Brandsensations

What do you see as the hottest trends in marketing technology today?

I think one of the hottest trends is social network ad spending. According to eMarketer, advertisers worldwide will spend $23.7 billion on paid media to reach consumers on social networks this year. Also according to new figures from eMarketer, a 33.5% increase from 2014. By 2017, social network ad spending will reach $36 billion, representing 16% of all digital ad spending globally.

Currently the hottest social networks for marketers are Facebook (now dominating the paid social advertising landscape globally), Twitter (expected to take up 8.8% of global social network ad spending), and LinkedIn (4.2% global social network ad spending last year). When looking at future predictions, I believe that marketers need to look at Instagram ads. Instagram ads are managed and purchased through Facebook’s ad platform so Facebook’s extensive targeting and custom audience platform can be applied to Instagram users–a very easy as well as powerful and potent transition for current Facebook advertisers.

Think about the last time you tried to solve a jigsaw puzzle, or Sudoku, or a crossword. It felt a lot different than UX design, right? Well, it shouldn’t.

When someone gives you a puzzle, you assume there is a solution, and you try to figure it out.

The concept is simple.

You get some rules. You get a bit of information. Then you work on it until you have all the right things in all the right places.

You will look at it different ways. 

You will try and abandon different approaches.

You will look for evidence about whether you’re right or not.

If something seems to be working, you will do more of it. If your approach isn’t helpful, you will step back and re-evaluate.

And puzzles have built-in feedback! If you put a puzzle piece in the wrong place, nothing else will fit. Or if you have the wrong answer in your crossword, it will make the other questions impossible to solve.

But you always assume it can be solved. Who would make a puzzle with no solution, right? (Psychologists would, just to fuck with you, but that’s a topic for another day…)

So eventually, you finish it. Good for you!

That’s the Puzzle Mindset™.

****

Lazy designers treat UX like a menu.

They choose from what they can see, instead of actually being creative.

When a restaurant gives you a menu, you pick from the options that are listed. You might compare the options with each other or you might look for something that’s trendy or familiar. That only considers what you can see.

If some designers did Sudoku the way they design UX, they would write zeroes in every square and then be proud of how nice it looked. Just rows and columns of the same number. Organized. Clean. Minimal. And totally pointless.

Then they would post it on Dribbble and get 100 likes.

****

The solution to a puzzle is not a matter of taste.

Lazy designers want to “decide” when the puzzle is finished based on how they feel about it. Like a 2 year old.

Unless you also throw temper tantrums in the supermarket and put toys in your nose out of curiosity, you can do better.

What’s the point of doing a puzzle if you’re going to ignore the rules and the clues?

(In this metaphor, the rules and clues are the users!)

When you make the text smaller because it looks better, or when you hide your menu behind a hamburger icon so your hero image is cleaner, or when you add a parallax effect because “holy shit, that looks awesome!”… you’re filling your Sudoku puzzle with zeroes.

It doesn’t solve anything, it just looks nice.

Too many designers approach UX this way.

But that’s not even the big problem.

****

Your mindset determines the next step.

The real problem is when you fail.

With a Puzzle Mindset™ your failures are just a step toward a solution. You will assume that you failed. You won’t assume that the puzzle is impossible to solve! 

You’ll just try something else. 

However, if you’re reading from the menu and you don’t find what you want, you will blame it on the menu. You will assume it can’t be done.

That’s a problem. A big one. Because you will give up, choose whatever is closest to what you had in mind, and move on.

The real solution might never be designed.

****

In real life, design is a job.

We are expected to deliver something. And the Menu Mindset makes it easier to look busy and ignore the results. People leave you alone, and the paychecks keep coming, am I right?

But if you want to build a reputation that opens doors in your career, gains the respect of your colleagues, and makes users more effective: treat UX like a puzzle.

Solve it. Every time.

The linked article is good and bad. And the image that goes with it is all “how” and not even a little bit of “why”. Thinking about data-driven design as a step-by-step procedure can be dangerous. It ignores the design-thinking of science.

****

It’s true that most A/B tests are useless.

Almost every company runs A/B tests when they shouldn’t. That’s the main point of the article, and they’re absolutely right about that. 

Maybe you’re curious about something random, or you’re just trying to settle an argument between sales and the developers, or maybe you imagined a bunch of different use cases and you’re not sure which ones actually matter (i.e. - fictional personas).

In real life, those are all probably going to give you results that are “inconclusive”. In other words, it doesn’t matter whether you use the A version or the B version of the design, because neither of them makes a useful difference.

However, there is an assumption built into that, and the article misses it too.

****

A/B tests don’t design themselves. YOU design them.

Which means — if you’re getting mostly useless results — you are doing useless tests. There is nothing wrong with A/B testing! You’re just doing a shitty job of it.

I rarely get inconclusive A/B tests. Every test has a hypothesis. Every version will support a different conclusion.

Like, maybe 2 out of 10 won’t answer my question, and in those cases the inconclusive result was one of the possibilities I was checking for. It means something if it’s inconclusive! And my second test pretty much always gets a result I can use, because I can avoid repeating what the first test disproved.

If you’re not testing a specific hypothesis, then your A/B test is a waste of time. You’re hoping, not testing. Even if you get a huge result, you don’t know why. 

If you are testing a hypothesis, then even an unimpressive result is new information!

If your hypothesis is that people find it hard to read two columns, so engagement will be lower in a two-column layout than a one-column layout, run an A/B test and find out! That’s a good hypothesis. 

If your test shows that both layouts are roughly equal for keeping people on the page, then your hypothesis might be wrong. You either need to design a new experiment to test it in a different way, or come up with a new hypothesis about why your Time-On-Page is lower than you would prefer. 

But if you learned that your hypothesis might be wrong, that’s important information!

See? If you do the science properly, you get useful results all the time.