A new proposition, no playbook, and a role that grew to fill it
Senior leadership asked a cross-functional project team to launch Roll, a 30-day rolling broadband contract aimed at students, healthcare workers and home movers. There was no precedent, no template and no defined design lead. I joined as the team's digital and design specialist to handle the website, and ended up owning the entire creative and digital delivery: the page and persona journey design, the campaign's visual identity across web, social and print, the enquiry route, conversion tracking, and the reporting that connects an online enquiry to a confirmed sale.
Roll · hub plus three persona journeys · live on kcom.com
A concept from the top, and a wide-open remit
Roll started as an idea from senior leadership: a 30-day rolling broadband contract at £49.99 a month with no upfront fee and no long tie-in, built for people who need speed without commitment. It was a genuinely new proposition for KCOM, so there was no launch playbook, no design language, no page templates and no agreed way to measure success.
The project team was deliberately cross-functional: one person each from marketing, product, sales and customer improvement, and me from the digital team. That mix covered the commercial side of the launch well, but it meant I was the only designer and the only digital specialist in the room. Every visual, every journey and every measurable touchpoint of Roll would come through me.
My job on paper was the website. In practice, the project had gaps that sat between everyone's job descriptions: who defines how Roll looks and feels, who designs the audience journeys, who owns the creative across channels, who proves the campaign works. I took those gaps on deliberately, because each one was work I wanted to be doing and could evidence. By the end I was leading the creative and digital delivery across the team and reporting performance directly to the people who commissioned the idea.
A promising idea with nothing built underneath it
Roll had a name, a price and a clear commercial logic. Everything else needed designing and building. Three problems defined the scope:
No identity and nowhere to live
Roll had no visual language, no page, no URL and no place in the site architecture. Before anything could launch, the proposition needed to look like something: a design system that could flex from a phone screen to a printed leaflet, and a home on the site that could scale if the proposition grew.
Three very different buyers
Students, healthcare workers and home movers all want the same contract for different reasons. A single generic page would have flattened those motivations into one message and converted none of them well. Each audience needed its own journey, designed around its own motivation, without tripling the design and maintenance overhead.
The sale happens where the web team can't see it
Roll enquiries convert through telesales, offline. Without deliberate tracking design, every online enquiry would disappear at the point of handover and the campaign would be impossible to evaluate. Solving this became the most technically valuable part of the project.
One visual system, from phone screen to print
Before the first page was built, I designed the system everything else would use.
The visual identity. Working within KCOM's brand, I developed Roll's campaign identity around a single idea: freedom without friction. The rolling-contract scroll, the flame motif and the recurring arrow gave the proposition an instantly recognisable visual shorthand, bold enough to stand out in a social feed and simple enough to survive die-cut print production. The "No strings, just speed" message hierarchy carried through every asset.
Designed for every surface at once. Roll launched simultaneously across web, Meta social ads, print leaflets and out-of-home, so I designed the system format-first. Core assets were built in Illustrator and Photoshop to scale between a 1080px carousel slide and a physical leaflet without redrawing, and the layout grid on the web pages mirrors the hierarchy of the print pieces. Someone who saw the ad, picked up the leaflet and landed on the page experienced one product, not three campaigns.
Persona journeys as a design problem. The student, healthcare and mover journeys share one template but lead with different imagery, motivation-led headlines and proof points. Designing the difference into the content layer rather than the structure kept the set consistent, maintainable and fast to extend.
Designed to clear sign-off. Every asset was built from pre-approved brand copy and components, a deliberate design constraint that meant the full creative suite cleared brand and compliance sign-off cleanly, without rounds of rework eating into the launch window.
One proposition, built in deliberate layers
I sequenced the work so that each layer supported the next: identity, page, journeys, enquiry, measurement.
The Roll page, designed and built
I designed and built the core Roll product page end to end in Umbraco: hero, proposition, pricing and the route into enquiry. As both the designer and the builder, there was no handover loss; what I designed is what shipped, and I iterated it against performance data after launch rather than leaving it as a one-off.
A hub plus three persona journeys
I designed a Roll hub with dedicated journeys for students, healthcare workers and home movers. Each journey leads with that audience's actual motivation (term-time flexibility, shift patterns and relocations, moving without a 24-month commitment) and shares a common design template so the set stays coherent and maintainable.
An enquiry journey that knows who is asking
Each persona journey feeds a Jotform enquiry route that carries the persona through hidden fields. When an enquiry reaches telesales, the team already knows which audience it came from, and so does the reporting.
The campaign creative across every channel
I produced the full creative suite: three Meta carousel ad sets tailored per persona, a die-cut print leaflet for events and letting-agency drops, and the supporting web assets, all from the visual system above so every touchpoint reinforced the same product.
Conversion tracking, then reporting
I implemented GA4 event tracking across the whole journey: UTM capture, enquiry starts, form submissions and call clicks, each verified in DebugView before anything went into a report.
