Content Marketing is twenty-first century biz-speak for collateral materials. The term includes any media intended to help sell a product or service. From old-fashioned press releases to Twitter memes, they take advantage of the low cost of Internet distribution and targeting by micro-market. Does content marketing work?
Before you fall for the breathless description of the magic of content marketing, it’s worthwhile to look back to the days before the web made the printing press obsolete. In history is the antecedent for almost every category of content. There is a set of communications that for a hundred years have worked for both producers and consumers. With the rise of radio and TV came a new set. The web has spawned a few new ones, but not so many completely new forms.
- Specification sheet
- Parts list
- Business card
- Usage story
- White paper
- Advertisement (standalone)
- Read advertisement, endorsement, and sponsorship
- Paid placement
- Outbound call center (later robocalling)
- Twitter ad (140 characters)
- Meme (postcard)
- Webinars and their recordings
Content marketers imply that things have changed in the Internet age, that consumers have a new appreciation for sellers’ collateral materials, and that marketers can measure the contribution the content makes to sales.
Certainly, commercial content helps in various degrees. Consumers read it or watch it or otherwise consume it with various amounts of satisfaction. All of it, to some extent, contributes to sales. To imply that consumers enjoy this stuff is a stretch. Some of it, perhaps. Most consumers would not choose to spend spend free time reading specification sheets or parts lists. Watching ads for pain relievers is the price we pay for watching broadcast television.
At one end of the scale is producer-sponsored content people will spend their own time to consume.
Pillsbury covered the table with its Pillsbury Bake-Off, a sponsored television show, a series of cookbooks, and recently, the winning recipes on its website. These stories feature the use of Pillsbury’s BEST flour.
Arm & Hammer uses writers to spread stories of baking soda in hundreds of household uses. These stories have been standard fare for decades.
Procter & Gamble Productions introduced As the World Turns and The Edge of Night in 1956. Soap operas have supported advertising campaings ever since.
Mutual of Omaha’s Wild Kingdom debuted in 1963. It is probably the only reason people recognize the insurance company’s name. Its success is indisputable.
Specification sheets, industrial catalogs, and instructions are crucial to commerce, but most people wouldn’t say they enjoy reading them. When they do they need them, they need them now.
It seems consumer catalogs do capture the imagination of consumers, or they would all end up in the junk-mail trash. They linger and do stimulate sales. Sears Roebuck named its Christmas catalog The Wish Book.
At the bottom of the list is content that the company would likely not produce, nor the buyer consume, were they not required. Examples include Material Safety Data Sheets and limited warranties. In most cases, this content is required only after purchase of the product, so one would have a hard time arguing that they stimulate sales. Without them, perhaps, sales would be illegal, but that’s not a marketing question.
What we are really interested in: Advertising
All of those other forms of content are just noise as far as the marketer is concerned. The marketer wants to know which ads are effective. He wants to know which ad results in sales when shown to which consumers and when. The problem is, in the ad business everyone is a salesman. Most of what they tell each other is bullshit.
The smart analyst didn’t study marketing. She studied biostatistics. She learned why case series, longitudinal studies, and cohort studies don’t appy to advertising. Proof is only found in a randomized control trial. Show the same ad differing in one aspect only to two groups selected at random. If their buying performance is different, more different that one would expect as a matter of chance, then the campaign is effective. All of the other study techniques, case series and prospective cohort studies are flawed pretenders. Marketers know the randomized control trial as an A/B test.
A typical A/B test compares two different headlines promoting the same copy. Half of readers are presented at random with the first headline, half, the other. Then you measure which one gets more clicks. If the sample and the difference are large enough, you’ve proven one headline better than the other at getting clicks. The science is indisputable. The researcher’s question, though, is dubious. Sure, the better headline got the user to click. It is a big jump to assume that reading the article is a predictor of buying the product.
The magic word in online marketing is conversion. As a general matter, it is the movement of a user from one stage of the buying process to another. The most important conversion is from non-customer to customer. A more mundane conversion is from page reader to button clicker. When you listen to a presentation including the word conversion, listen carefully for the implication that a conversion is a sale. Often you will discover it is something less meaningful, a request for information or some other activity without a dollar value.
Members of the Content Marketing Institute report website traffic as their leading measure of campaign effectiveness, over sales and quality of leads.(Pulizzi 2016) It’s unclear whether is is more difficult to measure the sales generated by a web page, or whether website traffic is simply a happier statistic to report.
Cynics will tell you that crappy researchers report the findings that show the most astounding results, not the ones that that prove a conjecture. Increasing web site traffic is not the same as increasing sales or profits.
Website designers make plenty of claims about the purposes of a site: “customer engagement”, to automate customer interactions and the like. All of those claims are supported by high traffic numbers. The person paying the bill for the website will tell you that the goals are simple: (1) to attract more clients, (2) to retain existing clients, and (3) to reduce the cost of serving clients. The first and second goals are advertising. At the end of the day, measurement of website success is not traffic, but sales.
If you want to prove that your content marketing campaign is working, you need to capture the identification of each bit of content a user consumed. Then you look at sales to that user. You set the machine learning algorithm loose on the data. The first question is the one that any experimenter asks, “Can we reject the null hypothesis, that buyer behavior is unaffected by the specific content consumed or the amount of content consumed?” Assuming that buyer behavior is affected, then we want to see what content made that happen.
If you don’t hear these kinds of words from someone selling you content marketing, swipe left and move to the next vendor.
Pulizzi, Joe. 2016. “B2B Content Marketing – 2017 Benchmarks, Budgets, and Trends—North America.” Content Marketing Institute. Content Marketing Institute. https://contentmarketinginstitute.com/wp-content/uploads/2016/09/2017_B2B_Research_FINAL.pdf.