IMPROVING SANTANDERS 
CONTENT DISCOVERY 
EXPERIENCE.
FIRST STEPS
INTRODUCTION
Santander sit at the very top of their sector within commercial and business banking, rubbing shoulders with the likes of Lloyds TSB and Barclays within the competitor discussion. Like most banking organsations, digital innovation sits in the forefront of their success.

Within recent workshops, Squiz and Satander outlined a backlog of key improvement areas. A glaring gap between Santander and their competitors had been visisble for all to see for some time. It was clear to see that use of the search module within Santanders global nav operated in a very basic way, and didn’t offer the assistance that Santanders competitors could within their respective search experiences. On top of this, when landing on the Santander SERP (Search Engine Results Page), there was no ability to filter your results or great clarity surrounding refining your search.
CHALLENGE
DEFINING THE
WHY
Santander sit at the very top of their sector within commercial and business banking, rubbing shoulders with the likes of Lloyds TSB and Barclays within the competitor discussion. Like most banking organsations, digital innovation sits in the forefront of their success.

One of the key points of discussion surrounded the type of account holder a user is. This often helps banks paint a generic picture as to what kind of needs a user may have. A personal account holder for example, may be considered to be somebody that only has a vested interest in managing personal affairs. Their concentration is largely focussed on receiving their wage, and managing their monthly finances. A select account holder is percieved to be an account holder that may embrace the perks and benefits of being a prioty customer. They pay a little more into the maintenance of their account in exchange for priority services such as customer care priority, gadget insurance, holiday insurance, and other perks. A business account holder is percieved to be somebody that relies on their account for the financial maintenance of not on themselves but a potential business and workforce.

One of the common ground topics regardless of the type of users in discussion, is the necessity for ease of data capture. All users, no matter what their purpose, will rely on their banks website in order to discover areas of information that they will struggle to find elsewhere.

Within recent workshops, Squiz and Satander outlined a backlog of key improvement areas. A glaring gap between Santander and their competitors had been visisble for all to see for some time. It was clear to see that use of the search module within Santanders global nav operated in a very basic way, and didn’t offer the assistance that Santanders competitors could within their respective search experiences. On top of this, when landing on the Santander SERP (Search Engine Results Page), there was no ability to filter your results or great clarity surrounding refining your search.
CURRENT
GLOBAL SEARCH
MODULE
CURRENT
SERP PAGE
CHALLENGE
CURRENT USER 
FLOW
SOLUTION
ASSIST, ASSIST,
ASSIST !
We spent a short period of type validating a couple of hypothesis in order to prototype the proposed solution in discussion. Given the nature of banking, we established very early on that each and every solution we proposed would be implemented at different time periods beyond sign off given how Santander operate their website. Whilst this was far from ideal, our hypothesis had to be broken down into micro-solutions to allow roll outs to be subtle but impactful. 

In order to improve how customers discover important areas of information effectively, we proposed the implementation of a search engine concierge built within the search module. The premise was that when a user activates the search module, as they type the concierge would use our Funnelback Products algorythm to populate the concierge module with predictive results. The results will all sit under their categories, and the more you type, the more refined your concierge prompt will be. This solution could then effectively skip the need to land on a search engine results page altogether, and provide the user with a simple, effective UX solution to a dated method of finding important information on the fly.

We also suggested the addition of a filter module within the SERP, given that the only way to break down information on the page at an early stage was via high level tabs. A filter module would further reduce the need for users to paginate through multiple layers of information.
SOLUTION
PROPOSED USER 
FLOW
SOLUTION
PRODUCT 
IMPLEMENTATION
As apart of the ideation provided, it was important to conduct considered decisions surrounding the technology we proposed to execute the solution.
Within this project, Funnelback was the product to execute the proposal. This was a product Santander had been using previously to power all of their search and SERP operatives. Within this round of design and implementation, we proposed an updated version of the product that included all of the latest innovation we have been working on over the previous 6 months.

It was important to communicate how Funnelback would work in unison with Santanders servers, and we provided a detailed document to breakdown how this relationship would function.
VALIDATION
DEFINING THE
WHO
As previously mentioned, the Santander web infrastructure is a single source of information for all tiers of online and on-site bank account holders. Users of the website could be quickly grouped by account type, but the overiding trend that we discussed surrounded the fact that everyone should at some point have a strong case to search for a valuable piece of information based on their own account or situation on the site.

Through multiple rounds of interviews and research, we established effective ways of resolving user pain points, and boiled our user type down to three of the most popular types of account holder. Through doing so, we were able to validate our early phase ideas.