Closing the gap between the click and the sale
The hardest technical part of the project was measurement. Roll converts through telesales, so the standard web analytics picture ends at the enquiry. I designed the tracking so it doesn't.
Every enquiry carries its source with it. UTM parameters and persona identifiers pass into Jotform through hidden fields, so each submission arrives at telesales with its origin attached. Call clicks fire their own events against the campaign's 03 numbers. Form submissions confirm through the dataLayer rather than URL guesswork.
The whole thing was built to be simple and reliable.
Every mechanism uses standard, supportable tooling (GA4, GTM, Jotform hidden fields) so it keeps working without me babysitting it, and so anyone auditing it later can follow exactly how a number was produced.
Instrument first, scope tightly, verify, then build
The reporting request arrived as three words: "deliver a report." I turned that into a defined piece of work.
Define what "deliver this" actually means
I scoped the report with the team before building anything: which questions it answers, which numbers are canonical, and what it deliberately excludes.
A journey per buyer, one report
The report breaks performance down by persona, matching the structure of the site and the creative, so the business can see which audiences respond rather than one blended figure.
Tracking before reporting
I refused to report on events I hadn't verified. Every metric in the report maps to a tracked, tested event.
Confirmed in DebugView, then published
Each event was validated in GA4 DebugView with real journeys before it appeared in any output.
Stitch the funnel, label the limits
The report follows an enquiry from source to telesales outcome via the Jotform trail, and it states its own limitations plainly: where matching is confirmed, where it is probable, and where the data cannot say.
A measurable channel, and a model the business reuses
Roll moved from a leadership idea with nothing behind it to a live, fully designed and instrumented acquisition channel with persona-level attribution, and the design system and reporting model are now available for every future campaign. The clearest proof of the model in numbers so far comes from the Black Friday landing page, built on the same campaign-page approach.
Roll: from concept to channel
The shift isn't a single headline metric, it's an idea turned into a product the business can actually run, read and grow.
| Dimension | Before | After |
|---|---|---|
| What existed | A name, a price, an idea | A full visual identity, live hub and three persona journeys |
| How customers reach it | Nothing in place | Web, SEO, social, print and OOH, one design system |
| Audience targeting | Undefined | A designed journey per buyer |
| Enquiry attribution | None | Persona-level, automatic |
| Online to offline visibility | Disconnected | Jotform trail from enquiry to sale |
| Creative and delivery ownership | Unassigned | Single owner end to end |
The Black Friday page validated the wider campaign-landing-page model so clearly, including a backlink earned organically when an industry news site picked it up before launch, that it's now embedded as standard practice in the FY27 SEO strategy.
One campaign, every surface it had to live on
Roll launched across web, social, print and out-of-home at the same time, so the creative had to hold together everywhere from a phone screen to a poster. I designed one visual system, one message hierarchy and one set of master assets, then adapted them per format rather than redesigning per channel. The section below shows the system in situ: the persona journeys live on kcom.com, the die-cut leaflet, and the Meta carousel ads.
Print leaflet · die-cut
Meta ads · carouselThe detail that made it mine
Persona-aware enquiry capture
Hidden fields carry the persona and source through Jotform automatically. Telesales sees context; reporting sees attribution. Nobody has to remember to do anything.
Designed and built by the same hands
Because I both designed the pages and built them in the CMS, nothing was lost in handover and design intent survived all the way to production, down to spacing, type hierarchy and interaction detail.
Verified before published
Every tracked event went through DebugView with real test journeys before it fed a single reported number.
Documented for handover
The whole setup, from the design system's asset library to event names and hidden-field mappings, is documented so the next person can maintain or extend it without reverse-engineering my work.
From a launch to a repeatable growth engine
Reporting is live
Performance by persona goes to the team on a regular cadence, built on verified events.
Optimisation is ongoing
The persona journeys are now a testing ground: design, copy and CTA changes measured against real enquiry data.
Match rates are monitored
The Jotform-to-telesales trail is tracked over time, and confidence in the funnel grows with the sample size.
The model is a template
The visual system, the hub-and-persona structure, the hidden-field attribution and the instrument-first reporting approach are all reusable for KCOM's next proposition launch.
The toolkit behind the work
A pragmatic stack chosen to keep the build self-serve, the tracking reliable, and the reporting honest, with no developer dependency for the measurement layer.
The lessons that outlast the project
Gaps are a route into where you grow
The unclaimed work on this project (identity, journeys, creative, measurement, reporting) became the most valuable experience I took from it. I now look for those gaps deliberately.
Design the system before the pages
Building the visual identity and template logic first meant three persona journeys and a full multi-channel creative suite could ship in one launch window, and stay consistent afterwards.
Fix the measurement before you report on it
A report built on unverified events is a liability. Instrumenting first cost days and saved the credibility of every number that followed.
Owning something means documenting it
Ownership that lives in one person's head is a risk. Writing the setup down, from asset library to tracking config, turned my work into infrastructure the business keeps.