Balancing Optimization and Web Development

Posted by | Comments

Tags: , ,

Kevin Eichelberger, founder and CEO of Blue Acorn, says that optimization is not given enough weight in design and development. He notes that in many companies today, optimization is done after the fact, when ideally, optimization techniques should be incorporated throughout the decision-making process.

In an episode of Landing Page Optimization, Kevin and SiteTuners CEO Tim Ash discuss making optimization and web design and development play nice.

Choosing the technology platform

Kevin and Tim agree that it makes sense to first define a holistic set of requirements that takes into account the optimal user experience and then find the platform that’s going to best suit those requirements.

Tim says this is so you’re not constrained by the platform which will ultimately result to poor user experience.

Kevin points out, though, that not enough companies go through the user experience before the request for proposal. He shares that in Blue Acorn’s case, they work with primarily with Magento, which has a great deal of flexibility that enables them to deliver on requirements as they come up even if it’s after the fact.

Making user experience part of the early stages

In cases where the platform is chosen before designing the user experience, one possibility is merging the user experience team with the optimization and data team. Kevin shares that those teams will collaborate during the initial discovery and design processes in creating use cases and tasks and finding opportunities for optimization.

It’s also important to test in the current site. As you think about what might be a better user experience, through research, use case development, and user experience flows, you’re creating hypotheses that would be great to validate on the current site. This could be in the form of conversion optimization or A/B split testing on the current site to validate the hypotheses before they get thrown in with the new site.

Tim notes that you can’t test fundamental changes, only localized changes. Tactical tests can be done within a product detail page or the checkout process; however, it would be hard to test a complete rethinking of the buying process.

Kevin says that for larger wholesale changes, you’ll need user testing or qualitative type of feedback points of prototypes.

Measuring changes

After the new site is launched, sometimes companies just start making some arbitrary changes to the site.

Kevin says that everything, at minimum, should be measured, and that with today’s technology, the most accurate form of measurement is testing. He says it’s a step-by-step evolution and that sometimes you have to start by testing the big user experience changes or things that are simple to test to get some momentum.

Tim points out that measuring everything doesn't necessarily mean testing everything. If you’re going to make a change, collect data of how things were before, and then measure what happens after the change.

Kevin Eichelberger

Blowing up the site

Kevin shares that he prefers not to blow up the whole site unless there’s really a need to. He suggests looking at a variety of data points and ultimately, the overall performance of the site:

  • Is it a stable base that we can work off of?
  • Does the platform have the functionality, capabilities? Or is it limiting us from creating a robust user experience?
  • Does the design have a lot of the best practices and fundamental elements that will allow us to continue to build upon? Or is it just a terrible starting point to begin with? 

He adds that his company also has to take the client’s budget into account, and whether they’d rather pump everything into one big redesign or spread it out over the course of a year in a results-oriented optimization program. He says a full redesign has the potential for bigger impact, but that there’s also the risk that the site that might perform worse.

Tim counters that iterative testing will not you to the same spot as a wholesale considered redesign and that you can’t get to an optimal experience by just tinkering. He says that while a study has shown that 40 percent of ecommerce website redesigns perform worse than what they replaced, this failure can be attributed to lack of focus on conversion for user-centered experience. Redesigns are usually a visual facelift or a re-platform of putting the same things on a different platform. Companies who are focused on visitors’ needs tend to have a universally much higher performance than what they replaced.

Kevin shares that the evolution of web design is getting more around user experience. He says focusing more on creating a better user experience, and using data and testing as part of the process to understand what is a better user experience beyond a creative director’s opinion will lead to better results in full redesigns.

webmaster-radio-1.jpgListen online or download the "Testing and Website Redesigns with Kevin Eichelberger" podcast from

First Air Date: July 28, 2014