Prudential - Corporate Pensions, outcome revenue increased from £3m to £15m a year.
I redesigned 7 pension scheme websites for three UK public corporate bodies servicing 100,000s of employees, built online joining capability and set up the framework to measure the revenue generated by the websites.
Summary process: Stakeholder interviews, requirements workshop, personas, analytics, sketching, wireframing, prototyping, user testing
Revenue increased from £3m to £15m over 6
To understand what stakeholders thought the problems were I interviewed them, attended sales presentations and listened in on calls. I mapped out issues and opportunities on a journey map.
- Fragmented user journey between touch points and information overload.
- Online joining and topping up did not exist. Users had to print a pdf from the websites and post it.
- Pension plan management was not easy.
- No analytics and no user testing.
There was a role for digital to play but limited resource. Consensus was hazy so it was up to me to set direction.
I wanted to understand the team’s goals, issues, areas of opportunity, and agree a roadmap of priorities.
I presented my insights from the user journey map and stakeholder interviews and asked the team to mark up the map with opportunities and issues. I wanted to prioritise areas of focus. I then asked the team to map ideas / opportunities / features to an Impact vs Difficulty/Cost matrix
- We agreed that the websites should play a much larger role in generating revenue than they currently did.
- They would need redesigning to convert better and we set up revenue targets.
- Persona work and user testing was required and web analytics needed to be set up.
- I was keen to build online joining and topping up but this was deemed a longer term goal by the team.
We needed to know how the websites were doing, so that we could measure changes in performance as a result of changes we were making.
- I set up analytics tracking in Webtrends so we could see visits and user journey flows.
- I then liaised with the operations department and provided unique call centre numbers so we could specifically measure what call centre business each site generated.
- I also arranged back end processing changes so that we could attribute sales that came from visitors who downloaded a pdf application form from the websites and posted it to us.
- We could now differentiate these from paper application forms sent via other channels.
At the start of the project the websites were generating around £3m in annual revenue.
We needed to understand who our key users were. What were their thought processes around pensions? What was their context around the problem? What triggered them to visit and take action and what stopped them?
I pulled together customer data that we already had. Segmentation profiles, data from Salesforce, industry studies, brand studies.
I held a workshop with client managers, some of the marketing team, call centre representatives to create proto personas.
Based on the proto personas we quickly redesigned the websites. We would carry out user testing afterwards. I sketched out flows and got the team to critique them.
We started to see better conversions but not as high as we’d have liked. We still needed to user test to see what was really going on.
Analytics told us that more users were getting to the Get started page. They appeared ready to join but drop offs remained high.
Either they hadn’t read the content that should have prepared them for joining (so they weren’t really ready), or something suddenly something became an obstacle. We were getting more calls – so possibly the sites were driving more business, but users were less keen on printing a pdf and posting it in.
This would be a good time to build online joining functionality.
In the meantime I had also been investigating how we could build secure online joining / topping up functionality. Users could only print a pdf from the website and post it in. It took several months of liaising with different departments and to very limited effect. Cost, security and politics seemed to be the issues.
I’d noticed that the operations department had been using an online survey for their presentations in the workplace. We already had the licence, so cost was not a factor, and it provided the correct level of security we needed. I suggested that we could use this solution and adapt it for our purpose. I got the go ahead and worked with a Salesforce developer and operations manager to link the front end to Salesforce and back end database. I designed the front end of the forms ensuring they satisfied the operations and compliance requirements, whilst at the same time being user friendly.
We got the forms up and running and whilst they were not pretty (we were limited in functionality and display settings) the impact was immediate.
- Conversions went up
Previously – PDF form submissions / PDF opens were around 7%
Now – online form completions / starts ranged from 30% to 50% depending on the customer base.
- Revenue per visitor went up.
- The proportion revenue via online forms grew compared with revenue via the call centre – so we were saving money too.
- Overall revenues via the digital channel grew.
No user testing had ever been carried out on the websites. So we needed to learn how users proceeded through key tasks and where they dropped off. How did they approach thinking about their pension and how did they view Prudential’s offering? What information did they need and when did they need it? What was their thought sequence?
We recruited 12 participants based on the persona work and we asked them to complete four key tasks – research, use the calculator, apply, or increase their existing contribution.
User testing insights
Feedback was illuminating.
1 Product value and mindset:
- Users were busy, opening hours did not work for them, words like risk put them off, and they had limited knowledge about pensions.
- They did not understand the value or benefits of the product and how it fitted in with their main pension plan.
- The language did not speak to them or their situation – either ‘not right now’ (younger) or ‘ am I tool late?’ (older).
2 Navigation, order and content did not match mental model
- There was no clear stepped journey that matched their mental model in terms of task flow.
- Too many rabbit holes an inconsistent navigation and too much content made it difficult for users to get back on task.
- A large drop off – users abandoned after 15-40 minutes. All said they’d call someone or speak to someone to get advice about the product.
- Users could not find the existing members link.
- The case studies were interesting and should be more prominent. The calculator needed to follow the case studies more clearly. It needed to be simplified to make a better case for the product. Seemed like too much risk for little reward.
3 Barriers to applying
- Making an investment choice was introduced too late and there were too many pdfs to read. It looked complicated and scary.
Along with recommendations from our usability study I sketched out some ideas and got the team to brainstorm and critique them. I also looked at my older sketches to see where there were areas of unnecessary complication. We agreed a single page flow was the simplest way to illustrate the journey.
- Match users mental model with single page flow – product explanation and benefits followed by case studies, then calculator, then investment choice then apply.
- Ensure users understand the key benefit of the product up front.
- We should also provide easy access to case studies, which users found helpful, and ensure there is a clear link to the calculator.
- Simplify the calculator and ensure it supports the value proposition and does not detract from. It’s got to look like a ‘good deal’. And finally, cut down any rabbit hole links.
I created a prototype of the sketched solution for testing in-house. After further tweaks this was handed off to developers.
I was responsible for all copywriting and stakeholder management to get it signed off from legal, compliance, actuarial and expert stakeholders.