Design too begins with grand theories. Theories of aesthetics, assumptions of what others will like, and predictions of what will succeed in the marketplace. While the quest for great product experiences might be data, it’s worthless without starting with human empathy.

Marissa Mayer, recently appointed CEO of Yahoo.com, used to be the VP of search products and user experience at Google. Mayer was one of the key leaders behind many of Google’s key products. It’s well-documented that Google loves data and relied on engineers and data to make design-related decisions. For Google, design is a science. In this FastCompany.com article, Mayer shared how Google uses data to cut through the political nature of subjective design choices (like color).

Microsoft also sees design as a science. Behind the colossal failure of the KIN, data informed every step of their development—ignoring what the real world would want. This is a classic case of inside out thinking; logical and provable on a spreadsheet, but not necessarily true in the real world.

Unfortunately for Google, Microsoft, and the many other analytically-run companies, this practice of using data to drive innovation and inform decision-making does not necessarily ensure success.

It’s really a matter of process and timing.

Successful products must first be fueled by user empathy and intuition. Data is best utilized to inform refinements after the fact—once you collect information about real world usage.

The biggest process challenge to designing successful Web sites comes from the two extremes. There’s the “damn the facts” camp that lets ego and individuals dictate what is good design. Then there’s the “all-data, spreadsheets = reality” camp where decision-by-numbers creation almost always falls flat. I’m sure data-only thinking mitigates risk, but it does more than mitigate progress.

These two ideologically-opposed approaches only get it half right. Bringing intuition and analytics together in the right order makes all the difference. Microsoft bragged about the mountains of data and research they did while creating the KIN. But it was a case of the cart leading the horse. Apple relied on intuition for building a better mobile phone when producing the first iPhone. Steve Job’s smart phone market research boiled down to his belief that “they all sucked.” But it was data that forced the notoriously stubborn Steve Jobs to change course and allow outside software developers to create apps for his iPhone that fully realized the $100 billion dollar eco-system.

At Wondrack Design, there are a few ways we successfully bridge this typical chasm of left- and right-brain thinking. One before launch and one after. Both are tried and true methods for managing successful design.

1. Rapid prototyping
A method borrowed from product design. This method makes it relatively easy to bring ideas to life enough to analyze their impact and anticipate what the final user experiences and outcomes may be.

There’s a major flaw in research-first thinking that’s obvious to designers. While there’s no hard evidence attributing this quote to Henry Ford, the message is ever true: “If I had asked people what they wanted, they would have said faster horses.” If a company relies solely on research to innovate, they’ll never really achieve innovating thinking.

2. A/B testing
A method extended from direct marketers for testing versions of a Web page, experience, or element to gain statistical insights into key performance indications. It may be hard to measure design, but A/B testing makes it easy to test design’s effect on a company’s key performance indicators.

Designing is a process focused on making something better. It follows the arc of human interaction with technology. It’s best served by intuition that empathizes with the user combined with leveraging data that informs improvements.

Update Dec 12, 2012: Google redesign efforts show us how they do it.

Recently, the leader of a large redesign project proclaimed that speed was the most important criteria. “Done is better than perfect” was posted on our project room wall. Cute. As a professional who has launched over 150 sites, this reference to speed made me cringe more than playing soccer against kids three grades above me. This redesign project felt like a twelve- to fifteen-month project but we accomplished it in six. The speed was a challenge and many team members got burned out along the way.

But why gripe? The site redesign launched as planned and is a consensus improvement over the previous site. Though I wonder, does speed equate to success? Will the site’s quantitative goals be met when it has been designed by hurried and “consistently inconsistent” decisions?

At the start, a pile of research was done, a strategy was developed, and leaders were full of good intentions. Designers were relegated to skin wireframes and push a lot of pixels. What happens when designers are not given access to copywriters to develop personas that fill the site’s product pages? The task to procure over a hundred photos was more akin to playing a game of Minute to Win It than actually building visual coherence.

In the end, the quantity of decisions made and work delivered was impressive. But how good were those decisions? They weren’t tied to real business metrics, conversion goals, and the like. Rather, they were made with one thing in mind: speed.

Getting it done fast made people feel good—accomplished. Deadlines were met. But getting through it was difficult and came with a cost. Will the new site increase conversions, decrease customer service calls, or increase cross-selling leads? Does it set the brand up in a new, more effective direction? Since decisions were not tied to metrics like this during the design process, it’s difficult to say.

