+47%
Dynamic package bookings, FY22 vs FY19
+53%
Profit growth vs FY19
3m
People sent on holiday in 2022
-1%
Market growth over the same period
Between FY19 and FY22 the package holiday market shrank by roughly 1%. loveholidays grew 35%, matched the passenger volume of the category leader it had spent years chasing, and lifted profit 53% above pre-pandemic levels, after a period when holiday searches dropped 80% and cancellations peaked at 95%.
The company had professionalised. The brand had not caught up.
| Company | loveholidays, a UK package holiday OTA |
| Role | Lead Product Designer |
| Team | Director of Design and research, with branding agency Koto on identity and Message Mechanics on positioning |
| Timeframe | October 2022 – May 2023 |
The problem: The brand still read as the scrappy disruptor, and gave teams nothing to make decisions with. Marketing, product, content and partnerships were each interpreting the company for themselves.
The goal: One brand system that worked as hard inside the organisation as it did in an ad: recognisable to customers, and operable by the teams building the product.
My scope: Design lead on our side of the brand development, and owner of the implementation across the product experience.
The internal framing was three phases: startup before 2019, scale-up from 2019 to 2022, and scale-out from 2023 onwards. Each phase had broken something the next one had to fix. The startup years built a fast, founder-led selling machine with real structural risk underneath it. The scale-up years fixed the platform, the process and the people.
Scale-out was the phase where brand started carrying weight. Three horizons were running at once, optimising the core, expanding the product and expanding internationally, against macroeconomic headwinds, a smaller marketing budget than competitors and a smaller team. Customers were dealing with a cost-of-living crisis and limited annual leave; they found choosing a holiday complicated.
That combination made this a brand system problem rather than a campaign. Fewer people, less money, more surfaces, more markets. Everything the brand produced had to be reusable.
1. The company had no shared definition of itself
The first task was not visual. It was agreeing what the company was, in language specific enough that different teams would make the same call without checking. That work was done with Message Mechanics, and it had to settle before any of the identity work began.
The framework came out as a single stack:
- Mission: to open the world to everyone
- Vision: to become Europe’s #1 holiday provider, by building the most loved packaged holiday experience
- Proposition: the smart way to get away
- Pillars: unlimited choice, unmatched ease, unmissable value
- Personality: friendly, positive, assuring; a personal holiday expert rather than a discounter
The three pillars are the useful part, because a page, a campaign or a feature can be argued against them. That is what separates a brand statement you can use from one you can only admire.
Make it Count
The internal rallying line, chosen because the constraint was real: less budget and a smaller team than competitors, against bigger ambitions
2. A rebrand is the easiest place to lose brand equity
loveholidays had built up real equity in a particular kind of warmth and optimism over years of trading. The temptation in any rebrand is to trade that for whatever looks contemporary, spending a decade of recognition to buy a year of feeling modern.
We held three things as non-negotiable continuity: the joy and optimism of a holiday, colours that were both ownable and accessible, and a clean, simple wordmark. Holding that line through the development was a job someone had to do.
3. Nothing would hold unless it survived the product
A brand refresh gets attention once. A brand system has to survive repeated use, and the hard test is the confirmation screen, the service reply and the empty state rather than the billboard.
The product already had design foundations. Any new brand had to land against them, and the gap between a specified type ramp and the one already sitting in the codebase is where a rebrand reverts to what was there before.
My title and remit: Lead Product Designer. Design lead on our side of the brand development, and owner of the implementation across the product experience.
On the brand development. I worked with branding agency Koto alongside our Director of Design and the research team. I sat in the steering meetings where direction was set and signed off the briefs before they went out. Most of my contribution at that stage was holding the returning work against what loveholidays was, making sure it read as an evolution of the brand rather than a replacement, and that what came back could be built.
On the implementation. I audited how the new guidelines would land against the design foundations already in the product: what mapped cleanly, what conflicted, and what would need rebuilding rather than restyling. Then, as teams rolled it out, I checked what shipped against the brand approach rather than assuming a document would enforce itself.
Why this needed a lead. The brand work and the product work were being done by different people to different deadlines, and neither group was positioned to see the other’s constraints. Someone had to sit in both, able to argue for brand continuity in a steering meeting and to spot a drifted component in a live search result. That is the role I was in.
