Heat Map Testing - How should it affect modern web design?

I'm an in house UI/UX/Web Designer at my company. Before hiring me our company had an enterprise level web consulting firm on retainer that we are "letting go of" soon. We are still using them primarily for feedback at the time until our contract is up.

In the past, and particularly with some current feedback, this firm has referred a good bit to the importance of designing pages for the "F-pattern" that heat map tests have proven users tend to absorb web-contents with.

My question is, how important and relevant are those types of Heat Map Tests to today's web design? The tendency seems that our consultant agency thinks we should "play it safe" and stick to the research. I.开发者_开发技巧E. - build on what is proven.

While that seems sensible, I think it is lacking in logic from the standpoint that the Heat Map Tests and the "F-pattern" come from how users tend to read blocks of text, and not as much with the interface. And indeed, if you see a heat-map image where there is an image in the text, usually the image has a longer dwell spot whether it fits in the "F-pattern" or not.

So should the "F-pattern" dictate site layout and interface design, or does it refer less to overall design and more to content absorption?

The benefit of heat map analysis comes from understanding how people read your site. The "F" pattern tells you that English speakers read left-to-right, top-to-bottom. Duh. Heat maps will tell you if they can find the navigation controls, or if a side-bar distracts from content.

Jakob Nielsen wrote an interesting blog post about how people ignore content when it looks like an ad. The site he studied followed the "F" pattern perfectly and eye tracking data showed people saw the content, but didn't read it--they only scanned the first couple words and decided (wrongly) it was an ad. That would have been tough to figure out without heat maps of the actual site.

Use heat maps as an analytical tool, not as a design principle.

My non-actionable hyperbolic opinion is that it's a whole lot of hooey they use to convince you they're doing something useful in exchange for the obscene amount of money they're demanding. The art of composition and most other design principles do not lend themselves well to left-brain analyses, but unfortunately, nobody pays artists $300/hr to talk about their feelings at board meetings, so you end up with these quacks shilling irrelevant statistics the MBAs can pretend to understand instead.

The proof is in the pudding. If the design change increases your sales (or whatever you want the website to do for you) then keep it. It worked. If it doesn't, switch it back.

Unless the site design is a radical departure from a "standard" layout, it's tough to refute the evidence of how people tend to scan the page contents.

I think that going against human nature is a massive losing bet in the long run.

That being said, you should run your own studies on your own design and content. Get real data that you can use to support changing (or not changing) the design.

In my opinion heamap testing it's only one of many testing phases a project should go through for the analysis data to be perceived as proven. Each case should be treated individually, so you won't be caught red-handed creating copies of the same website. Furthermore, each business has different principles, and while generalizing is sometimes good while you're working on a generic site for a wide audience, it is imperative to perform multivariate testing (sometimes called "split-testing" or "A/B testing"), to make sure that you model your content structure accordingly.

Stakeholders hire experts to do this, but due to the fact that they're longing for appreciation and input into the work we're doing they might want to try and push their personal opinions on the subject, and therefore creating an abomination which is hard to control and doesn't generate revenue.

Do as many tests as you can before handing over the results to your bosses or stakeholders as this will be scientific data with which you can back up your statements and shut their mouths.

Heatmaps analysis is a powerful tool when used accordingly. But if you stick to old concepts that might become invalidated in few months time you'll be digging up a whole under your feet. Innovation is what drives this business and research is a good way of finding that golden ratio.

I'm using heatmaps only alongside other methods of testing. Rarely on its own.