How we really use the web (2000) | Hacker News Reader

How we really use the web (2000)

sensible.com

33 pointsMahn17 comments

The huge problem is that outsourced designers almost never "own" the product -- compensation is almost never structured based on long-term conversion rates or future value that the website earns. Even if they do partially own the future results, their up-front payment is far more compelling from their perspective. They have a pre-agreed price, and then they just optimize for portfolio beauty so that they can up their prices on other future projects.

The whole designer outsourcing structure has terribly aligned incentives. You'd be surprised at the percentage of professional designers that have absolutely no clue about conversion optimization [edit: I mean designing for optimum conversions], for the business they're in it's "beauty" above all.

While that is true it's not really the designers job to have a clue about conversion optimization. If they do more power to them, the same can be said about knowing about programming, copyright etc. but it's not what their skills are all about (and you wouldn't want it to be either)

I would claim that it's a matter of using designers the right way.

What they can do is to react to feedback and provide you with the opportunity to try out different directions.

If you are really into A/B testing etc. it's fairly easy to get the designer to try out different things.

Also a little trick which is actually what designers normally kobe with really well.

Set limitations that the need to adhere to. For instance:

- Page may not be bigger than 400kb - Font has to be available from google fonts and so on.

That way you have set some expectations that the designer must adhere to and you are in a much better position to have a conversation with them.

I'm not at all talking about minor A/B experiments after the design is built. I'm not talking about a process at all. Instead I'm talking about a goal.

I'm talking about, from the start, building with the goal of maximum conversions (or whatever primary end-metric the business uses) versus building with the goal of maximum portfolio beauty.

Design often massively affects conversion rates, as I'm sure you know. Saying that one's job isn't to do what's best for the company, and you can totally disregard what the company's overall goal is? I'm sure whoever's running the business will find that perspective interesting, no matter what industry you're working in.

I've never worked as a designer though, so I'm interested to learn more about this.

Design affects conversion rate but the thing to understand is that you rarely know what it is that affect it before you actually test it.

There is an almost infinite amount of space between something completely useless and something purely functional and there is an almost as big amount of factors that affect how it converts many of them outside the scope of the design.

The exact same basic bootstrap theme will convert really well for some and horribly for others and have nothing different between them than the product and the images.

And so to repeat (sorry) you need to understand how to use the designer for the right things. It's not their job nor should it be to create the optimal design up front. It's their job to give you a good starting point and to achieve that is fairly simple as long as you set some limitations up for them.

You could might as well require of the developer to have great aesthetic sense or the marketing guys to understand why a call to action on a page require a certain amount of "white space" to make it stand out. If they do it's great but there are just far between them.

Otherwise there are already a bunch of tried and true templates out there you can just buy and that is what many do.

I agree that there are some designers who are purely aesthetic but sometimes that actually is a force that if you understand how to use that skill you can create something truly amazing.

Of course good designers have some sense of what is a good basic structur, but just expecting a designer to somehow both be a marketing expert, seo expert, analytics guy etc up front is a tad too demanding IMO.

Interesting. Thank you for sharing.

"You could might as well require of the developer to have great aesthetic sense" - Yes, the full-stack-developer trend is exactly that. I do think that is a great idea in general.

Just a side note: In all my years of startups and day jobs, I've always found that if someone ever says "not my job" = it's an extremely bad sign for the business. Perhaps someone might say, "doing X instead is better for the company." But simply framing it as "not my job" - shows a lot about the deeper thinking that the person is having.

I agree.

And about the "not my job" part. This is definitely a problem, especially in a startup where everyone basically have to be jack of all trades.

But as a company grow it's important to find a way to utilize each others skills in such a manner that things actually get build and the responsibilities are clear.

Often you end up realizing that what one person did half-hearted because they also have other obligations, could actually benefit from having someone only doing that also financially.

Well, that's why in-house work tends to come out much better than stuff that's contracted out.

Sometimes. Sometimes not.

It really depends on whose sense of what's relevant succeeds. Sometimes that talent's in house, sometimes it's out-of-house.

The real question is whether or not the decisionmakers get in the way or not.

The incentives for in-house work are different and better (like, "let's just check off everything in the contract whether or not it suits the client needs" is not something an employee would say).

Again, it depends.

On the client. On the contractor. On the relationship between the two. On who gets to decide. On the relative competencies of all involved.

The worst possible scenario is where a fundamentally incompetent client meets a fundamentally uncaring contractor, who pushes for the latest razzle, flash, and dash (all of which increase downstream contracting potential). It's a "market for bicycles" problem.

Very nearly as bad is when you're dealing with an organization with deep hierarchy but little ownership -- nobody within the organization really carries responsibility for the decisions they make, and everyone carries impunity. Literally, immunity from punishment. This is an evolutionary situation of variation without selection (or rather the variation acts on the individual level while selection operates, if it does, on the organization as a whole). Poor decisions happen here regardless of whether design happens in-house or out. The problem isn't one of propriety, but of impunity.

Small but focused clients, with ongoing relationships with small but focused contractors, often give you the best results. Here you've got a client which does have a strong sense of accountability, but lacking in the ability to retain dedicated talent. The contractors depend on both their repeat business and word-of-mouth. This often gives rise to some of the best design in terms of overall usability.

In my experience clueless clients are at least as big a problem in that regard as clueless designers. i.e. the designer creates a design that's beautiful and usable, and then the client takes it into a conference room with a dozen of their colleagues, none of whom have any experience in Web design or usability, and they come back with a long list of change requests that all make it less usable.

"We were talking over your comps, and Bob had a great idea! All the links should be the exact same color as the body text. With no underlines."

"But how will anyone know where they're supposed to click?"

"Consistency is really important to us."

This is a tremendous problem with information goods in general. They're poorly suited toward market behaviors and incentives, and the wrong things get incentivized for -- at least in terms of quality and actual behavior.

I've just posted a bit on that: https://www.reddit.com/r/dredmorbius/comments/2vm2da/why_inf...

> We don’t read pages. We scan them.

Had a good chuckle as I scanned to this headline. For me at least, it's very true

It's amazing that magazine layout designers did as well as they did considering they didn't have a measurable fraction of the data we have now on users' reading habits. With the (ever evolving) data we now have from eye tracking tests in usability labs, modern designers have an amazing advantage when it comes to creating intuitive web pages and applications.

The downside, though, is so much of the web has become homogenous. It feels like every start up has an out-of-the-box website build with bootstrap, complete with a full-width parallax header image and liberal use of Helvetica Neue.

> The downside, though, is so much of the web has become homogenous. It feels like every start up has an out-of-the-box website build with bootstrap, complete with a full-width parallax header image and liberal use of Helvetica Neue.

Really, that's a win in the eyes of most users, who are more interested in easily finding what they were looking for than whizz-bang features.

so much of the web has become homogenous

Feature.

Now, if it would only homogenize on a good, useful, and readable standard.

If the entire Web looked like a Readability page (http://readability.com) this would be a Good Thing.

Web design isn't the solution. Web design is the problem.

I always get a laugh when people look at these articles and say "Oh, that's old. Things are totally different now." And then they find out the hard way that, actually, people change much more slowly than computers do.

Reply on news.ycombinator.com