Web development: The great convergence myth

I’ve always learnt that converged systems are the best way to go. In theory they are: An integrated site should ultimately allow you to do more with your site’s content and resources.

It should save you time because you are not going backwards and forwards trying to get one Content Management System (CMS)* to speak to another or one paradigm to work within another. There should be a cost saving, because you’re now working with one system, not eight.

Well that’s the theory. The reality is that it doesn’t work like that. Not even close. The ideal vision of a converged CMS and website operation is idealist bunkum.

In reality, it works differently. Many expanding online businesses face questions of whether to build their own or outsource to other platforms. Depending on the size and focus of the business, an online publishing operation’s core system should probably be inhouse. Then, from a business point of view, the better call is to “outsource” your blog system to a specialist blog provider or your social systems to a social networking service. If you need e-commerce, jobs or classifieds platforms, you wouldn’t build your own either (unless you were a specialist provider planning something innovative).

It also depends on the size and type of your operation too. Smaller to mid-sized operations are more likely to outsource. Larger organisations would either build their own at the start or outsource at the beginning as a proof-of-concept. It may depend on how core the outsource function is to your operation too. There are some features where building your own never makes sense. An example of this is blogs. Why go to the effort and expense of building your own blog system, when there is something out there better than you could ever dream of building? (It also depends on whether you are going to be a blog service provider or just use the platform to blog. The former may push you to do your own).

What this means is management of your website will eventually become a patchwork of CMSes and separate systems. The trade-off is that your outsourced systems won’t be as integrated with your core system as you may like, but you’ve won on three fronts: time, cost and cool functionality from a specialist system.

Your tech guy may hate it, because from his point of view it’s inefficient — and a converged system is a complex thing of beauty allowing him to fulfill his role of the tech god you can’t live without.

But it isn’t solely about the tech. It’s about the business and community you build around your site. You make a trade-off: build my own at cost and time vs use another specialist platform at probably a tenth of the time and cost.

Cross platform sites bother me less and less these days. This is the era of RSS and APIs — advanced integration between disparate systems can be achieved fairly robustly without them needing to be part of the same system.

For many media operations the same applies for the different systems managing their websites and traditional media products, such as their newspaper or magazine. The theory goes that you should try converge your web and print systems. For an often-meagre gain, this integration comes at much expense and time. Personally, I’d rather shoot for achieving basic integration on a RSS and API level and keep the systems separate. (This is another post all on its own).

And if it really bugs you, the issue of multiple systems and CMSes can be mitigated by centralising them on a single page, allowing a level of auto-authentication for your staff. If uniformity is a need, you could address this on the front-end by tweaking the look & feel. You could even have a centralised bar across your CMSes, bringing them together on a usability level.

My advice: If it makes sense, outsource it. Convergence is utterly overrated.

* Can we stop calling these “Content Management Systems”. They should do more than manage content, but manage your business. Prefer “Business Management Systems”.

Comments (9)

  1. Elan wrote::

    Good post,

    I agree on the tech level but I think the real biz question is who owns the user and who gets to monetise them when it comes to outsourcing platform models.

    Tuesday, July 14, 2009 at 11:49 am #
  2. Must say that over the while I have come to agree, focus on the core and then just use the right tool for the other jobs whatever plaform or provider. Watch out for ‘lock in’ however.

    An exception to the rule might be for strategic purposes.

    Tuesday, July 14, 2009 at 11:55 am #
  3. Alan Hammond wrote::

    Good post. I agree with you that if you are a media company you don’t really want to own the technology. Why spend money and hours (which is more money) to develop something that is already available for a monthly fee? Rather spend the money making your content, marketing and user experience better than the competition.
    I have seen companies with large departments of permanent tech staff. They end up with a large, ongoing expense and I don’t think they are going to get the best technology.
    The employees are going to be biased towards the systems they know and use. Things change so quickly in this business management need to be able to move swiftly to a new platform if that is better for the company.
    Of course when you are as big as 24.com or Avusa there are addtional concerns about stability and scaleability. But for the majority of smaller sites this isn’t such a concern. Although if you are outsourcing to international firms they may be more used to the type of scale that only the larger SA companies reach.

    There isn’t really an issue about who owns the user. Of course the company owns the user not the outsourced CMS provider.

    Tuesday, July 14, 2009 at 12:10 pm #
  4. Andries wrote::

    Couldn’t agree more, but insight and knowledge are often also centralised around these huge tech departments with their specific preferences.
    What bothers me, however, is that very few understand the CMS-demands of their business. Often publishing companies end up with content solutions not in line with their high-volume publishing needs.

    Anyway, in agreement on keeping core in-house and getting the rest where it’s the best…ain’t no use in reinventing the wheel. Especially if your market are flying or walking.

    Tuesday, July 14, 2009 at 12:51 pm #
  5. Tracey Steyn wrote::

    Natural selection, a process which drives evolution ensures that helpful traits are passed on from one generation to the next. In this process species diverge, an evolutionary tree is just that – it has a starting point and branches out to become more and more diverse as time passes. Convergence of species does not happen – the branches of the tree don’t join up to reform a trunk, because it does not lead to optimization. I firmly believe that this is true in the digital world as well. We should invest in technology that is best for the job we need it to perform, keep things simple and at best ensure that all our tools co-exist in an optimal ecosystem.

    Tuesday, July 14, 2009 at 2:03 pm #
  6. Johan Nel wrote::

    I like the naming convention of BMS instead of CMS. A proper BMS system should be able to push relevant content, leverage user behaviour, incorporate CRM data (if you are an online retailer), run analytics and finally accommodate your ad serving tool. Reporting and optimization in a dashboard view – that’s workflow heaven

    Wednesday, July 15, 2009 at 8:47 am #
  7. Tim wrote::

    So why is your dev team continuing to invest in a blog platform when WordPress-MU is free and good? Not too sure what you’re trying to say when you have to the power to follow your own advice and don’t seem to be doing it?

    The “not invented here” attitude needs to be stamped out, but I don’t think you can make sweeping generalisations about when a technology should be built or bought.
    In the case of high-volume online publishing, an inhouse CMS is a strategic asset.

    Sunday, July 19, 2009 at 1:40 pm #
  8. matt wrote::

    @tim …looks like you missed the part where i say it “depends on the size and type of your operation”.

    I think the point that’s being made here is that its ok to have multiple systems (as opposed to a single, converged one) — whether that be your own or outsourced

    Sunday, July 19, 2009 at 9:54 pm #
  9. Tim wrote::

    Ok, guess I didn’t see where your cut-off point was for “big” or “small”.. I consider 24 small compared to the WP community.

    Tuesday, July 21, 2009 at 7:58 am #

Trackback/Pingback (1)

  1. […] Here is the original post: matthewbuckland.com » bWeb development/b: The great convergence myth […]