Great call on Themeforest. The added benefit of using a prefab admin skin is that you can actually build your entire concept before you need to think about paying a designer. Many of the themes have a built-in grid system (if they aren't using 966 outright) so that you can build UI workflows after creating mockups. I have made this mistake in the past but would now never hire designer if I hadn't already built a working prototype of every screen, and for that I use ThemeForest HTML5 skins.

As multiple designers will attest, I am more than happy to pay for design work (nothing we have up publicly came from Themeforest or 99designs). But unless your concept absolutely depends on graphical experience --- and look how shitty Pitchfork's site was for years and years and years in the trendiest most superficial demographic imaginable --- graphic designers are a terrible thing to block your project on.

Rather than wrack my brain and carefully come up with a scoping document laying out all the pages and functionality in my next application --- a document that will be out of date exactly one customer meeting after I write it --- I'm just going to point my next designer at a $12 Themeforest theme and say "do this, but way better".

I don't mind that it costs $5-10,000 to get "better than Themeforest", but I do mind losing weeks of release time sweating whether it's time to pull the trigger on a designer and what exactly to have them do.

99designs is a different story. I don't have the designer's moral problem with spec!work but the misgivings people have about quality and IP infringements are based on real issues. I still think it's a good tool, but I can see why some would avoid it.

Avoiding Themeforest though seems crazy.

The designers' aversion to spec work seems more like an attempt at collective bargaining than a true moral stance to me. I'm not saying that collective bargaining is necessarily a bad thing, but the success of 99designs and the like makes me think that they won't succeed in this age of globalization.

Exactly. The article presents tools for minimal viable products for which Themeforest make sense. It is only if the project gets traction and eventually money that the logo and theme should be worked out more seriously.

I recently got an item from themeforest (just a launching page). Even though it looks rather fancy, modern, etc.. (and I could never have made anything that nice looking by myself), I've found that integrating it into my project and tweaking it to be just right has taken far more time than developing something decent from scratch.

This doesn't square with my experience at all. You must be much better at HTML/CSS than I am (not at all unlikely).

For my part, it takes me roughly two hours to:

* Copy all the static assets to their Rails-appropriate locations

* Take the "demo" pages and get them to display reasonably as Erb templates of a Test controller

* Convert the HTML to Haml

* Move content to layouts/application

* Break each feature in the demo out to a partial

There is no other two hours I can spend that gets me as much UI functionality, so I'm a fan.

I have done this three times, and each time found the work of integrating a Themeforest template to be easier than integrating work from graphic designers.

I am careful to work with templates that are HTML5, CSS3, jQuery, and that use a grid framework. I can imagine there are many hundreds of themes there that fail one or more of these checkboxes, and that might thus be a nightmare to work with.

Cosign, I find it takes me the same amount of time to fold in a full theme into an ASP.NET MVC site using Razor. It's all just markup in the end, and whatever angle-bracket tax you're paying for server-side includes. Normally the most time intensive part is replacing paths to statics in CSS and HTML.

new to HN so not sure who sees my replies or who don't, but I should've clarified - what you're saying is what I agree with, in terms of prefab themes.

it's just the spec work that gets me. that stuff isn't being sold multiple times like prefab themes. college students do spec no matter how much I yell at them and then end up (in terms of time spent) making so little an hour that it's just atrocious.

but, time and a place, etc.

and the type of work I do is very front-end heavy.... tends to go from mvp to beta launch quickly... and we have designers on staff. so, just spoiled. bad bias!

I am advocating prefab themes.

The price point for a very solid HTML5/CSS reasonably-semantic gridded admin theme is $12 to use and $600 for unlimited license. Even the license is a tiny fraction of the cost of any design project.

I am very specifically saying that samey prefabness isn't a liability for web app UI; in fact, I think UX outcome is likely to be superior, since samey prefab admin themes that people will like enough to pay for are all the Themeforest designers tend to build. They're developing palpable expertise.

The liabilities of these themes --- for instance that they are highly non-optimized --- are things that (a) do not matter until you have lots of customers and (b) are not particularly expensive to remediate with contractors after-the-fact.

The liabilities of the DIY approach --- often, a devil's bargain between either waiting weeks or months to get a pro designer to pick up and complete a project, or going to market with a manifestly poor design --- are much higher.

I think for most app devs, not using Themeforest is an expensive, irrational point of pride. It's not good business.

This is a fairly recent revelation for me, so I'm being noisy about it.