Turning a text-only search box into a visual path to the right product
After the replatform, Cole-Parmer’s basic text search was ready for an upgrade. Its results were too broad for scientists who search by specification, so buyers hunted up and down through categories. I led the project from pitch to launch, and the catalog had more than 80,000 products with complex variants.
Search revenue rose 29%, average cart size grew 25%, and click-through rate rose 10%. This is how it came together, including a CMS limitation we worked around.
What changed
Each line jumps to the part of the story behind the number.
Broad results kept buyers seesawing through categories
Basic text searches returned results that were too broad. They kept users on category pages they had to move up and down through to find products, which added friction and slowed purchasing decisions across thousands of scientific products with complex variants.
The platform could already search. The problem was what a search gave back, and how much work it left for the buyer.
A visual typeahead could put products, categories, and clear labels in front of someone while they were still typing.
I championed it, then set the standard for building it
Getting the project moving came first, and doing it so each person could work without constant check-ins came next.
I championed the project and recruited others to a shared vision. I spelled out the requirements, guided designers and engineers to their own research, and helped them hone their solutions toward the results we wanted.
I also set the standard for how Bloomreach AI worked on the search results pages. Once stakeholders and contributors were on board, I shared the high-level plans with each contributor so they could execute without too many touchpoints.
What success looked like
Four goals, all aimed at getting buyers and scientists to the exact product they needed with less hunting.
- More ways to reach a productExpanded the product options users had within search so users could keep or clear their past search terms.
- The right products firstSurfaced the most relevant and most-purchased products in typeahead results so users can get items into the cart quickly.
- Search beyond a few keywordsScientists want equipment that fits specific parameters so when they searched, honed results had to show.
- Detail in, detail outIncrease conversion efficiency by partnering with buyers and customers who use search phrases with product specifications, and use them to push search to return detailed results.
What solutions had to work within
The solution had to fit inside what already existed and development time was limited, so every decision had to count.
- The existing global header and search infrastructure
- 80,000+ products with complex specifications and variants
- Limited development bandwidth that required focused, high-impact choices
What I found and how I worked
The data and the catalog both had surprises. Some were limits to work around and some were quiet problems nobody had flagged.
The CMS only recognized product families, so we surfaced the best-fitting item inside each family.
- Mapped the whole funnelI analyzed behavior in Google Analytics, Bloomreach analytics, and Glassbox to see how users moved from search to purchase.
- Surfaced items inside familiesWhen product parent families were the only things the CMS recognized, I directed the methods we used to surface the right child item in search.
- Caught 240V listed ahead of 120VInconsistent item listings inside product families put 240V equipment first, which showed European products at the top too often when US buyers needed 120v equipment.
- Used what was already thereI worked within the existing global header and search infrastructure, which avoided extra rebuild work.
- Stayed close to how buyers searchI worked embedded with buyers’ search habits to keep moving search away from simple category names and toward longer, sentence-style searches. That surfaces honed product results instead of broad ones that make people hunt.
A search that kept getting better
The visual search redesign met its targets and left behind a self-correcting loop of stakeholders committed to constant improvement.
The updated search drove a meaningful increase in search revenue. It showed how much reducing friction in product discovery matters for a highly technical B2B audience. It also built on the replatform that came before it.
Need your digital strategy and design pain points solved? Let’s talk.
I’m looking for a team to grow with a team long term in a remote or hybrid in the northwest Chicago suburbs.