[OT]Custom CMS vs Learning Existing Myraid
Aaron Cooper
I’ve always thought that this was a no brainer, but as I’ve been investigating future competition(for me) in the industry, focusing on the guys who are doing really well, I’ve discovered that some of the more prominent development houses maintain and develop using their own CMS over sourcing or training staff to become pros at one of the many Open Source CMS offerings like Drupal, Joomla, Expression Engine or Modx.
The pros of going open source are obvious:
1. Most of the hard work is already done
2. Security fixes are usually picked up by someone else, and fixed by someone else, not on your payroll
3. In most cases Development is continuous and swift
4. Support is plentiful and free
5. Chances are that a client request not included in the base code is covered by a module
6. May be possible to find staff with experience in the particular CMS
The pros of a custom solution are arguable depending on your familiarity with any OS CMS, and the size of your organization:
1. Flexibility – You know the code (but you still have to train others)
2. Total development control
3. Total IP ownership
4. Development can follow client feedback easily
I was hoping to kick off a discussion on this. Because to me, the pros for spending time and money becoming an OS solution expert (on one chosen CMS) far outweighs the pros for a custom one. I’m simply at odds to work out what benefit these companies find in offering their own CMS to clients.
I must say that the Flexibility of rolling your own is second to none. There have been times where a mod to Joomla has been far from trivial – but then I’m no expert in this CMS. Couple that with the fact that if I built my own, and I do end up working ON my business rather than IN it, I’m going to have to train PHP programmers on my fancy CMS in the same way I would have to train someone in Drupal, or find someone with Drupal experience.
So why would some of NZ’s most successful Web Development firms offer their own CMS to clients and at bargain rates too? (This whole wonder has come about from a friend having a site built by *bleep* and not only is the CMS custom, it’s mustard, and the bill was surprisingly cheap). And yes, as far as I can see, they use PHP.
Cheers
Aaron
Cameron Junge
I’d say tie-in.
Same goes for some Open Source products as well. I’ve used & done some sites in Plone… now it’s Open Source, and it’s a damn good product, but in NZ there’s not a huge amount of support for it, and when the hoster I originally used sold off their clients I’m now being told that the current hosters are screwing them with the hosting charge while not providing much support. Considering I don’t support the clients directly anymore, it makes me feel pretty stink that they’ve being screwed.
Use something that not many ppl know, and you’ve got a good tie-in. They’re forced to use you, &/or your hoster. Moving is a pain, they either need their site rebuilt or they need to find someone who is willing to pick up what their site uses.
Cameron
ph...@huntdesign.co.nz
are actively winning customers from several other web development firms
who used a custom CMS. In my experience very few of the custom CMS
solutions can offer the range of functionality available in the OS CMSs.
Several customers have approached us having been burned by the attempt to
tie them in on a custom CMS. Our selling point has been that the only
lock-in we have is to continually perform well, meeting or beating their
expectations. They like the openness and the lack of lock-in to their
data. We like the fact that customers stick by us because we are good at
what we do, not because we are holding their data to ransom.
On 12/12/2007, at 12:17 PM, Cameron Junge wrote:
I’d say tie-in.
Same goes for some Open Source products as well. I’ve used & done
some sites in Plone… now it’s Open Source, and it’s a
damn good product, but in NZ there’s not a huge amount of support
for it, and when the hoster I originally used sold off their clients
I’m now being told that the current hosters are screwing them with
the hosting charge while not providing much support. Considering I
don’t support the clients directly anymore, it makes me feel pretty
stink that they’ve being screwed.
Use something that not many ppl know, and you’ve got a good tie-in.
They’re forced to use you, &/or your hoster. Moving is a pain, they
either need their site rebuilt or they need to find someone who is willing
to pick up what their site uses.
Cameron
From: nzp...@googlegroups.com [mailto:nzp...@googlegroups.com] On Behalf
Of Aaron Cooper
Sent: Tuesday, 11 December 2007 11:36 p.m.
To: nzp...@googlegroups.com
Subject: [phpug] [OT]Custom CMS vs Learning Existing Myraid
I’ve always thought that this was a no brainer, but as I’ve
been investigating future competition(for me) in the industry, focusing on
the guys who are doing really well, I’ve discovered that some of the
more prominent development houses maintain and develop using their own CMS
over sourcing or training staff to become pros at one of the many Open
Source CMS offerings like Drupal, Joomla, Expression Engine or Modx.
The pros of going open source are obvious:
1. Most of the hard work is already done
2. Security fixes are usually picked up by someone else, and fixed by
someone else, not on your payroll
3. In most cases Development is continuous and swift
4. Support is plentiful and free
5. Chances are that a client request not included in the base code is
covered by a module
6. May be possible to find staff with experience in the particular CMS
The pros of a custom solution are arguable depending on your familiarity
with any OS CMS, and the size of your organization:
1. Flexibility – You know the code (but you still have to train
others)
2. Total development control
3. Total IP ownership
4. Development can follow client feedback easily
I was hoping to kick off a discussion on this. Because to me, the pros for
spending time and money becoming an OS solution expert (on one chosen CMS)
far outweighs the pros for a custom one. I’m simply at odds to work
out what benefit these companies find in offering their own CMS to
clients.
