Almost all supermarkets share a common layout. Many competitive products (toothpaste, for example), have similar packaging. Do supermarket and toothpaste companies lack imagination?
It’s possible, but a different explanation is more likely.
Smart businesses apply science to marketing. Relying on psychological research, these businesses adapt marketing strategies to maximize revenues and profits. When companies unlock the innermost secrets of how and why people buy things, interesting patterns begin to emerge.
For example, there’s good empirical data showing the best times and days to send marketing emails to maximize opens and click-through rates. However, as people have grown to more heavily use mobile devices, the science of email is gradually evolving. New research suggests, contrary to conventional wisdom, that many brands can benefit from sending email campaigns at night.
How can you apply scientific wisdom to improve marketing for your business? Let’s look at two approaches.
1. Let data drive your decisions.
Many marketers develop campaigns based on intuition. Guerrilla marketing campaigns fit this mold. A marketer believes, based on experience or a “gut” feeling, that a stunt might work, and they invest time and money to execute it.
Similarly, landing pages are often designed based on aesthetic look and feel, not on their ability to optimize user conversions. Paradoxically, the best looking designs are not always best. Sometimes, aesthetically better designs simply don’t convert as well.
In contrast, marketing as a science looks to optimize campaigns and marketing tactics to maximize returns on investment. It has become easier and more practical to apply science to marketing because marketing technology has exploded. For example, smart companies routinely A/B test landing pages in an effort to optimize conversions.
A number of years ago, for example, major publishers were losing print subscribers and wanted to find ways to convert print subscribers into digital subscribers. Many experimented with the decoy effect, also called the asymmetrical dominance effect. The decoy effect occurs when people tend to have a change in preference between two options when a third, asymmetrically dominated option is presented.
One of the best examples of the decoy effect was an old subscription page of The Economist.
The first option at $59 seemed reasonable. The second option at $125 seemed expensive. The third option offered options 1 and 2 (web and print) for the same price.
Dan Ariely, author of Predictably Irrational: The Hidden Forces That Shape Our Decisions, tested this phenomenon with his MIT students. When presented with all three options, zero students chose option 2. Most chose option 3. When the second option was eliminated, most students chose option 1 (online subscription only).
Data can be very useful, as the above example, shows, but can have its own biases, as The Harvard Business Review cautions:
Data and data sets are not objective; they are creations of human design. We give numbers their voice, draw inferences from them, and define their meaning through our interpretations. Hidden biases in both the collection and analysis stages present considerable risks, and are as important to the big-data equation as the numbers themselves.
More importantly, as Albert Einstein famously said, “not everything that can be counted counts, and not everything that counts can be counted.”
Tips: To optimize how you use data to help you make decisions, you must ask the right questions and focus on the relevant data. For example, if you’re wondering why or when your customers are leaving your site, consider what data you have that can help you answer those questions. You can look at customer complaints, payment history, the funnel customers follow when browsing your site, poor customer service experience, frequency of usage, etc.
2. Create and execute controlled experiments.
Do you remember having to write a hypothesis for your science experiment in school? If you have school-age kids, you’re probably helping them do this now.
The goal of a hypothesis is to help explain the focus or direction of the experiment. A hypothesis is a prediction.
An experiment is structured as follows:
- Formulate a hypothesis
- Design and execute an experiment to prove/disprove the hypothesis
- Analyze the results
- Accept, reject or refine the hypothesis
Experiments can help you apply lean marketing principles to conduct quick, low cost tests to develop and scale your marketing strategies.
Let’s walk through an example. A few years ago, our friends at Basecamp wanted to test various design concepts for one of their software products, Highrise. Here’s what they wrote:
We have assumptions about why some designs perform better than others. However we don’t know exactly why. Is it the color of the background? Is it the headline? We hope more iterative testing of the winners will help us get that information. If you have any theories please add them in the comments.
The team created a variation on their original design and A/B tested that variation. They found that the newly designed long form had a 37.5% increase in net signups compared to the original form. That’s a terrific improvement in conversions.
One very useful place to run experiments, if you run an online e-commerce business, is your pricing page. People have long assumed that customers want more choice. It turns out that for most customers, that assumption is wrong. People are more likely to purchase when their choices are limited.
Tips: If you’re struggling to figure out what to test in the first place, we recommend you read this helpful post from Optimizely – 71 Things to A/B Test. We use Optimizely at crowdSPRING to help with our A/B and multivariate testing and recommend them. Another good option is Visual Website Optimizer.
How do you feel about marketing as a science? Do you believe marketing is mostly art or science?
