Business Tech: Marketing - Part 2
Brought to You by International Spectrum, Key Ally, and Readers Like You
When I give a hot dog vendor money, I expect a hot dog. Money spent on marketing rarely has that sort of an unambiguous return. Sponsoring events, buying air time, peppering the Internet with click-ems, these are all ways of getting the message out. How do we know if the message is being received?
The short answer is that we don't. The long answer involves metrics, which is our NEW AND IMPROVED topic on marketing for this issue.
Now Enhanced with Punctuality
I had a boss who had no real idea how to measure my effectiveness as a programmer. When review time came around, I found the conversation revolving around the metrics he could apply:
Was I punctual?
Was my work area neat?
Was my vacation time scheduled enough in advance?
How many sick days had I taken?
Nowhere in my review were the more relevant questions, like:
How many million dollars did the company save due to my efforts?
How many employees where retained due to my morale boosting?
The included questions were ones that were as direct as, well, buying a hot dog. He picked things where my relationship with the result was clear.
The omitted questions were fuzzier. While the company did save millions, the credit has to be shared among a lot of people. Was my part .1% of the solution or 99.9%? Marketing is almost always fuzzy.
Did sales go up because of the Halloween ad, or because we added three new resellers that month?
Did our spokes model contest raise our sales, or was it the price drop?
Conversely, we have to ask if dips in the sales projections are due to marketing. Is it our sponsorship of Outside Underwear Day, or just the malaise in the economy fueling our downturn?
The old saying is that experience is what you get when you don't get what you want. In that same vein, metrics are what you use when you can't measure what you want. Since we can't measure the exact value of each marketing dollar/euro/what-have-you we spend, we fall back on metrics.
51% of All People are in the Majority.
One of the better metrics tools is to get the customer to tell you why they buy. The most direct way is to ask them. Those older folks among us remember when survey calls were relatively infrequent, and likely to get answers from us. Nowadays, I find myself feeling annoyed when asked, just because so many people are asking. I get them on all five of my phones, in the stores, on the street, and by e-mail. It has become too much.
So, what is a marketing person to do if asking isn't received well? Motivating people to do things is a core marketing skill, so marketing motivates you to respond. See how many of these types of motivators you've been targeted by this week:
Fill out our questionnaire for a chance to win an iPod.
Every completed form will be entered into our monthly drawing.
Free subscription, if you qualify.
Your response will be rewarded with a $10 gift card.
Say "You Heard It on WKAI" and the refills are free.
Putting our IT hat back on, we need to start thinking about how to support these self-defining techniques. Of course, gathering data is of limited value unless we are prepared to present it, so we need to look at reporting, dashboards, cubes, and other analytic output.
The big question is "How?" If the questionnaire is web-based, data collection is easy, but data validation is now in IT's hands. If it is on paper, then collection is a bit harder on us, but at least it will provide jobs for clerks.
We can't look at those as the only options. Maybe the form needs to live on a touch screen. Maybe it needs to run on my Android mobile. For marketing, these are logistical questions, which are secondary to the content of the questionnaire. To IT, these are fundamental questions which need to be answered first.
BOGO Knows
Another way to get customers to self-track is to offer encoded coupons. Offer a code of 1002NYT in the February NY Times and a code of 1003EM in the March e-mail and you know which marketing piece brought the customer to your door.
Dynamically generating coupons lets you uniquely identify each coupon's origin. I give numbers MV0001 to MV0900 to the people who flyer cars in the parking lot and I give the next 300 numbers to the folks who hand out papers on Eighth Street. Now I know that putting them on cars got a 2% return, but street hand-offs got 10% despite my giving out far fewer pages.
If I apply the same thing to the web, I can link the on-the-fly dynamically generated coupon code to the Google Analytics — or other data — that I had at the moment that the customer acquired the coupon.
Being able to connect the dots makes our metrics progressively more useful. We are still in a fuzzy place, but we are de-linting as fast as we can.
Limited Time Offer
Since the best metrics involve incentives, and incentives reduce revenue, we are left with the worry that we are marketing ourselves to the poor house. Remember that discounts are given to the people who qualify, not just the people who would not have bought without prompting.
When I sell a $100 item on a 20% margin and I give 10% off to a customer I would have already gotten, I've cut my profits in half. Conversely, if I make the sale because of marketing, I make $10 I would have missed. So, marketing by discount risks leaving money on the table.
Another problem with marketing by discount comes when we train our customers to stop buying. How many of us have had this conversation:
"I'm going to buy a new TV."
"Don't be stupid, wait until next week, there should be a holiday sale."
So, when you get people get used to discounts, you may be rewarding them for delaying purchases. One way to reverse this trend it to reward volume, not timing. Good examples are:
Progressive discounts: Save up to 20%: 5% on orders up to $50, 10% on orders over $100...
Pay a one-time fee of $99 and get free shipping all year long.
Join our savings club and earn more discounts the more you shop.
Bring a friend, and you each get a discount.
Limited time offer.
I was in Florida recently and overheard this conversation:
"Why did you go to that restaurant?"
"I had a coupon that was expiring."
Fair Use
In all fairness, sales are not the only reason for marketing. Coca Cola doesn't just put ads to sell soda. They also do it to create a more esoteric thing: brand value. I used to work for a company which made ties with the Coke logo on them. People didn't buy them because the ties tasted like soda, they bought them because the brand itself had a value separate from the product. So, while we have been talking about sales based metrics, there are many other kinds of fuzzy factors to consider.
The problem with branding is that marketing is often too successful. Pass me a Kleenex and I'll lay back in my Easyboy while I point out that brands can become so identified with a product category that the manufacturer has trouble defending their ownership of the name. Xerox had to run ads at one point stating: You can't Xerox a Xerox on a Xerox. But we don't mind at all if you copy a copy on a Xerox® copier because the name was in danger of falling into common use and ceasing to be a controllable brand.
Between too little measurement of success and the danger of succeeding too well, marketers have their work cut out for them.
OUT OF BUSINESS Everything Must Go!
All good things must come to an end and this topic, but wait there's more! Next issue, we will talk about modification. It will change the way you look at reading columns.