At the end of any worthy game, everyone knows the score. I’m eager to see what the score will be compared to the cost of all this blinding speed.

Sound familiar? They should. These conventional intentions are worthy end goals and evident in many successful experiences we have online. Unfortunately, I’ve found that most redesign projects struggle to stick to criteria that have the greatest impact on simplification and ease of use.

To ensure you end up with a site that meets these goals, spending more time defining your intentions—and how to measure them—goes a long way. If you don’t, you open yourself up for a messy end product that underperforms. Here are some typical culprits that prevent a successful redesign:

1. Expectations: making things simple can be surprisingly hard, frustrate the impatient, and undermine strategic decision making.
2. “Feature-itis:” the phenomena of piling on features to fend off misguided fears of being perceived as simplistic. In my experience, more features rarely makes for a better experience. The right features make it easier to use, while too many make it harder.
3. Ego: individual tastes trumping facts can undermine decision making that’s useful for everyone else.
4. “BS:” contradicting input, unnecessary politicking, lack of specificity, and unrealistic project management (among other compromises) will usurp your project.

These derailing aspects will undermine the process for building a refined user experience, creating simplicity, and determining your ultimate performance factors.

Setting the right criteria begins with debunking some assumptions. Designing Websites is not the same as developing an ad campaign or designing spreads for a brochure. The criteria for simple and easy-to-use Websites are experiential factors—not just creative, visual, or technical factors alone. And success takes leadership that understands.

You’ve all heard the phrase different answers depend on different questions, right?

Successful advertising campaigns answer the question: What perception do you want the audience to have when they see this?

Successful graphic design answers: What do you want the audience to think when they see this?

Successful Web design answers: What do you want the audience to do when they see this?

Things that are more interesting get read. Things that are more interesting to look at get noticed. Things that are easier to use get done.

How many times has a new internal application within your organization, like time tracking, only made it harder to do what you wanted to do? Easier-to-use means it’s more enjoyable to use, less invasive, more intuitive, quicker, and yields what it’s designed to do. Usability of applications or Website is not hard to measure if you build it into your project plan and initial definition for success.

Design, as a visual art, is more difficult to measure. However, the results design yields can be quite apparent when you frame the right expectations and measurements up front.

What’s your current site yielding? Is it really simple and easy to use for your potential customers? Can it be improved to yield greater returns? My bet is yes. There isn’t a business site, e-commerce site, registration path, or call to action that wouldn’t benefit from being looked at, measured, tested, and ultimately made simpler to achieve greater results.

Websites and advertising are different mediums; best staffed and managed accordingly. Remember (circa 2000) when the Pets.com sock puppet ads won lots of awards, then the company liquidated ten months later? In the end, the campaign’s creative brilliance solved the wrong problem. Like any great campaign, Pets.com focused on getting your attention; however, it was at the expense of building a successful user experience. Amazon.com never won any major creative award, but has done something much more valuable: earned the trust of millions of customers and generated billions in revenue for more than a decade running. Unlike Pets.com, Amazon focused on creating the user experience that you, the customer, prefer. The page design is far from perfect. They aren’t as interested in awards as they are in delighting users. So does that mean less creative wins? Depends on how you define it.

I believe creativity follows the same rules as humor—both depend on incongruities to succeed. To paraphrase John Morreall, author of Taking Laughter Seriously, running into your next door neighbor at Heathrow Airport would make you laugh because it’s highly unexpected to run into him so far from home. Seeing your neighbor pull into his driveway any weekday evening isn’t funny because it’s expected. We judge creativity in a similar way. If it’s an obvious pattern, we discount it. If it’s different and breaks a pattern, we celebrate it. (Jackson Pollack would have made a horrible Web designer.)

If you accept the incongruence theory for creativity, I’ll assert it’s a misguided standard for making good decisions related to Web design. While humor can be effective in getting people’s attention, it’s not for creating user experiences that build trust, meet specific expectations, and measure performance. This pitfall of applying the same creative parameters behind effective advertising to Web design assumes the same (creative) tasks carried out by the same (creative) talent. But they’re different: good advertising gets attention, persuades, and drives traffic. Good Web design empathizes with and engages users, and converts traffic into sales. The creativity involved in building a site’s architecture, user flow, and a successful user experience is more akin to product design vs. advertising. The creative challenge is finding new ways of celebrating the obvious and actually supporting patterns. When it’s successful, “it just works.”

