Case Studies
Tiny snapshot of work - better to talk to me, there's so much to cover - let's cherry pick what you need to know....
SeedLegals:
Rapid mobilisation: zero to £40k monthly in three weeks
The problem
SEIS/EIS Advanced Assurance is our first key product that new customers buy. Founders face a catch-22 - to apply, they need to demonstrate they're actively seeking investment - but they don't have committed funding yet so they can't finish the application.
Unless they use a 3rd party that HMRC is happy with,
We had a solution to help founders meet investors, working with one of our partners. But it was baked into a regulated product - not yet approved to go forward.
What we did
The COO spotted the opportunity to extract this piece as a standalone product - one that didn't need regulatory overhead.
The CEO, COO and Sales did a back of fag packet definition. My job was to take that decision and make it real. I had no PM so I became the PM.
In three weeks, I wrote the epics and stories, got the product designed and built, and coordinated across the business - marketing, engineering, CX training, support materials. The CEO drove the marketing messaging. I made sure everything else was in place to actually ship.
Impact
-
Live in three weeks from idea to deployment
-
£40-50k per month in revenue
-
Founders can now demonstrate active investor-seeking as part of their Advanced Assurance application without waiting for committed funding
SeedLegals:
Reducing operational costs through service design & AI
The problem
When you build an MVP, it works, it sticks - but what looks slick to the customer can be another story under the hood. Lots of manual interventions - spreadsheets, admins working via the backdoor into the system to make things happen that the product can't yet handle. That's fine when you're testing and experimenting. The deal is that once it sticks, you go back and fix it.
SeedLegals, like a lot of scale-ups, kept shipping new features (the right call) but hadn't gone back to tidy up the manual processes left behind. Some of those will always be there but a lot of them could and should be automated. It was one of those situations - everyone knew about it but no one knew the size of the prize or where to start .
What I did
No one had sized the problem- so I started there,
Service design blueprints. We plotted out blueprints for each CX team and the associated product. Identifying every time suck and manual process. I also ran separate sessions with the CX team and the Legal team, walking each through their own side of an escalation process that was causing huge amounts of delay and time to solve the issues. When you lay out what a team actually does - step by step, including the workarounds, the opportunities become impossible to ignore. HMRC form submissions were a clear example - something customers could be doing directly but weren't, which meant downloads, emails, even snail mail and CX time.
I've run blueprints in a lot of places. I've never seen teams engage with them the way these teams did. CX hadn't felt they'd been heard before. The blueprints gave them a way to show the weight of what they were carrying and they helped distribute that information across the business.
Processes
Once they could see each other's side of the process laid out, they fixed the processes themselves. I gave them the tools to see the problem but they solved it.
Cost reduction via efficiencies - but evals reduce speed to market
It led to an entire programme of activities - initially taking the top priorities and automating. SQL where we could and then AI for the places we had complexity. The Evals were tricky - it's not a quick fix and this is data where you have to be 100% correct. A caveat for hallucinating isn't going to work on this.
The switch now is with MCP integrated, rather than building custom product features for every manual touchpoint, we've shifted the approach. The CX team now have AI tools that let them handle these tasks themselves. The capability is in their hands rather than sitting in a backlog (albeit still on going).
Impact
-
Escalation time reduced - more time to service customers
-
Manual workarounds being replaced systematically - SQL, in-app and AI tools reducing task time
-
Estimated saving: the equivalent of 3+ headcount freed from support
-
Speed to complete tasks (such as HMRC submissions) 200% improvement
-
Law graduates doing legal work, not product support
SeedLegals:
GTM co-ordination for a big bet
The opportunity
Every start-up has big bets. For SeedLegals, one of them was adapting the UK product for the US market.
Others had tried and failed. We spent a year validating that we had something that could work before we committed. Once we were confident enough to go, we had to get the go-to-market right.
Entering a new market isn't just a product and delivery challenge. Sales readiness, CX capacity, marketing, localised copy, pricing, positioning - all of it needs to be in place at launch, not assembled afterwards.
What I did
I created a structure that made coordination possible: shared GTM plan, dedicated PM role, defined cross functional lanes.
Operating model design for a high stakes launch
I brought together product, engineering, marketing, sales and CX around a unified GTM plan. Each function understood the target customer, their positioning, and their lane.