I must say that the Flexibility of rolling your own is second to none.
There have been times where a mod to Joomla has been far from trivial
– but then I’m no expert in this CMS. Couple that with the
fact that if I built my own, and I do end up working ON my business rather
than IN it, I’m going to have to train PHP programmers on my fancy
CMS in the same way I would have to train someone in Drupal, or find
someone with Drupal experience.
So why would some of NZ’s most successful Web Development firms
offer their own CMS to clients and at bargain rates too? (This whole
wonder has come about from a friend having a site built by *bleep* and not
only is the CMS custom, it’s mustard, and the bill was surprisingly
Jochen Daum
Cameron Junge wrote:
>
> I’d say tie-in.
>
> Same goes for some Open Source products as well. I’ve used & done some
> sites in Plone… now it’s Open Source, and it’s a damn good product,
> but in NZ there’s not a huge amount of support for it, and when the
> hoster I originally used sold off their clients I’m now being told
> that the current hosters are screwing them with the hosting charge
> while not providing much support. Considering I don’t support the
> clients directly anymore, it makes me feel pretty stink that they’ve
> being screwed.
>
> Use something that not many ppl know, and you’ve got a good tie-in.
> They’re forced to use you, &/or your hoster. Moving is a pain, they
> either need their site rebuilt or they need to find someone who is
> willing to pick up what their site uses.
>
Some comments on that from the viewpoint of someone who has chosen
Joomla for this exercise.:
- Customers are tied in with a system, but a rebuilt is smaller in cost
than you may think. We have rebuild many sites from proprietary CMSs and
this often is done when other changes are done or the web design/html
code (table to table less) is done.
I prefer to have clients stay with us based on the service as well.
However, there is also a certain amount of lock in when you offer lots
of solutions that other people say no to, because they would have to
hack the OS CMS.
Also, a lot of clients are looking for consulting and guidance in
regards to a CMS implementer, the chosen technology quickly fades in
this respect. This often comes with the unwillingness to learn the more
advanced features of a CMS and delegating this to a service company.
In regards to Camerons question:
> The pros of going open source are obvious:
>
> 1. Most of the hard work is already done
>
> 2. Security fixes are usually picked up by someone else, and fixed by
> someone else, not on your payroll
>
> 3. In most cases Development is continuous and swift
>
> 4. Support is plentiful and free
>
> 5. Chances are that a client request not included in the base code is
> covered by a module
>
> 6. May be possible to find staff with experience in the particular CMS
>
> The pros of a custom solution are arguable depending on your
> familiarity with any OS CMS, and the size of your organization:
>
> 1. Flexibility – You know the code (but you still have to train others)
>
> 2. Total development control
>
> 3. Total IP ownership
>
> 4. Development can follow client feedback easily
>
I woudl add 5.:
- You can easier productise your offering, i.e. offering standardised
templates and only minor changes to each installation necessary.
> I was hoping to kick off a discussion on this. Because to me, the pros
> for spending time and money becoming an OS solution expert (on one
> chosen CMS) far outweighs the pros for a custom one. I’m simply at
> odds to work out what benefit these companies find in offering their
> own CMS to clients.
>
I agree with this. You also ride on the wave of marketing that is done
by the bigger CMS providers in the market.
>
> I must say that the Flexibility of rolling your own is second to none.
> There have been times where a mod to Joomla has been far from trivial
> – but then I’m no expert in this CMS. Couple that with the fact that
> if I built my own, and I do end up working ON my business rather than
> IN it, I’m going to have to train PHP programmers on my fancy CMS in
> the same way I would have to train someone in Drupal, or find someone
> with Drupal experience.
>
> So why would some of NZ’s most successful Web Development firms offer
> their own CMS to clients and at bargain rates too? (This whole wonder
> has come about from a friend having a site built by **bleep** and not
> only is the CMS custom, it’s mustard, and the bill was surprisingly
> cheap). And yes, as far as I can see, they use PHP.
>
As I said above, they standardise the offering and can sell it quicker
andmore often.
Jochen
Nick Taylor
> FYI, all of my sites over the last few years are based on Drupal and we
> are actively winning customers from several other web development firms
> who used a custom CMS. In my experience very few of the custom CMS
> solutions can offer the range of functionality available in the OS CMSs.
A couple of times recently I've gone with Wordpress with custom modules
that I've written myself... because for a lot of companies/people, the
terrain is shifted and what they need is not a website with a news
section attached, but a blog with a website attached... and I'd almost
be tempted to do a spot of soothsaying and say that The Day of The
Website is passing in favour of one-off applications surrounded by a
network of active presences in misc social-networking sites.
Depending on what the business is of course.
Personally I'm no longer interested in writing CMS systems. They're no
longer interesting from a problem-solving point of view... and they have
a funny way of turning into never-ending, unpaid support jobs.
Nick Taylor
> A couple of times recently I've gone with Wordpress with custom modules
> that I've written myself... because for a lot of companies/people, the
> terrain is shifted and what they need is not a website with a news
> section attached, but a blog with a website attached... and I'd almost
> be tempted to do a spot of soothsaying and say that The Day of The
> Website is passing in favour of one-off applications surrounded by a
> network of active presences in misc social-networking sites.
