7 comments posted · 1 followers · following 0
nice post! As someone running a reasonably successful independent agency, I'm clearly a bit biased, but agree that there seems to be a huge gap.
Two other forces which influence pricing:
One price driver is offshore development - that drives down prices for small, simple projects like iphone apps where the project objective can be really clearly described & communicated. To be fair- in your example of $160k maybe you're not comparing apples with apples. It sounds like there was a lot of definition required, and then the client was able to suddenly define it (perhaps completely differently) to get a cheaper price.
The other pressure is around CMS style sites. Not so long ago, big agencies could build static CMS sites for $200k+ as you've outlined. Today, I can buy a theme for $30, load it in to Wordpress, add my content and I'm done. It looks pretty good, I can get an HTML developer to customise it if I need, and it saves me having to come up with a design brief - I just pick the one I like.
This puts pressure on pricing if all you're delivering is a nice looking static site with CMS abilities - which is often what big "campaign" style sites are at the end of the day.
For the actual test, we've used xOs - http://www.accenture.com/Global/Consulting/Market...
We also use Google site optimiser for simple tests, but that has the disadvantage of having all the variations inside the Google tool (so it is harder to build and manage)
I was wondering if you had any opinion or data on whether delivery of video via a CDN had a negative impact in terms of SEO.
From what I've been told, there are potentially two issues:
1. The content isn't from my server
The page content is coming from myserver.com, but the actual video content source is on another server. So the "google juice" for that video content goes to the server holding the video rather than the site.
2. The video content comes from different locations
Using a CDN the content is often served from different locations. I'm in Sydney Australia. Sometimes the video I'm watching is getting served from Sydney, while other times it is coming from Hong Kong or San Francisco. This is a problem if geographic location is important for SEO efforts (which it is for us - we make sure that servers for a particular country are located in a datacentre in that country)
However, from a development point of view, using a CDN has a huge benefit- particularly for video content. So, I was wondering if you'd run into this issue before and had any comment or thoughts on the matter
I've just checked- seems to be all OK. Make sure you "save as" root.fla
to answer your first question, about how it can look: there is no connection between the ruby bit and the way the front end looks. I'd say this is the case for most modern frameworks, with the exception of a few that force particular classes or IDs in HTML - which can be limiting.
re: scaling - it depends on exactly what you mean by scaling. We moved to Rails as we saw this to be an advantage. Our largest Rails experience to date has done ~4 million unique people a day, and there are plenty of large sites and applications that do a lot more than that in terms of people visiting.
Scaling might also mean being able to handle an ever increasing bunch of transactions. Twitter has had some very public scaling issues, but from my understanding this seems to be to do with their database architecture rather than the framework.
Apart from there being lots of things to consider (db architecture, web & app servers, etc), there are also design issues such as how much of your page needs to be dynamic vs static, and how optimised your page is.
Hope that helps
yes, our comment code was a bit unloved- sorry for the hassle. I've replaced it with a shiny JS version.
To reply to your comment - I think it is more that many of the new testing tools and ideas come from the "Ruby crowd", and so are quickly adopted by other people working in that space. Take Cucumber (http://cukes.info/) for example. The irony there is that it plays with the other kids like java, .net etc.
I agree having quite a few automation tools "in the box" helps a lot, and that almost all documentation refers at some point to testing (so it is pretty hard to avoid). But you could argue that the MS .Net environment comes with lots of automation tools, without the same kind of thing (embracing tests and automating more) occurring. I suppose one difference is that Rails encourages a developer to at least try to have a poke around under the hood, while .net (for example) is more "use this GUI - you don't need to know any more".
But then again, maybe we're just totally lazy...
I think a more useful thing at this stage is to document how it might behave - eg: the text "includes meals and drinks" appears in the field, but disappears when you click in it.
We've found things like default state, values etc are best put directly in to a prototype. Depending on the development framework, things like what's in a dropdown become less relevant as they can be sorted out in the data model (eg: in an MVC framework like Rails)