For a number of years, Squiz have powered Santander’s search experience using our prestigious ‘Funnelback’ product. Funnelabck is an industry leading search tool that uses advanced machine learning to to deliver personalised, engaging search results and content analytics. Squiz recognised significant areas of improvement within Santanders search global search experience as well as their SERP, and delivered substantial UX improvements.
Sam Marks - Design Director
Toby Margetts - Strategical Director
Thomas Evans - Sr. UX Designer
Jake White - Account Manager
Sonia Pitton - Project Manager
Donna Moore - Santander
Adam Towell - Santander
What I Did
As a part of their Funneback Search retainer package, I worked with Santander incrementally in analysing search performance, using data to inform and execute design led innovation.
The data in hand was a clear indicator of a pain point, and one that the digital transformation team received a very quick green light to work with Squiz in actioning at pace.
An internal design created to showcase the final Funnelback product utilisation.
We crunched the numbers
We interviewed various users using different account types, gathering valuable qualitative data to help us formulate insights and validate early assumptions.
Observed User Testing
We developed a script and recorded 100 users (using different account types) completing tasks that we had set them.
We implemented the Hotjar heat mapping into the front end of the website to further enable us the opportunity to analyse user behaviour, as well as monitoring a GA dashboard implemented in the early stages of the 1st iteration.
We conducted in depth analysis work to understand how Santander’s competitors approach search and information discovery.
A solution stream session held in the Squiz offices in Shoreditch.
Information discovery is often a time sensitive task, and one of the key takeaways from this understanding was that users looking to discover information or retrieve information related to their finances were often looking at pace.
When searching for sensitive information, users will lose focus quickly and resort to telephone contact.
It is common knowledge that pagination is a hinderance. This was echoed even more so with a large banking organisation, showing users will 86% of the time never venture to the second page within the SERP.
Amidst the search process, it was heavily commented on that the search engine results page didn’t feel helpful or successfully engage users.
It was mentioned that options to refine results were minimal and caused immediate drop-off.
The existing search module lacks intuitive guidance, relying on users to input the right information without assistance or search prompts.
Often users may not know the correct terminology within the search process, extending the period of time spent searching through multiple pages through the SERP.
This often causes loss of engagement, leaving users resorting to telephone or email contact which can prolong time in discovering information.
When users land on the search engine results page, they’re often left scrolling through multiple pages without any helpful filtering.
Whilst the SERP does contain filters, users are only presented with top tier filter options that often don’t deliver the level of contextualisation users require.
The search engine results page is aesthetically dated and doesn’t respond wonderfully to different device sizes and orientations.
Visually the SERP sits within a container that inherits a visual reminiscent of a 90’s website, showcasing little responsive reaction to modern devices.
How might we
How might we
Create an advanced search concierge module to effectively reduce the number of steps in the user search journey. The module should include auto-completion to assist the speed ion which information is presented, and to aid mis-spelling. Improved search categorisation to help first time and return users search by department or category.
The application of a smart, prompt driven concierge was positioned to empower users within their search to discover content skipping the SERP. Prompting covered users just using key words or incorrect spelling.
Advanced SERP Filtering
The application or SERP filters to immediately help users refine search terms, dramatically deducing the number of results in relation to well refined prompts.
Responsive SERP Design
Improving the responsiveness of the SERP, allowing for seamless consistency regardless of the device used would build trust and familiarity for repeat users.
To build a responsive SERP page as well as enhanced components to support greater filtering, we referred to Santanders design system (under strict orders) to make sure all works were delivered cohesively and in keeping with the corporate messaging and delivery.
With the core concept signed off by Santander's digital team we commenced with the implementation of the smart search module, using Santander's design system for corporate adherence from a technical and brand perspective.
Using the power of our Funnelback product, we were able to deliver an experience leveraging the power of multi-source indexing to expand the net reach, and smart search suggestions that deliver related results with ever letter typed.
The end result reducing a reliance on users landing on the search engine results page, dramatically increasing the speed in which they discover important information.
We delivered a vastly improved search engine results page to accommodate users that purposefully or accidentally hold a reliance on SERP use.
One of the earlier prioritisations was to ensure that the skeleton ui kit was carried forward to be response across all devices and orientations. The previous used Funnelback architecture to deliver an excellent technical solution, but left a lot to be desired where the general UX was concerned.
The addition of a vertical filter module was added to assist users in refining and contextualising their searches, aiming to reduce the levels of pagination they were required to sift through.
It goes without saying that the creation of materials housed on the internet requires an adaptive approach surrounding orientation and responsiveness. The challenge with the Santander feature updates was to compile solutions that could be signed off by the design systems team once they sit consistently with Santander's tone of voice and interaction library.
With Santander using Funnelback as their primary search product, data analytics and strategy meetings needed to be a little more regular.
Squiz's relationship with Santander exists solely through search (the Squiz Funnelback product), but we were able to deliver evidence that other elements of the website that weren't in the scope of the project play a big role in how search is received. Using this information, Santanders digital transformation team carried takeaways that could inform the beginning of a wholesale design system update. Using the Funnelback search analytics dashboard, Santander recorded dramatic improvements in how users discovered important information, and we continued to monitor the results to date.