And to derail my own post, and cross-pollinate it with a conversation
that's going on right now in one of my UK lists... this question arose:
"Which sites do you think a company should sign-up to register their
online presence/brand name? also to include sites that at a later date
could be useful"
An agregation of the answers being :
Myspace
Facebook
Google : analytics, gmail, video, blogger, feedburner
Youtube
brightcove
Twitter
Yahoo : Flickr, Upcoming
Amazon : associates
Linkedin
del.icio.us
Technorati
Vox
Digg
Hotmail
Dopplr
openID : can get one via technorati, vox, wordpress amongst others
Clangingly off-topic and all, but you might be interested etc.
Mark Rickerby
But you can't implement a unique design for free on any platform... If you're trying to directly deal with a particular requirement that a customer has, then oftentimes, you need to build something complete anyway, whether it is within a CMS or not - so unless you're already using that basic CMS functionality for other parts of the site, it's not going to be significantly less work having to custom code your design into the CMS framework than building it from scratch in a general purpose framework like Symfony or Cake. In either case, the work is in understanding what the customer needs and delivering a design that meets this.
Personally, I think that while there excellent products available that solve the problem well, there is still a lot of untapped potential in UI design for CMS interfaces. Frameworks tend to focus too much on the CRUD, and CMS tools tend to focus too much on the Modules and flexibility. But it's easy to miss the point that it's not what these tools can do technically, it's how the underlying model is presented to the user that really counts. In any case, I think there are many interesting design and IA problems in this area, and there are lots of interesting ways of solving them, regardless of the back-end technology.
sigurdm...@gmail.com
However, I have to say that with your cons, none of these
1. Flexibility - You know the code (but you still have to train
others)
Short term yes, your quick hacks and changes here and there make for
cost effect changes. But then your system will get unmaintinable over
time, and the client will get frustrated when things become more and
more hard and expensive to upgrade. Importantly, the client will
suffer from having a small number of people fund and progress their
site. They will need to pay, often solely, to improve it, rather than
just upgrade to the newer (free!) version, and find the newer version
substantially faster/better/more secure, etc. So, I'd think of the
client's perspective on this one.
2. Total development control
As above.
3. Total IP ownership
It's more sustainable for the client to take open source code, extend
it, and give the appropriate parts back. More sustainable for the
client, and for the overall project. An example may help. The client
may, for instance, need a special "poll" module, never before built
for the global open source framework you're using, and you convince
them to release that as a official module. Others pick it up in the
open source ecosystem, and over time it grows and matures, and can be
rolled back into the original client's code easily. The poll would
never mature like this in the closed source world. Since you gave
birth to this module, you're give direction to it, so you can keep it
relevant to your implementations. And ofcourse, this process of
maturing and extending the project is sustaining the project itself.
If you find yourself infected with the GPL and believe the code you're
working needs to be held back from the open source world, then choose
a BSD licensed package. That's one reason we release the SilverStripe
web development platform and CMS under the BSD license. We hope you
choose the above ecosystem-friendly option for all the mutually-useful
parts of your code, and have the option of still using SilverStripe
for your competitive advantage-generating code.
4. Development can follow client feedback easily
easily, but you're moving at a snail's pace in comparison to a project
with hundreds of people programming it. Look at how much we and the
other global open source projects got out of the Google Summer of
Code. More than any single NZ web project ever put into the reusable
core of a code base, I'd say.
I'd suggest web developers becoming heavily involved in the
development of an (larger than then) open source project, in which
case you can reap the benefits of a large, standardised platform, and
the inertia behind it, and also be able to incorporate your client's
wishes into the core project for the places where you cannot do this
as your own independent customisations/modules. Open source projects
are hungry for quality contributors, and spend a lot of time agonising
about how to get great people extending them; so you'll be very
quickly accepted if you are committed and competent.
Mark's points about using your own are valid if you're doing a niche
system, but often there is a framework or product that gets you half
way. I'd suggest focusing your business such that you can become
highly adept at a few key languages, frameworks, technologies, and
building solutions using that knowledge, rather than being spread out
thinly and having a shallow but wide knowledge to be able to do merely
adequate jobs for a diverse range of situations.
I also think that NZ web developers are slow to follow their US leader
counterparts into specialising into recognised standardised
frameworks, and while we continue to favour roll-your-own in
situations where an appropriate framework IS present, we're ripping
off our clients by spending lots of time on the basics, and
subsequently giving NZers a mediocre web experience... when instead
all our top brands and progressive organisations (and govt!) should be
living and breathing highly interactive social web applications and
self-serve portals. Why must we keep advising "oh, so you're a company
that does x, y, and z and your website needs upgrading? Well find a US
counterpart and that will show you have you the ideal way for your
website to work". We're a small smart country that ought to be
demonstrating innovation to the rest of the world, not the other way
around.
Aaron Cooper
through it today.
So I'll still be hammering away at Drupal for the time being.
Cheers
Aaron