I put a dedicated project manager in place - new for SeedLegals - specifically to hold the GTM coordination together, which freed the product manager to focus entirely on getting the build right. Post-launch, the product manager stayed on to run product iteration with the CEO.
Impact
-
A coordinated US launch, with all functions ready at go-live
-
The product has been live and iterating for over a year
-
Almost at break even - lean team, costs controlled, market being tested properly
-
A template for how SeedLegals approaches big bets now
SeedLegals:
Building a design function - small but perfectly formed
The problem
When I arrived at SeedLegals, designers worked in isolation from the development teams, completing work in full before handing it over - almost a waterfall approach. Design reviews happened at the end, which meant a single CEO session could unravel weeks of work. The team had learned to be cautious of reviews.
There were no shared artefacts: no personas, no customer lifecycle maps, nothing the team could point to and say "this is who we're designing for." Usability testing had been sporadic and used expensive online platforms who don't actually have our customer base.
The brand was fun but aging fast and - if we're honest - a little goofy.
What I did
This is my bread and butter. I analysed how they were working versus how they should be. Startups notoriously cluster inexperienced but enthusiastic people together to achieve a goal and, after a while, they need some help with the direction for delivery optimisation - while still brinign the innovation layer.
Integration. Embedded the Designers in the development squads. We create the triad - PM, tech lead and designer - per team; to work together on problems.
Focusing on UCD - We built a customer research panel to facilitate usability testing. Work with the CX team - who knew the customers better than anyone. I built a full set of design artefacts: personas, the founder and investor lifecycle.
Tools/Speed to deliver
We built a pattern library that underpins the product's visual consistency and reduced the time from design to delivery.
More recently we've layered an AI prototyping capability on top of it: designers can now produce realistic, interactive prototypes quickly, which means stakeholders can feel a feature before it's built, and engineering gets stories with enough fidelity to work from cleanly.
Processes Feedback arriving at the end of a design cycle is expensive: changes are late, rework is high, and the team builds in the wrong direction. I moved CEO involvement earlier: vision-setting and co-design at the start rather than critique at the end.
Brand cohesion- company wide For two years I made the case for bringing in a dedicated visual designer. When I eventually got the hire, one person transformed how the brand shows up across every channel. She has created consistency across the product, the website, social and marketing - and built templates so the marketing team can produce content that holds to the same visual language without needing to go back through design every time.
Impact
-
Ultimately the team barely need me for day-to-day decisions now. I am their sounding board.
-
The team is faster, more confident, and better connected to the business.
-
Design is collaborative with significantly less rework. Everyone can contribute.
-
Prototyping and vibe coding is on brand and can be handed over to engineers
-
Onboarding of new employees is faster as they understand the customer and the lifecycle.
-
Brand maturity and consistency across channels. You can recognise SeedLegals without seeing the logo.
Furniture revolution
Marks & Spencer: Online and store: Reinvent the furniture dept directly increasing revenues
The problem
Marks and Spencer has a Home business – part of which is selling sofas. Online interaction was low and fewer and fewer stores were selling the full range. The entire service needed an overhaul – beyond purchase; from sending swatches to delivery.
What we did
I put a small crack team of 2 designers to work collaboratively with everyone involved with the furniture chain. Our user researchers went into store and uncovered the personas and their buying behaviours.
We brought all the departments together - from buyers to logistics - shared the research and workshopped a service blueprint map to find problem areas and identify opportunities. We then sized the work needed to rectify and the group prioritized what to work on first.
Developers worked with the designers to create a sofa configurator and make it easier to choose. We knew from the research what information they needed – including “will it fit” and how to measure.
Impact
-
Customers could view in store but then make decisions later. We introduced digital tools in store to help customers visualize their new sofa – while they could touch the fabric and bounce on the seats.
-
From release there was 16% uplift in conversion (£2.5m - £3m incremental sales pa)
-
Bonus - happy connected teams who had never worked as a cross-business collaborative team before. It was fast and rapid - unheard of in a behmoth like M&S
-
Side effect - through this work we realised that pricing for larger options had been incorrectly calculated, losing the company money - so - that got rectified!