Our main research methods used: qualitative interviews (interviewees provided by Santander), user surveys (created by my team, distributed by Santander), market insight gathering, and competitor analysis.
VALIDATION
USER STORIES
Off of the back of our user testing, I collated proportionate feedback and narrowed down each of our account holder/user groups to two definitive sticky sticky notes per group (using a validation chart). In doing so, I was able to identify common user stories.
VALIDATION
PERSONAS
VALIDATION
DESIGN 
ASSUMPTIONS
PERSONAL ACCOUNT HOLDERS

· Users aren’t finding key information quickly.
· Users dislike using the SERP.
· Users want to find specific contact details.
· Users want to manage personal finances.
· Users don’t feel assisted.


SELECT ACCOUNT HOLDERS

· Users often land on SERP when looking for priorty benefits, but faily to find desired information.
· Users find using the global navigation is fairly efficient, but often struggle
 to find contact details.
· Users feel the SERP is dated, and disconnected from the rest of the site.


BUSINESS ACCOUNT HOLDERS

· Users find contact numbers are overly tricky to find.
· Users find gaining information abroad at pace can be frustrating.
· Users would like to find business information at pace.
DESIGN
SKETCHES
I led multiple sessions between the client and the Squiz project team in order to deliver  ideas. Due to the face that participants ranged from London, America, Australia, and Spain, we utilised Murals whiteboarding capabilities to collate and validate ideas.
DESIGN
LO-FID
WIREFRAMES
DESIGN
MID-FID
WIREFRAMES
DESIGN
BRAND
DISSECTION
In order to achieve complete consistency within the build phase, I worked within Santanders comprehensive design system.
Often, even small enhancements can stand out like a sore thumb, be it a pixel off in padding, or a visual language miss. The complete adherence to the design system ensures a completely seamless integration at a small or large scale.
DESIGN
CONCIERGE
DISSECTION
With the application of a concierge, the journey from question to solution is rapidly decreased. Technology such as autocomplete further reduces the time in which it takes to land on a solution. As previously mentioned, user assistance is the name of the game and although the SERP page still remains important, if we can reduce the necessity to use it, we are creating a simpler user experience.
DESIGN
SERP FILTER
DISSECTION
Whilst projecting the application of a concierge that will often negate the need to use a SERP page, it was still important to recognise the importance of the SERP page in general. For users that may type things incorrectly, phrase words or sentences wrong, or simply not quite understand their subject matter as well as they could, the SERP page offers a second source of clarity.

With the glaring issue being that the SERP page only offered minimal filtering in the format of top tier tabs, the application of a filter bar/dropdown gave users great opportunity to refine their search results.

With statistics showing that users become disinterested when scrolling beyond the first 5 results of SERP listings, it became apparent that Santander needed to give their users more of an opportunity to refine results.

Through multiple filter selection, the user is narrowing down an enormous percentage of potential listing results, with top level tabs still retaining their purpose in defining the user type.
DESIGN
COMPONENT
PROTOTYPE
To enable the client a greater feeling for look and feel, I created a clickable prototype in Adobe XD. Whilst the purpose isn't to directly project interaction, it offers the client the opportunity to pass around a tangible piece of work. This is super important to secure sign off from the different departments within the organisation.
DESIGN
COMPONENTS IN
THE WILD
REFINE
PROTOTYPE
VALIDATION
To enable my team the opportunity to gain further validation from users, we set up a testing environment using our usertesting.com environment.

Through uploading the work I had created in prototype mode, I was able to set up a validation environment allowing users to navigate their way through the proposed changes, answer select questions, and leave feedback on their initial thoughts.


Search Concierge Questions (Desktop & Mobile seperate)

- What are your initial thoughts on the concierge assisting you?
- What makes it easy?
- What makes it hard?
- Do you prefer this to the existing search module?


SERP Filter Questions (Desktop & Mobile seperate)

- What are your initial thoughts on the SERP filter?
- What makes it easy?
- What makes it hard?
- Do you prefer this to the existing SERP page?
CONCLUSION
NEXT STEPS
Off the back of testing and validating the prototype, I uploaded the design work into an environment that displayed all asset values to allow the development team to quickly carry designs over into cutups.

I prepared a document for Santander summarising our findings within the design phase, and included thoughts surrounding quick wins that we could add into the backlog.

Findings for backlog:

- Improve SERP cards from an accessibility standpoint.
- Utilise full-width header and footer components projected in design on SERP prototypes.
- Apply more brand to SERP.
- Imagery within SERP cards.
- Apply concierge within a hamburger menu, inclusive of imagery.
- Reduce use of tabs, and rely more on automated assistance and the new filter module.