The criteria you set and manage for creativity throughout the Web site creation process ensures your site’s success. Criteria focused on getting attention and novelty will underserve the user experience (like Pets.com). Follow the Amazon example—focus your criteria on empathy, engagement, and analytics to create a successful user experience that generates results.

It’s not the budget or message that undermines more advertising, email, or promotional campaigns; it’s the underperforming site design. Despite conventional thinking, success is not measured in hits, but in sales conversions, increased productivity, or other metrics.

“A black cat crossing your path signifies that the animal is going somewhere.” (Groucho Marx)

Too often, people read into things, make illogical predictions, assume outcomes, or bet on unfounded hunches when there’s evidence to guide decisions. Consider this recent experience my company had with a retail client’s path to increase their site’s income from one of their products. Average monthly figures: 1,200 hits and $300 net income from a conversion rate of 6.8 percent.

Initially the client engaged an ad agency to implement an advertising-centric strategy to increase hits. The ad campaign (and very minor page copy changes) successfully increased hits by 32 percent. At the same time the conversion rate dipped to 4.7 percent, resulting in a net loss. The client then considered doubling their ad budget. Fortunately they didn’t and turned to us for help instead.

Our experience has helped us to recognize these patterns when working with clients:

1. Objectives are based on assumptions (e.g., hits = sales).
2. There’s a default to budget for advertising (vs. fixing the user experience on the site).
3. Decisions are guided by incomplete data (e.g., measuring the wrong things).
4. Most sites underperform (e.g., inefficiencies are being ignored).

To break the client’s “spend and hope” cycle, we helped them look at their site metrics differently. It started with how objectives were set and pursued. The result after revisiting their stats and conducting A/B design tests (without advertising): 1,300 unique visitors and $1,200 in profits. Conversion rates jumped to 10.4 percent.

In a nutshell, this client originally succumbed to the pitfall of measuring the wrong things and tracking superficial metrics like hits. Such metrics make people feel good because they are easily accomplished, but they rarely achieve the ultimate goal. This incomplete site strategy suffers from what author and Inc. contributor, Norm Brodsky, may call the sales mentality: pursuing sales without minding more important factors, namely profits. Our approach was to focus on actionable metrics, like conversion rates, to employ a profit mentality.

Not only was there financial gain from our work, it was a more cost-effective endeavor than advertising. Building a better page design with SEO-friendly code is a one-time expense. Performance is not dependent on ongoing media budgets or retainers.

There’s no denying that people enjoy voicing opinions, but that can be costly when making design decisions. Each doctor and engineer has deep expertise in a narrow discipline, but just an opinion outside their field of study. Our deep expertise, guided by Design Analytics, helps organizations make better design decisions about improving site performance to achieve specific goals. Our profit mentality can save you from misguided objectives and incomplete information that would otherwise undermine your pursuits.

On the other hand, Google spends a radically large amount of time, resources, and energy on quantifying everything, including design. They famously tested forty-one shades of blue to determine their hyperlink color. They view the world as a pile of data to be analyzed. Google’s view of the design process is empirical and akin to most MBA decision-tree curriculum. But these two approaches, subjective taste and analytical study, are incomplete on their own. To paraphrase Jim Collins (author of Good to Great), the “Tyranny of the OR” limits each of these design approaches.

The case against intuitive decision making (subjective taste) is that it’s a prediction based on assumptions and people are notoriously bad predicting events beyond their control. (Just ask any oddsmaker. Our egos seem hard-wired to believe otherwise.)

Alternatively, purely analytic decisions are void of empathy that affects how people feel and respond toward things. Also, analytics are reactive. It’s hard to have effective insights without enough data to benchmark and test. Google’s forty-one color study had the luxury of billions of online transactions to measure, compare, and analyze.

The key to successful and effective design is to combine the two—ala Jim Collins’s “Genius of the AND“—in the right order. A good UX designer can anticipate user needs and create a more enjoyable experience, but must rely on analytics to find out for certain and fully realize the goal.

The only way to predict the future is to invent it. That takes courage, and a tuned sense of what people will want without asking them. To ensure your goals, it also takes the discipline to leverage the readily available data to measure the results of your design choices.

Quantifying design may be difficult; however, combining intuition and measurable outcomes to inform better design decisions will enable you to predictably achieve your goals.