Argos:
High revenue impact through cross business collaboration & interpreting user data correctly
The problem
The product listing page for the business was cluttered with ads and deals. They felt it was the best place to promote the BIG deals. (It's not - it never is - that's like walking through the door of a shop and having 20 leaflets thrown at you - most of which have nothing to do with why you came here today).
For users it’s a navigation page – it signposts them to where they want to go.
Research (quant & qual) showed this was not the opportune place to place deals. The sweet spot for deals is while the user is solving their specific need.
What we did
User tracking showed people clicking on the deals – showing intention & obviously a need for the promotions. The business felt that proved it was the right design.
Users' (in lab) feedback told us they weren’t interested & it was distracting. They wanted relevant and in context deals – ie while they are purchasing.
We collaborated and compromised – limited to 2 ads at the top and 6 at the bottom. Made it easier and quicker to get the users to their destination and still allowed the business to promote. (Note: Argos users are task focused - it's not a destination site - so you have to make sure you promote at the right time, in the right place. They won't proactively come tothe site to seek them out except on specific sale type days - Black Friday etc. ).
Impact
-
A/B tested the new approach on 2 of the least popular categories to see if it worked.
-
It had a 23% uplift in basket – extrapolated across all categories this would increase revenue substantially across the year.
-
Reprioritised the backlog to update all PLPs to provide clear purchase pathways across all categories – directly driving bottom line revenue - by hundreds of thousands!
BTFG:
Transforming multiple legacy systems into 21st Centrury Web and mobile app
The problem
BTFG financial advisors used multiple spreadsheets and legacy software to pull together their customers' investments and pensions.
Panorama was a full scale implementation. It brought all of the investment products for both advisors and their clients into a single webapp and mobile app with a complete rebrand.
What I did
I ran a multidisciplinary team – UX, UI, VD, Research, Copy across 26 streams. It was in total a 4 year project to the first launch and then ongoing iterations. A consistent, simplified, intuitive - and - I would say - beautiful interface.
I advocated Agile methods & successfully worked to embed them across the organization.
I drove design led methods – and we covered the walls of the - previously strict - banking environment in so much paper they had no choice but to be enveloped in the process!
Impact
Full suite of wealth management tools in one place for BT customers - the first of its kind. It included your investments, pension, trading and all your financial recommendations.
Consolidation means tighter stickiness for customers - more useful recommendations which means more efffective upsell / cross sell, increasing basket size and overall revenue per head and CLV.
A suite of easy to access tools for Financial Advisers to forecast and recommend investments with prospects and customers.
“At last I have something I can show my clients”
SEEK: Removing silos - aligning digital with marketing for consistent touchpoints
The problem
Arriving at SEEK - an iconic dot com brand in Australia - the website was completely disconnected from the brand.
The brand is strong and humorous. The experience on the website was devoid of it's personality and had wandered down a strange illustrative route.
(I had the same issue with Marks and Spencer - Marketing never considered the digital channels with their new campaigns. And M&S campaigns are BIG. Digital was always an after thought - and they never understood that it wasn't just a case of swapping out images).
What we did
Aligning with the marketing team to understand the brand values, we refreshed the website, removing those elements that didn't portray the brand effectively.
The best part was taking some of the current campaigns and using them for 404 and error pages.
(at M&S - I worked with Marketing so we were included at the beginning of campaigns and on our side, we made sure digital had the flexibility to adapt and showcase/integrate the campaign)
Impact
Instead of irritating users when pages didn't load - we made them smile.
Consistency of brand across the touch points; a website with some personality and, internally, cross functional harmony.
(For M&S the impact was more substantial - it meant a consistent approach for campaigns across stores, media and digital. It was a large project - making the digital channels flexible to adapt to imagery, tags and themes without a full redesign).
Stanford Medical: Helping frontline Covid medical staff
The problem
Stanford Medical wanted an app for frontline covid staff so they could recalibrate or use techniques to help with their mental health.
Beyond a wellness app this was based on cognitive and behavioural psychological techniques.
What we did
I did the wireframes (in the UK) and worked with the design firm (in India) while working with the client in the USA.
I was brought in to help rescue the app – the development team (also in India) hadn’t manage to realise an effective workflow.
It’s one of my favourite projects – because you don’t often get to be part of an app that is doing good. It also used my psychology background and that helped me gel with the team in Stanford and empathise with what we needed to achieve.