Our Portfolio of Successes

While we do conduct all forms of generative and evaluative user research, including competitive research, interviews, surveys, and usability testing, observational research remains our most successful generative method. The following list describes just a few of our generative research successes:


“Proflowers.com is one of the most successful e-commerce companies and we owe a good part of that success to the user-experience processes and designs that Larry and his team provided.

— Bill Strauss, President, Proflowers


One of our earliest projects is our best case study of a successful user research effort. The executives of an e-commerce site commissioned us to design a new approach to buying flowers online that would disrupt the florist market. They were surprised that our results didn’t look like anything they had seen before. Needless to say, they were skeptical. 


The executives initially suggested a number of key competitor websites to use for usability testing. Instead of running tests on these websites, we visited brick and mortar flower shops to observe how people bought flowers without relying on a specific technology. Remaining technology agnostic exposed the user’s cognitive processing behaviors better than watching how they struggled with a poorly designed website.


Very quickly, it became plainly obvious why these other online florists were suffering. They were solving the wrong problem, very well. All of the existing florist sites expected their users to create bouquets. The problem was that most people buying flowers were men. What do men know about flowers? Roses, and that’s pretty much it, so any site that relied on men building bouquets was doomed to fail.

We observed that buyers performed the similar behaviors, every time. They wouldn’t ask the clerk for specific flowers, such as Lilies or Daffodils, they simply said “I forgot my wife’s birthday and need flowers to get out of the doghouse. What’s a good bouquet to say I’m sorry?” So men weren’t there for the flowers, they were there for a reason or an occasion! Our UX strategy was obvious.

Proflowers doesn’t sell flowers, they sell occasions.

Given that men know nothing about bouquets, especially knowing which bouquet is right for each occasion, we designed the site to avoid that knowledge requirement. Rather than requiring men to build bouquets, or even to know which bouquet to buy for a specific occasion, we organized the site around the occasion. Men knew WHY they need the flowers, so all they had to do was select the right occasion and they were presented with a selection of bouquets perfect for that occasion.


Proflowers launched their new site just in time for Valentine’s Day, 1998. A few weeks later the executives called and wanted their money back. The new site was a miserable failure!  

We reviewed the new site and noted that the developers took obvious liberties with our original design. The developers admitted that our design didn’t look like any of the other florist sites so they changed the design to make the new site look just like the other sites.

The Proflowers executives were livid and told the developers to rebuild the site following our designs. The rebuilt site launched in time for Mother’s Day and has had a top-ten conversion rate every month for 20 years. 20 years! This unintended A/B comparison proves the power of good user research.

Unnamed Enterprise App


The client was just another entrenched, second-tier company, which had relied on their 30-year-old design approach with declining success. (Sorry for the lack of specifics, but this is still under an NDA) Their product was indistinguishable from its competitors and they needed to shake things up a bit in order to earn more market share. What we identified became the buzz throughout the industry. It was so novel that even though the competitors have become aware of this new approach, they don’t have the specific knowledge to address this new unmet need.

Research & Insights:

The early observational research clearly indicated that the product served only a fraction of the total user task environment. The original product was designed to operate on a set schedule, but there were mitigating issues that occurred almost daily that required tedious schedule changes. Once the situation returned to normal, it was just as tedious to revert back to the previous schedule. Consequently, this process invited numerous user errors that constantly frustrated the users.

This issue was common to all of the competitor products. They were all making the same mistake. Based on the research findings, we changed the design approach to focus on accommodating constant schedule changes.  This change incorporates numerous predictable schedule changes and the ability to revert back to the standard schedule with ease.


Initial user feedback indicates that this new design approach promises to disrupt the industry and demand a greater share of the market. The buzz around the industry is so strong that potential buyers are holding off until the product releases. When was the last time you heard of a new design that disrupted the market BEFORE ever being released?

Vanguard Mutual Funds


In their first attempt at a self-service investment management website, Vanguard Mutual Funds created a site design that reflected their business operations with 5 different silos. Each of the 5 investment vehicles had it’s own site, such as Mutual Funds, Stocks, Bonds, etc.. Users were required to visit each silo separately to find all of their Vanguard investments and manually add them up to determine their current value. This frustrated the users and generated exorbitant support costs that diluted the intended cost benefits of the website. Vanguard required a site that fully supported their customer’s needs without intervention by customer support representatives. Our research delivered an industry standard model still used today.


Our observational research in the user’s homes identified that users invested in multiple types of instruments in various combinations depending on their investment objectives and each objective had a different investment profile. For instance, the husband and wife might each have individual retirement portfolios that included mutual Finds and blue chip stocks, a semi-aggressive college fund with stocks, bonds and money market holdings, and a more aggressive vacation home savings plan with other stocks, money market funds, etc. The existing website required the user to visit each of the Vanguard silo sites and copy all of the values into a spreadsheet to determine what their current net worth might be.

Unfortunately, this still did not serve the main objective of determining how well each portfolio was doing with respect to their investment goals. 


Users had difficulty navigating the 16,000 pages of the Vanguard website and, given the similarity of the cryptic fund and stock names, they weren’t always sure they found the right data. Moreover, the spot prices they were shown did not indicate the progress towards their investment goals. Additionally, some customers bought stocks that were already part of one of their mutual funds, often over-exposing them to that stock.


The solution strategy was to turn the site into a wealth management agent that collated all of a customer’s holdings into separate portfolios based on their objectives, then plotted the progress for each portfolio, indicating how each were progressing towards their individual goals, and, most importantly, providing action items to address any anomalies, such as over-exposed or profile incompatible holdings. This solution has been in place for 15 years and has become the de facto interaction model for all other major investment firms.