Vanguard's UK Personal Investor site was moving to a new content platform. I led the design of a focused rebuild: three high-value pages, redesigned rather than simply migrated. Conversion improved on every one of them.
The site was moving to a new platform. One option was a like-for-like migration; another was a full redesign. I recommended a focused rebuild of the pages where evidence showed the greatest opportunity.
A business will not willingly reopen a site that is converting. The migration was the only moment this one would be open, and spending it on a like for like move would have closed it again for years, with every known client pain point carried straight across.
Vanguard has 600,000 customers in the UK and 50 million worldwide. Redesigning every page at once means changing a site that already converts, with no way to tell which change did what and nothing to fall back to if it drops.
Everything after these was execution. These were the key decisions I led with the team, and they shaped the work that followed.
Migrating the markup untouched was the cheap option. I made the case for a rebuild: the rebrand had landed everywhere except here, the new component library meant every other team would inherit whatever we built, and the known client pain points were all in the pages we were about to copy.
I took the scope out of opinion and into Adobe Analytics instead. After the homepage, the ISA and SIPP product pages were the most visited on the site, so three pages was the smallest change with the largest effect, and a scope the team could validate before extending it further.
The industry splits investors into confident and nervous. Fifteen interviews said there is a third group and it is the biggest one: people who have invested for years and still would not call themselves confident, and people who understand it perfectly well and simply have no time. Neither is served by a beginner's guide or a fee table.
No single one of these would have moved a business this size. Together they were hard to argue with, which was the point.
The analytics, not a workshop. Where people go and where they leave is what chose the three pages, and it is what made the result measurable afterwards. Picking the pages by evidence is also what made the conversation with the business short.
Trust does not come from one source. It takes reputation, assurance, safeguarding and communication at once, and satisfying it is the single biggest gain available.
People's idea of who counts as an investor is narrower than the truth. Demystifying it would let far more people in.
Being proactive and easy to reach matters, and once someone is in touch the service has to build a relationship rather than close a query.
An excellent digital experience across devices is simply expected now, and close to a deal breaker where it is missing.
We were addressing less confident investors and more confident investors. What we were missing was the investor who has been investing for a long time and still does not deem themselves confident, or who simply has no time. That person is not a beginner and does not want to be spoken to as one, and they are not an expert either. It is the reason the pages explain rather than reassure, and the reason fees are stated in full on the page instead of linked to from it.
Content design and product design worked on the low fidelity wireframes together, so the words and the layout arrived at the same time rather than one being poured into the other.
Content design in the room from the first sketch, so nothing had to be rewritten to fit a box that was already built.
I built the kit and put a design system behind the high fidelity work. Same instinct that later became Nebula: decide it once, in one place, rather than in every ticket.
The first round's issues were resolved and new pain points appeared in the second, which is normal and worth saying out loud rather than tidying away.
We prioritised what made launch in the open, so the cuts were a decision people had watched being made rather than a surprise later.
Three pages, each with a different job, each measured on its own. The pattern underneath all three is the same: explain the thing before asking anyone to buy it.
The old hero gave a visitor nothing to do. The new one puts the two things most people came for at the top, and then answers what Vanguard is, what it offers and how to invest, in that order, because that is the order the interviews said people ask them in.
Conversion from 3.9% to 5.4%, a rise of 38%.
The page that carries most of the new money, and the one where trust is either built or lost. The FAQs are not invented: they are the questions the client service team answers most often, moved to where people are already asking them.
Conversion from 8.1% to 16%, a rise of 96%.
The hardest of the three to explain, and the one with the largest gap in the old site. Self employed people had nowhere to go at all, which came straight out of the client services session rather than out of the analytics.
Conversion from 15% to 25%, a rise of 72%.
Released in January and measured in Adobe Analytics against the same period before launch. Conversion here means visitors who go on to open an account.
Which was the whole argument. The site does get redesigned, in the order the evidence pointed at rather than all on one night. Three pages were never the ambition, they were the amount of change worth betting on before anyone had proof, and the proof itself.
The most valuable part of this work was using analytics and research to make scope a shared decision, rather than an opinion. The evidence gave the team confidence to invest where it would matter most.
Out of scope requests are what derails a project this size. A plan everyone has already seen is much harder to quietly expand.
Paper, low fidelity, anything. Every round we ran late cost rework a cheaper round would have caught. And I would push sooner on the pages beyond these three: the evidence to do it was there by launch.