It’s well known that simple-to-use site work better than complex one but that’s not what I’m talking about here. I want to cover briefly the major components of a web-site and specifically where and how they should be developed. Many web-sites re-invent the wheel (for example a site search function?) costing in development resource and also in lack of function. Taking the example of site search, would it not be easier and better to provide Google’s site search on your site rather than pay for a development company to develop a custom search function?
Well, there are pros and cons for each approach so I thought I’d explore some of the major functions expected on a site and see which ones are, on balance, seriously worth considering but before I start this I thought I take a generic look at the advantages and problems or using third-party widgets on your site.
Advantages
- Best of breed is possible here, you can potentially have a widget that’s had thousands or even millions of man-days of development. That has got to be attractive
- Many of these widgets are hosted on the cloud, provided as SaaS maybe, reducing the load on your own hosting / server / communications infrastructure
- The cost model for most widgets it based around either low-cost license or medium-cost capex. Very rarely do you see an expensive capex widget
- Changing widgets can be pretty simple, allowing you to move with the times and adopt the latest, greatest, newest thing out there
- Using industry-standard widgets (Google search, Facebook Like) means visitors do not require a learning curve
- Ongoing improvements by the suppliers can usually be incorporated
- Browser and platform support is carried out by a third-party, reducing costs, overhead and (usually) lead-time
- Some widgets have APIs to allow add-on development
- Ease of adding new gadgets and widgets can quickly create a more interactive environment
Disadvantages
- Can be difficult to bespoke build some items, risk of shoe-horning the business needs into the widget and not the other way round
- Costs can mount up for simple license terms over time, especially if the widget has a poor RoI
- Managing lots of suppliers for different elements requires good project and cost management
- Suppliers come and go quicker, they stop developing a widget, or get bought, or go bust so you need succession planning
- Ownership of data can be a little blurred generating possible privacy and regulatory issues
- Easier for your competition to “copy” you be installing the same widgets
- Harder to differentiate your offering if you use standard tools
- Dependent on the uptime of the third-party providers (only recently Amazon EC3 servers went down and currently Sony’s Playstation Network is down!)
- Ease of adding new gadgets and widgets can quickly create a more dis-jointed environment
Overall it is worth considering a mix of approaches, some standard functions are actually better served as widgets (site search is a good example) but where the web-site is reflecting or providing key functions that differentiate you from the competition is is probably better to stick to bespoke and custom-built applications. By way of example; a good search engine is expected on most sites but is very rarely critical to it’s success (unless you are a search engine).
Before we dig into the actual widgets (part 2) it’s worth considering the Free vs Paid vs Freemium model for widgets. Obviously, a free service has the benefits of costing nothing to run but usually has the downsides of your videos being used as a revenue generation method via adverts, interstitials, sponsored links etc. Paid is a better option as that usually entails some service level agreement and sometimes the ability to “white-label” the widget. Freemium sits half-way in between and has seen its’ historical credibility waver a little recently but a great way to try something for free and then upgrade to the paid-version if it works for you. However, if the fully-paid-up version is only £50 a month then you’d be nuts to go for the free version if it costs only £50 to try out the real, industrial-strength, version.
(…contd)
