If you’re in the market for front-end redesign of your Web application (or eCommerce site), but you don’t know how far to take it, or how much to spend on the overall effort, you’re not alone.

Not every redesign needs to be a major overhaul; incremental or partial redesigns can add considerable value, in spite of their tighter budget. But, naturally, there are tradeoffs. Doing a partial redesign can constrain your effort. Partial redesigns are by nature evolutionary rather than revolutionary.

So how do you know when to go full throttle on a redesign, and when to take a more measured approach?

Defining Scope

In order to “right-size” a redesign, you need to determine:

• The most pressing problems or issues
• The impact on your business caused by solving those issues.

What’s the Current Damage?

If you haven’t done so already, you’ll need to apply due diligence to the current issues within your application. You may find you’re too close to your own product — especially if you (and everyone on your team) have contributed to it for any length of time.

If you are too close to it, get outside help — a consultant in usability or user experience can do a heuristic evaluation to help you identify the most pressing concerns. Customer feedback will act as a beacon for you or your consultant. What are your customers and users saying about your application?

You need a major overhaul if:

• Calls to customer/technical support call centers are growing at a faster rate than the customer base itself
  
  
• You are hearing frequent reports of ease-of-use issues — i.e., customers having trouble finding information or completing transactions
  
  
• You have poor customer retention
  
  
• You know you need to add significant features or content to stay competitive.

You may need a more focused redesign if:

• Calls to customer/technical call centers are centered on selected sections, pages, or flows
  
  
• Your content and functionality is at least on par with the content and functionality of your competitors
  
  
• Your user interface has been redesigned within the last two years.

What’s the Likely Impact?

What is the potential impact of fixing the above problems? This will drive the way you allocate your design dollars.

You need to determine whether your business can get a sufficient bottom-line boost by fixing the problems you have articulated. Your budget justification for a redesign will be based on whether solving the problem can help you achieve greater business success through increased sales, improved automation (and less call-center support), or improved customer satisfaction and retention.

Design can have a tremendous impact on the acceptance of software applications, websites, and online services. But unless a redesign is focused on solving issues, you’re just putting old wine into a new bottle.

Focus is Crucial

Everybody gets excited about full-scale redesign because, well, it can be a lot of fun. There are decisions about “look and feel” to be made, and it can be exciting to plan and implement new features and functionality. But a partial redesign may solve many leading issues at a fraction of the cost and is more appropriate in many (and probably the majority) of situations.

In order to be sure your partial redesign is as effective as it can be, you need to go for the core — focus on only the mission-critical elements that customers are complaining about or that you know need to change in order to drive change in the business.

Make sure your design team stays on track by focusing on the top metrics for success. These can vary widely, but typically can include improvements to:

• Click-through rates from landing pages or home page
• Registration rates
• Lead generation
• Completion rates on user transactions.

Bolster Your Effort

Incorporating some feedback from real users can help ensure the success of your design effort — whether it’s a large-scale or a partial redesign. Design work can be tested by prototyping and getting feedback in any of the following ways:

• Low-fidelity prototyping. Low-fidelity prototyping can be as simple as pages or sequences of pages executed in PowerPoint. These can be shown to users informally to get quick reactions to early design concepts, such as different layouts or navigational schemes.
  
  
• Screen mockups. This involves more professional layouts of pages in Photoshop or another graphics program. What’s useful about screen mockups is that users respond to them in a very realistic manner, and many even try to click on page elements. This gives you a pretty good gauge of what users are looking for, for a modicum of design effort. Screen mockups can be tested either in-person or through Web and phone-based interviews with users.
  
  
• Working prototypes. A working prototype looks and acts pretty much like the real thing. The code itself can be throw-away code, or it can be partially constructed — the main thing is that the user needs to be able to perform a few high-priority tasks so you can see whether there are any remaining issues to be tweaked in the design. Working prototypes can be tested either in-person or via remote Web and phone-based interviews.
  
  
• Pre-release code. It’s a good idea to test with a limited set of users (such as on a beta site) before going to your production release. (See also my December 2006 article, “Testing Multiple Models,” for more information on A-B and multivariate testing approaches).

Balancing Design, Development

What percentage of a budget should be devoted to design? And what percentage should be devoted to user testing? I get this question regularly from corporate managers at all levels. Design and user testing may seem more mysterious than other budgetary line items, but they don’t have to be.

Good, effective design can consume anywhere from 10 percent to 30 percent of your overall redesign budget, depending on the sophistication of your front end and the complexity of the development effort. (See Fig. 1.) User testing typically runs at only 5 percent of an overall budget — a modest amount, considering the rewards it reaps. (If you want to put more into user testing, however, your application will benefit.)

The total dollars involved in design and user-centered validation end up being relatively small compared with development dollars. Viewing them as percentages of the total product development budget is a useful way of seeing them in context, and ensuring you have enough of a budget to do an effective job of your redesign — whatever its scope.

Meryl Enerson is president and founder of Enervision Media, Inc., a user-centered research and design consultancy. She has assisted numerous software companies as well as Fortune 1000 corporations in evaluating and improving the user experience of their websites and applications. E-mail her at meryl@enervisionmedia.com.