look at this problem is to run through the results of a typical low cost buy:
You spend $200 on a local radio station.
You get 10 calls and 4 sales.
CICO Reports:
$200 Media / 4 Sales = $50 MCPS.If your Thing sells for $100, then a $50 MCPS may be just fine. Who do you congratulate - the guy who wrote the copy, the media rep, the TM company, or your web designer?
You might easily assume that $200 worth of radio will always produce the same results. So you spend $2,000 on more spots.
Oops! Only 31 sales!
$2,000/ 31 = $64.52
Who gets the ax?
Cash In : Cash Out (CICO) is a servant of modest intellect. He (I call him "Chico") can compare two stacks of money and tell you which stack is bigger.
What CICO cannot do is tell you why your radio spots worked, or didn't. Ultimately, you have to satisfy CICO, but along the way, he is remarkable inept at telling you what's working and what isn't. If anything burps, CICO tells you to bail. Very often he's dead wrong.
GM/C = X is a simple arithmetic formula that tells you what's going on at every stage of a media test or rollout.
It condenses three separate variables - response rate, media cost, and conversion rate - into one tidy statement.
G = Gross Impressions per call, measured in 000's.
M = Media Cost Per Thousand.
C = Conversion rate.
X = Media Cost Per Sale.
Here's how it works.
Your original $200 buy got you four spots. Each spot delivered 9,500 listeners. Your four spots produced
9,500 x 4 = 38,000 GIMPS.You got 10 phone calls.
38,000 / 10 = 3,800 GIMPS per Call.
Since we buy media in 000's we measure response in 000's. So, we move the decimal point over three places.
G = 3.800
You spent $200 for 38,000 GIMPS.
($200 / 38) = $5.26 per thousand.M = $5.26
Your conversion rate was 4 sales /10 calls = 40%.
C = .40
Therefore, since GM/C=X:
(3.800 x $5.26) / .4 = $50.00
You bought a medley of Fixed Position, ROS and streaming spots that ran at various times of the day and night. But CICO didn't bother to add up your Gross Impressions or check CPM.
If he had, CICO would know that you actually bought 294,550 GIMPS according to the most recent Arbitron ratings.
M = ($2,000 / 294.550) = $6.79
Your response rate and conversion rates were the same. You got 77 calls and converted 40% to a net 31 sales. (Do the math for practice.)
Therefore, since GM/C=X:
(3.800 x $6.79) / .4 = $64.52
Same results as CICO, but now you know why.
Your creatives and TMs did just fine. You just paid a little too much for media. Next time, check those CPMs!
But wait, there's more!
Let's assume you test the :60 on a local station and contemplate a rollout across the fruited plain. Network radio costs much more per commercial, but much less per thousand.
If you run that same spot on a network radio program like Sean or Rush at say $2.89 per thousand, what X would you expect?
(3.800 x $2.89) / .4 = $27.46
Clearly low network cpms reduce MCPS, all other things being equal. But they never are. Response rate, conversion and cpm are dynamic factors. GM/C=X lets you track all three simultaneously.
Using sixth grade arithmetic, you can solve the basic equation for any variable, given the other three.
GM/.C = X
.CX/M = G
.CX/G = M
MG/X = .C
POP QUIZ! If your target X is $35, M is $2.89 and
you can count on your telemarketers to close 28%
of sales, what's your highest acceptable G?
I give all my clients an Excel spreadsheet that lets us figure various combinations of G, M and C. Here's a snapshot that recaps the discussion above.
You can use the same method to track results from umpteen network programs that run in 150 DMAs every week.
EXAMPLE: You run two spots on Sean, one on Rush, and three on ESPN nationwide. You get 186,000 GIMPS in Dallas and 56 sales. You also get 38,500 GIMPS in Wichita and 18 sales. Which is the more responsive market, Dallas or Wichita?
Wichita! It boasts a much lower G per sale.
In practice, I add up the weekly GIMPs from 20 to 30 different programs for each of the top 150 DMAs. I get INQs or SALES from my clients and let Excel compute weekly G for every market in America. Does your agency do that?
Practical GM/C helped me guide Dinovite from $5,000 to $180,000 in weekly revenues in 18 months. Details.
But wait, there's even more!
In practice, telemarketer conversion rates tend to be fairly steady over time. Websites normally convert 1% to 3% of KeyWord searchers to customers, unless you Slickify your web site.
CPM is always subject to negotiation.
G is the key variable. It tells you if your creative is building interest or burning out.
RULE: Run The Same Copy Until G Turns Up.
In a local copy test track G daily. In a local or network rollout, track G weekly. G will likely behave erratically for the first few days, but it should dive as soon as a few F2s & F3s come on board.
As long as G trends down or stays flat, you're in your Optimum Effective Frequency range.
When G begins to rise you're edging out of OEF. Change copy. Take a breather - let the market refresh itself. If the natural Purchase Cycle in you category is 6 weeks and the middle of your OEF range is 3.65, you only need an EF of 3.65 every 6 weeks, not every week. Buy sparingly!
The graphic above summarizes G over 12 weeks for a Fast Response brand. Notice that G bears absolutely no relationship to the cost of media or the number of persons listening. It simply tracks the impact delivered by your message.
Like many smart employees, GM/C reports to CICO, who, if pleased, grunts and says, "Well, nice job there, young fella. You may have a future with this firm."
In the next section, you'll (finally!) learn how to plan a local copy test and a local or network rollout.
Intro/a> DR Radio v Print How Frequency Builds Response How To Select Stations
GM/C = X versus CICO How To Test & Roll Out Click-to & 1-800 DR
TECHNICAL NOTES:
Direct Mail pros will recognize the G is the reciprocal of standard percentage response. G = 1.600 is the same as .0625% response, but it's a little easier to read. G = 7.835 is the same as .000127632418.
Print buyers will realize that GM/.C also works for magazines or newspapers. Just substitute "Circulation in 000's per Inquiry" for G.
GM/C + TM = X. If you want to include Telemarketing costs in MCPS, determine an average [phone time + TM time & commission] per sale. TM training and initial set-up costs, like creative & production, are properly an overhead cost of business and should be amortized over time.
Single stations can give you a post-buy analysis that plots average cume frequency or cume P1 frequency at whatever points in time you choose. It takes them weeks, sometimes, to run the math, The OEM method is much easier. You can do it every day in about three seconds.
