One experience should not fit all.

If you set out to craft the perfect user experience for everyone, you have likely already failed. Perfection is a MacGuffin. There is never a single experience that satisfies every user. And trying to be all things to all people generally leads to being nothing very great for anyone.To use a non-digital analogy. Say you’re a clothing designer charged with crafting the next hot-selling men’s trousers. Some men swear by trousers held fast by a trusty set of suspenders. Other men are firmly dedicated to the use of a reliable belt. But neither faction is too keen to purchase any garment fitted with a belt and suspenders combo solution.As silly as our fictional trouser solution sounds, we often come across digital properties that attempt to provide anything and everything any visitor might want, which results in the same desirability as the belt and suspenders combo.

Maximization requires sacrifice.

It’s an easy trap to fall into—thinking that increasing options for people will increase your chances of converting that person into a customer/sale. But it’s far more likely in most cases that with every additional choice, you’re diminishing the potential to positively engage that person. It’s simply an example of Hick’s Law in action.Hick’s Law says that for any individual, given a number (n) of equally probable choices, with each additional choice (+1), the resulting complexity of the decision increases the average time (T) required for that individual to actually make a decision, geometrically.

T = b * log2 (n+1)

In plain English: you’re not doing yourself, your customers, their decision-making process, or your business any favors by increasing the complexity of choices within your navigation. The smart decision is to reduce the available user journey options down to those most desirable to people already predisposed to convert/buy. By proxy, that means diminishing or eliminating those journey options tailored to those who are outside of your optimal consumer. To paraphrase a favorite adage of political strategists, you shouldn’t waste time trying to change anyone’s mind. You should focus on getting the people who already support you to actually vote.

But how do you know what to sacrifice?

Even after careful consideration, incorporating user needs and feedback, the only real way to understand if you have designed an effective UX is by digging into your analytics.

Start by looking at the relationship between time on site and page views.

This won’t tell you if you’re winning the game, per se, but it’s a start to know whether you’re playing on the right field. There is no “best” quadrant here (though one comes close to being “worst”). Depending on the purpose of the site, what constitutes a positive visitor relationship varies quite a bit.Quadrant A, for example, is a positive indicator if your goal is to satisfy a user’s propensity for exploration (think: online fashion), but it’s a negative indicator if the purpose of your experience is to satisfy a visitor’s need to find a single piece of information or transaction, quickly (think: online banking site).

Quadrant B is a positive indicator if your goal is to satisfy a user’s desire to delve deeply into a single topic (think: academic journals), but it’s a negative indicator if the purpose of your experience is to drive impressions (think: advertiser supported news sites or blogs).

Quadrant C is a positive indicator if your goal is to satisfy a user’s desire to complete single tasks quickly (think: filling out applications at online lenders) but it’s a negative indicator if time spent is an indicator of value (think: education or tutoring).

Quadrant D, for example is a positive indicator if your goal is to satisfy a user’s need for a quick answer (think: Google), but it’s a negative indicator for just about every other purpose.Those are extremely limited examples of what types of businesses or sites logically reside in which quadrants. You’ll have to take time to evaluate which quadrant is most appropriate for yours.

Focus on your key performance indicators (KPIs) and ignore everything else.

Assuming you’re driving proper time/views statistics for your business, a logical data point to look at next is whether the people you’ve corralled in your quadrant are actually converting in a way that positively impacts your business. Are shoppers actually buying? Are potential borrowers applying? Are your indicators actually pointing toward increased performance?Understand that gains in areas that do not directly impact success aren’t gains at all. Increases on overall traffic are irrelevant if that traffic doesn’t consist of users who convert. Conversions don’t matter if those conversions don’t lead to sales, etc.

TL;DR—Strive to please some of the people all of the time.

In art circles, you’ll often run across the opinion that anything great is ultimately polarizing. As I’ve posited above, the sentiment certainly applies to UX design. The trick is knowing what customers and which of their behaviors will have the greatest impact on your bottom line and focus almost exclusively on satisfying them—even if it means disappointing everyone else.