Tag Archives: statistics

Real Numbers

As I said in this post and followed up with this post…when you do a survey and then publish the numbers, the numbers become real. When people are shown statistics, very few people will question the methodology that was used to get the numbers.

The survey/poll/etc says “X”…..so “X” must be true.

The power of published numbers is so strong that sometimes people don’t accept the real results. There’s a losing candidate in Florida who is considering a legal challenge to the election results because in his words….

“In this election, the results did not match the…pre-election poll, our internal polling, or our exit polling.”

The results didn’t match your surveys? So…the real numbers are wrong?

Of course, this is politics and there are other factors….but lots of businesses fall into this trap as well.

Surveys and research are good for getting a feel of the market. They’re good for testing the waters. They are not hard facts. Don’t get caught in the trap of thinking your research is the truth. You may be surprised.


Wanamaker finally gets it

–In stocks, it’s “Buy Low. Sell High”.
–In weather, it’s “not the heat, it’s the humidity”.
–In advertising, it’s “I know half my advertising is wasted, but I don’t know which half.” It’s a quote attributed to John Wanamaker who owned a department store in Philadelphia in the early 20th century. In one of my marketing keynotes, I throw his quote up on the screen and tell my audience if they don’t know which half is working, they’d better stop advertising because they’re wasting their money, time, consumer attention, brand equity, etc, etc.

Marketing is sometimes seen as an art, (which it is to an extent), but businesses take that too far. Are you too far? Apply your marketing philosophies to another aspect of the business…
–“I know half my customers are not paying me, but I don’t know which half!”
–“I know half my employees are stealing from me, but I don’t know which half!
If it doesn’t make sense in the other parts of the business, it doesn’t make sense in marketing either.

The trouble is tracking. Most companies (especially smaller ones) don’t track their marketing investment at all. Most that do track are doing it wrong. It’s no wonder that half (or more) of the marketing doesn’t work.

That’s why I’m encouraged by this news from Nielsen. Beginning this November, they will not only report TV program viewership, but also advertising viewership.

People have been preaching the death of the 30-second ad for some time now…(some better than others). I have always agreed that the change was on the horizon, but knew that old habits die hard for business. We’re now at the tipping point.

Today in the early 21st century, you have more tools and more opportunity than ever (and it’s growing everyday) to find out “what half of your marketing is working”. Why does your marketing still reflect a quote about a department store in 1912?


Blogs as a Marketing Research Tool

Alex asks the following question in the comments of my recent marketing research post….

“I know it’s not hugely scientific, but I’m sure with a little imagination a decent blog could provide businesses with very useful focus group style feedback.Comments from existing customers, passing traffic, a few people you direct to the site could all prove to be valuable.Is this plausible or is online research always a little dubious?”

I think it’s important to note the distinction between feedback and research. Feedback is fabulous. Businesses can gain a lot by just listening to customers. (But, you wonder why so many don’t.) I think feedback can come from anyone and everyone including your acquaintances, customers, and even your blog readers. Research, on the other hand, should be done with blind, random samples that take out most, if not all, of the factors that could taint the results – the essential feature in providing the custom assignment help. Feedback is something you can say and carry around in your head. Research is hard numbers on paper that you can use to prove points and make decisions.

Using online methods to do marketing research has a few problems….

1) You automatically have limited your sample to customers who not only have computer access, but are also computer literate/savvy enough to take part. If you’re reading this blog, you’re probably pretty savvy, but when I speak to groups, I still have to explain to intelligent successful businesspeople what a blog is. I have a friend who calls me all the time to explain to her how to insert a column in MS Excel. In many ways, we’re still on the left side of the adoption curve.

2) The audience you’re gathering feedback from is an important factor. For example. a software company developing a new program is likely to prompt more blog feedback than an auto manufacturer developing a new line of pickup trucks.

3) One of the problems with feedback from blogs is that a vocal minority could override the opinions of the majority. There are great masses of people scanning blogs everyday, but how many comments/trackbacks are there? Even the A-listers don’t have very many comments/trackbacks as a percentage of the actual traffic. When you rely on blog feedback, you’ve silenced a large majority of your readers.

In addition, the “blog as a focus group” model breaks some of the basic rules of focus groups. It’s not random…the participants are self-selecting themselves and already have a positive bias to you. And technically, you’re moderating the group by posting topics and leading them. It would be like the manager of a Pizza Hut grabbing customers as they walked in the restaurant to do a focus group about why people like pizza.

Is all online marketing research dubious? It depends. It depends both on the audience you’re trying to reach and what you’re trying to find out. If both those things fit into an online model, I think you’re fine. But the largest factor in any marketing research project, online or real world, is the methodology. Research HAS to be designed well with qualified random samples.


Lies, Damned Lies, and Statistics

bad marketing research exampleI see alot of misplaced gusto for marketing. Some of it comes in the form of bad print or broadcast advertising. And while bad ads or bad media buys are not as effective as they could be, at least they’re reaching some audience. They’re not causing the business any harm.

But one of the biggest marketing mistakes that businesses make that CAN cause harm is doing marketing research incorrectly.

Now, marketing research is GOOD. It’s one of the best things you can do to make sure that you’re getting the most out of marketing. The problem happens when people assume that they can quickly throw together a survey or focus group and get some data to work with. There are many little things that can be missed when designing a marketing research project that will drastically change the results. You need to have some idea of what you’re doing or hire a reputable firm to do it for you.

I’ve personally seen some really bad practices….

There was the business owner who wanted to do a “focus group”. He invited his friend, his friend’s wife, and some other people he knew. And then, he personally moderated the group. Shockingly, he was very pleased with what the group had to say.

There was the doctor who mailed out a massive survey to find out the age distribution and other demographic data in his city. He was shocked when I told him that his tax dollars were already being used to pay Census workers to gather this exact information and much more for him and that he could fully access the data.

I’ve seen hundreds of comment cards that were either designed to prompt an incorrect response or too confusing/long to provide useful feedback. The big sin with comment cards is usually the return methodology. It’s either a box located in the midst of employees who you’re commenting on…or makes you put a stamp and your return address on it.

And I’m sure you’ve seen numerous examples of one of the worst mistakes now happening in marketing research…the web survey. You do NOT have a good sample of people who come to a website to conduct accurate research. Take results from a web survey with a grain of salt unless your business is completely online.

The big trouble is that when you “do research” and then publish the numbers, it becomes real. When people are shown statistics, very few question the methodology that was used to get the numbers. Bad decisions are made with bad information. Then there’s a backlash of not trusting research…which should not be the case.

The truth is that marketing research is quick and easy to do wrong. It’s expensive and/or time consuming to do right. And it’s too important to mess up.


Black Friday

I’ve heard the fake statistic too many times today already not to post this…

The day after Thanksgiving is NOT the biggest shopping day of the year.

You hear this alot from people you know…as well as lazy news outlets who don’t check sources because the day after Thanksgiving seems busy with everyone clamoring at 5am for the next Baby-Wets-Sock-Em-Tickle-Me toy. But it’s just not true. While everyone may be out, they aren’t buying.

As far as retail sales go, the 2 weekends before Christmas show the largest sales figures and the highest sales day is usually either the last Saturday before Christmas or December 23.

Don’t believe me? Believe Google. Search http://news.google.com/news/search?q=%22biggest+shopping+day+of+the+year%22
or here’s a screenshot if you’re not searching on Friday http://twitpic.com/r6s85