Evolve the identity, don't replace it
The brief was to stand out in a crowded marketplace without throwing away what people already recognised.
What we built was a system rather than a logo:
- A primary and secondary palette, held to accessibility standards rather than picked for looks
- Typefaces with enough range to carry both a billboard headline and a booking confirmation
- Iconography and a graphic device, the yellow sun-ray shape that scales from a favicon to a tote bag
- A photography direction split into people and contextual, so briefs had a rule instead of a mood board
The device does the most work. It is simple enough to be cropped, rotated and printed at any size, and distinctive enough that a blue field with yellow rays reads as loveholidays with no wordmark present at all.
Write guidelines that specify behaviour, not assets
The guidelines turned all of this into something other teams could run without asking. Version 1.0 ran to just over a hundred pages across nine sections: strategy, tone of voice, logo, colour, graphic devices, iconography, typography, art direction and brand examples.
The useful half specified behaviour rather than assets. Colour ran from two core brand colours out to a secondary palette of tints, and then to a separate UI palette built for the product, adding functional success, warning, critical and information states plus seven surface tints a marketing layout would never need. Colour usage came with ratios per surface, so balance was a rule rather than a judgement call. Tone of voice worked the same way, with each surface given its own tonal mix. The devices were documented as mechanics tied to behaviour, and the lozenge stands in for search.
That UI palette is the seam my implementation work sat on, the point where a brand document starts making commitments a product team has to honour.
Make the system survive the unglamorous surfaces
The system had to work across paid and out-of-home campaigns, the app and booking flow, post-booking and service comms, partner communications, and internal decision-making. Marketing needed range, product needed clarity, and leadership needed enough shared language to hold a bigger market position without re-litigating it every quarter.
Two workstreams came out of that pressure. We commissioned content guidelines, because the rebrand had produced positioning, tone of voice and visual design but no copy style guide: grammar and formatting, accessibility and inclusivity, and writing for UX, CRM, social, service and legal. We also set up brand tracking as longitudinal research every eight weeks, measuring awareness, usage, attitudes, perceptions and purchase intent, so anyone arguing for brand-building could do it with evidence rather than taste.
Launched company-wide, then load-bearing for a new market
The brand went out company-wide at a Strategy and Brand Day in October 2022: the whole business in one room, with strategy and brand presented together rather than as separate announcements. By the Product and Tech all-hands at the end of January 2023, “we have a new brand” was listed alongside the platform facts: a 95-person product, tech and data team, more than 100 services, over 500 billion holiday offers processed daily.
The real test came next. loveholidays launched in Germany, opening for test bookings on 31 January 2023 and fully launching on 1 May. Before that, the team produced 176 slides of German market research covering segmentation, booking behaviour and accepted design patterns, and built the market on a configurable platform so the next one would cost less.
A brand designed six months earlier had to translate into a language, a market and a set of expectations it was never drawn for. It held because the work had gone past the logo into rules other teams could apply on their own.
Where the brand stands now. The tracking set up at launch has run every eight weeks since. The most recent UK read puts prompted awareness at 47% and consideration at 18.7%, each up 0.3 points on the previous quarter. That says where the brand is rather than what the rebrand caused: the tracking began with the relaunch, so there is no clean pre-rebrand baseline to measure against.
The brand only worked once we treated it as infrastructure
The identity is not what I took away from this. The brand became useful at the point it started carrying organisational judgement: what to build, what to say, what to repeat and what to refuse.
The palette, the pillars and the tone of voice all worked as decision shortcuts. They let a team smaller than the competition make more consistent choices, faster, across more surfaces, without a central group signing off each one.
Auditing the implementation taught me that. You can only tell whether a brand is doing that job by checking the surfaces nobody art-directs. The limit is worth stating too: a brand system does not create demand. It stops a growing organisation from spending scarce attention re-deciding what it has already decided.
What I’d do differently: set the implementation baseline before launch, not after. I audited against foundations I understood well, but without a documented before-state I could describe what had drifted more easily than I could prove how much. On a programme this size that measurement should have been a prerequisite, not an afterthought.