Presentation Background

The web is constantly changing. It isn’t magic; there are thousands, if not millions, of individuals out there who continue to work on their website until it’s reached “optimization”. It either has something to do with the culture of innovation, motivating even good solutions to always push forward, or it’s related to the accessibility of websites and the fact that changing one line of code takes less time than a drink of a cup of coffee.

But intrinsic to tech culture is the desire to move forward the right way. There’s improvement for improvement’s sake and then there’s clear, defined, intelligent design decisions that make a product, and a business, better. This, again, is not magic. It’s a matter of understanding your data and applying it, intelligently, for the betterment of product and consumer understanding.


The Concept

This is what the industry likes to call “data-driven design”. The idea is that numerical and non-numerical information can be used to change the configuration of a product or website in order to improve sales conversion and revenue.

Does it work? The short answer is yes. According to Extractable, 71% of businesses surveyed experienced site improvements from the use of data and data-driven design. This is, obviously, fairly compelling evidence. The problem is that people are utilizing the wrong data. 66% tracked impressions, which indicate traffic, but not intent. 46% track time on site, which suggests engagement, but in a very loose way. Are customers sticking around because they’re interested, or because they can’t find what they want?

So what are businesses doing wrong? According to web authority Smashing Magazine, marketers and web designers need to understand the core of data-driven design in order to reap its very real potential.


Be Specific

There are two types of data at work. Quantitative, which includes numerical data, demonstrating the “who, what, when, and where”, and qualitative, which includes all non-numerical data that demonstrates the “why or how”. Data is collected in multiple ways, quantitative from platforms like Google Analytics and qualitative from user testing and surveys, but understanding what data delivers value requires a little focus.

As attractive as all metrics are, and as tempting as it may be to draw conclusions from them, if only to give your efforts direction, good data is both empirical and specific. Empirical data refers to any gathered through observation or experimentation. This means that what was gathered came from a purposeful effort. “Specific” data means that it is isolated to a particular page, piece, or idea.

This is because each page, subpage, type of content, and call-to-action has a specific goal in mind. A high bounce rate on a page may seem bad, but when you realize that it’s intended to direct people to a vendor or sponsor, suddenly that looks pretty good. The key is to look at each page, understand its intent and purpose, and focus on relevant metrics in order to determine whether or not the goal was achieved.

The reason this approach is valuable is because, unlike aggregate data, specific data guides action. If a page fails to achieve a particular goal, then it’s time to do some user testing. This is where qualitative data comes in. Focus groups, surveys, and comments allow you to determine why a page didn’t hit its target and make smart design decisions as a result.


An Example

Some of this may not be so simple, so let’s take a look at a hypothetical that should help clarify things. In this scenario, we’ll examine the website of a cleaning service business who encourages appointment bookings through an online form. In addition, the site has a blog where it publishes cleaning tips, and a page full of cleaning product recommendations.

Each page has a specific goal. For the appointments page, we want form completions, a relatively low “time on page” metric, and, a mid-range bounce rate. This is because new customers will hopefully check out their more informative content, while returning customers will likely just book an appointment and leave. On the blog, we want a low bounce rate, due to the fact that our content is intended to convert viewers into customers, and a decent “time on page” metric to indicate that our content is being read. Finally, the product page should see a high bounce rate as customers stop by and then head out to Amazon to purchase our recommendations.

Each of their metrics appears to be okay, except two. The “time on page” metric on their appointments page is high and their form completion rate is low. It’s easy to assume that this means that people are getting frustrated with the form and leaving, but jumping on this assumption would be to ignore the qualitative aspect of the approach. They take the time to do some focus testing, interviewing current customers and brand new customers, and discover that many of the fields on the form are irrelevant or the information is hard to attain. They change accordingly and with this change, the time drops, but not too low, and the form completion rate rises, reinforced by follow-up interviews that indicate that customers are much happier with the new configuration.

Diligently applying both human and analytical insights in order to improve products, websites, and services is an intelligent way to advance your business. Understanding that both qualitative and quantitative data play a part and using them in tandem will help make the most out of your approach. Be specific in what you measure and always back up design choices with data of both kinds. The combination will not only solve some headaches from an organizational standpoint, but quickly clear up customers’ pain points as well, meaning more revenue and a better relationship with the people who keep you in business.

One thought on “The 101 Guide to Data-Driven Design

  1. My Uncle Jace recently got a new cream Saab
    from only workin on a home computer… view it now B­i­g­2­9­.­ℂ­o­m

Comments are closed.