5 Marketing Problems We Need To Solve Now
In the 20th century, marketing was a relatively sleepy endeavor. You came up with a “big idea,” sold it through the organization, shot some TV ads and put them on the air. It worked for the most part and brands became valuable assets, just like physical capital and technology.
Things started to change in the 80’s and ‘90’s. Cable TV created an explosion of channels and a massive fragmentation of audiences. Marketing was still about ideas, but became more numbers driven as issues like targeting efficiency and ROI rose to the fore.
Today, marketing has been altered beyond recognition, becoming more mobile, social and connected. The result is that marketers, rather than being a relatively monolithic group of professionals, have metastasized into a hodgepodge of specialists including designers, mathematicians and technologists. That’s creating altogether new problems to solve.
1. N = ALL
For a long time, the most common retort in marketing discussions has been “what’s your sample size?” followed by, “what was your methodology?” Whenever you’d encounter an idea you didn’t like, you could always gain the upper hand by attacking the data.
And the data always would be a small, supposedly representative sample of whatever we wanted to know about. We could glean insights by scaling the sample up, with the help of some statistical rules, in order to reflect the population, but if there was even a small amount of error, this scaling up would result in massive miscalculations.
In other words, our numbers were always wrong, sometimes off by a little, sometimes by a lot, but always wrong. Even if market research was done well, by the time a study was ordered, designed, conducted, reported and analyzed, months had passed, but we treated those graphs and charts as gospel (at least until the next study came out).
Big data is starting to change that. We’re now able to analyze entire data sets rather than mere samples. New technologies like the Web of Things are enabling us to collect data that we never deamed of before, including links in social networks, shopping patterns and conversations among thousands, if not millions of people (through natural language processing).
The result is that marketing organizations will have to fundamentally change the way they work in an “N=All” world where speed trumps brilliance. Strategy will need to become less clever and more Bayesian. Instead of “plan and execute,” we will need to execute, evaluate and revise.
2. Marketing Simulations
In the small data world, we dreamed up big, inspiring ideas. These were daring and brave, innovative and far-reaching. They were designed not just to “make the quarter,” but to transform the marketplace.
Unfortunately, our colleagues usually had big ideas of their own which were quite different than ours. So we would meet and argue about things like sample sizes and methodologies. Before long, the big ideas became progressively smaller and prudence would win out. Consensus would build around the safest course (usually a version of last quarter’s plan).
That’s beginning to change as big data and machine learning are enabling us to build marketing simulations. Instead of conference room battles, we can test our ideas in a virtual environment built by real world, real time data and if they work (the simulations are about 90% accurate), we can run with them.
In a simulated marketplace, we can dream bigger than ever. If the idea falls flat, all we’ll lose are some bits and bytes.
3. SEO for Social
The biggest single line item of 21st century marketing is search engine marketing. Just a blip on the screen in 2000, it has grown into a $20 billion business in the US alone.
However, as much as 94% of clicks are generated not by paid ads, but through “organic search,” the pages that show up when users search naturally. Perhaps not surprisingly, search engine optimization (SEO) has become a full blown science, with thousands of highly trained specialists dedicated to getting results ranked high on the page.
Social marketing, however, is still in the dark ages, with practitioners operating largely by instinct. As natural language processing becomes progressively more mainstream, that will begin to change. Companies like Networked Insights and OpenAmplify can quantify and analyze millions of conversations and provide actionable metrics.
If we are to join the conversation, it would be helpful to know what people are actually saying beyond what a fairly uniform group of social marketers are buzzing about. We now have the technology to listen effectively, industry practice needs to follow suit.
4. Marketing Integration
Back in the old days, marketers were a pretty monolithic lot. They were creative types who had an intuitive feel for the consumer mindset. Marketing was about ideas and so were they (well, their own ideas mostly, other people’s ideas were often deemed stupid and shallow).
Now, as I noted above, there is a dizzying array of marketing specialists and marketing organizations have followed suit. Take a look at the organization chart of any of the major marketing holding companies and you’ll see creative shops, data driven media agencies, digital guys, specialists for search, social, events and even marketing technologists.
The problem is that most of these people hate each other. Creatives believe that they “own the big idea” and often see everyone else as supplicant, media people think everything is “fluff” unless they can calculate the ROI of coffee cream and digital natives are exasperated that everyone else doesn’t just “get it” and write off their traditional skill sets.
These mindsets are as pervasive as they are destructive. Fortunately, there is a simple solution: No one should get promoted to VP or above without working in at least two functional or geographical areas. As long as executives are able to spend entire careers in safe, homogenous environments, marketing integration will remain poor.
5. The Skills Gap
By now it should be clear that there is an underlying trend to all four problems above: Marketing has become progressively more mathematical.
Unfortunately, for all the “by the numbers” hoopla, math skills in the industry remain atrocious. Very few have the skills needed to evaluate data effectively. There are no basic standards, little, if any, training on elementary concepts and, perhaps most alarmingly, very little awareness that math skills are important in a data-driven age.
Make no mistake, anyone who works with research needs basic math skills (e.g. the ability to use and interpret statistical functions in Excel). The fact that so few in the industry are familiar with material covered in an introductory statistics course is a huge indictment and leads to frequent misinterpretation of data.
And that is probably the biggest problem in marketing today. For all the conferences and webinars, fancy gadgets and neologisms, there is a very meager commitment to the basic skills that make up the human capital in the industry (writing skills are almost as bad).
If we want to change the world, we will first have to change ourselves.