If you want objects designed to decorate your company’s identity, that’s easy thanks to today’s software, the internet, and eager design school graduates. If you need to design a way to improve your sales performance, or reduce waste in your bottom line, that’s something else. I’ve noticed that whenever design becomes visual, it becomes trivial, superficial, worthy of trite contests to determine the best-looking solution. Note the most recent case from the White House—even irony can be ironic apparently. Plumbers, doctors, and accountants would never provide services for free with the hopes that you’ll like what they did enough to pay them for it after the fact. The thinking that this is okay in regards to design exists because clients ask, and designers and agencies abide.

The tragedy in this thinking is it undermines design’s real purpose: to improve how something works. The late Steve Jobs said this about Apple’s most misunderstood success factor: “People think it’s this veneer—that the designers are handed this box and told, ‘Make it look good!’ That’s not what we think design is. It’s not just what it looks like and feels like. Design is how it works.” This active view of design is the kernel of a ridiculously profitable company with a market value over $350 billion.

Most of you haven’t had the experience of working at a company like Apple, and some of you may never have purchased Web design services. That’s ok. When buying Web design services, effective buyers set specific and measurable success factors, like increase conversion rates from 2.3 to 4.0 percent, and give the designer the opportunity to design ways of achieving them. Effective project briefs avoid useless adjectives like “easy-to-use” or subjective terms like “clean.” They skip unrealistic feature wish lists and statements like “just like Apple.com.” An effective project brief is an honest assessment of the buyer’s current situation filled with information about what makes his or her business succeed and user data to support the project. Effective buyers understand that Websites that succeed have to be relevant to target users rather than what the CEO thinks after looking at it for a minute.

Regardless of any one person’s opinion, a clean-looking Website like Apple’s won’t ensure your success. Designing a better way for your customers to understand your product and an easy way for them buy (like Apple has) will. These differences in managing design are quite distinct, like the differences between nouns and verbs.

Advertising great George Lois is known for proclaiming, “Great ideas can’t be tested. Only mediocre ideas can be tested.” Lois helped spark the creative revolution of American advertising that has since fed the egos and lined the pockets of many smart, clever ad men and women. The mid-20th century formula for success was the Big Idea plus reach and frequency. A world where you could reach a high percentage of your audience in several key placements and the audience trusted your claims.

I’ll spare you op-ed about traditional advertising’s decline and the significance of post-Google advertising, or how measuring success by GRP is incomplete at best. (Quick note: roughly 96 percent of Google’s $29 billion revenue in 2010 came from advertising). This article isn’t about media buying strategies; it’s about challenging the notion that great ideas can’t be tested.

How did you test ideas in 1959? Well, we know George Lois didn’t have the amazing tools we have today. He lived in a world without tracking software, Google Analytics, or Websites like KISSmetrics, Clickable, or Kickstarter. His clients didn’t have the ability to automatically capture real-time customer shopping behaviors or track conversion metrics of multiple concepts, headlines, calls to action, button colors, button locations, etc.

The best they had included their two-martini-lunch-filled guts, accounting, and agency reviews. Don’t get me wrong, mid-20th century advertising in NYC was some of the most innovative thinking of its kind, though it’s marginally meaningful in today’s post-Google world.

But apparently, not much has changed in regards to testing ideas. What I’ve learned from working with certain ad agencies is they’re scared to death to test live ideas. When pushed why they won’t run a simple A/B test on a banner ad, defensive, irrational, and bogus reasons abound. One client tried to argue they couldn’t afford the $25 media surcharge for an A/B test. The media budget was over a quarter of a million dollars. Obviously there were other reasons.

I think it might be that testing ideas for banner ads, landing pages, or calls to action undermines the “agency knows best” mind-set. It dilutes the drama in the big pitch where the agency presents and sells the almighty Big Idea. Testing concepts requires the agency to park their ego outside and adapt a different problem-solving approach.

Agencies still do things no other business can do, and are capable of more than clients typically give them credit for. But it’s hard not to be critical, if not cynical, of the myopia agencies live in if they think their untested ideas are great. They’re not great until customers act on them and they help achieve the client’s objectives.

Like the radical ideas of George Lois and his contemporaries that changed modern business in the pre-Google world, new post-Google radical ideas are among us. Time to adapt and test their greatness, or watch them die.