Rationale: Real Trustpilot reviews show users are repeatedly confused by Ferryhopper's "Flex Cancel" add-on. One reviewer wrote that it was nearly impossible to understand that the fee structure wasn't a full refund. Another complained the 48-hour rule wasn't disclosed before purchase. This is a live, real-world UX copy problem that maps directly onto every core skill in the job description: simplifying complex information, auditing existing copy, reducing support tickets, and bilingual delivery (EN + EL).
Deliverable: A content audit of Ferryhopper's cancellation and flex ticket screens, followed by a rewrite of the most problematic copy in both English and Greek.
Phase 1 — Research & audit
I went through the Ferryhopper booking flow in both English and Greek myself. I took screenshots of every screen where cancellation terms, the Flex Cancel add-on, and refund policies appear. I also created an audit spreadsheet with columns for: Screen name, original copy (EN) (GR), my assessment of the problem, severity (high/medium/low), and my recommendation.
Then I looked at third-party signals. I found and read at least 40 Ferryhopper reviews mentioning cancellation, refund, or the flex option. I run a sentiment analysis, whose preview can be found here. categorized what users misunderstood. This gives you a research foundation — even without proprietary data, you can cite patterns from public reviews and frame them as "user signals."
Identify three to five specific problem areas. Based on what's publicly visible, likely candidates are: the Flex Cancel label and description at the point of purchase, the error state when someone tries to cancel within the 48-hour window, the email confirmation copy around what "Flex" actually covers, and any empty state or helper text around the refund timeline.
The basic findings of the sentiment analysis are summarized in Figures 1 and 2:
Sentiment distribution
Sentiment trend
Overall, Ferryhopper reviews lean positive (23 reviews) but have a substantial negative chunk (17), with very few neutrals (5). The peak positive period was September 2022, but negativity stayed persistent throughout. Common complaints are around customer support and cancellations, like the customer EliHughes who only found out about a cancellation at the ticket office and paid an extra €150 on the spot, or the customer Leonidas who waited 3 days for a support reply. On the flip side, happygoose751 praised the team for fixing a wrong booking without drama.
I further identified the most common complaint themes, which are summarized in Figure 3:
Most common complaint themes
Figure 3 makes it pretty clear: customer support is the #1 pain point by a mile (6 mentions), followed by delays (4). Cancellations and booking/ticket changes tie at 3 each, with communication rounding things out at 2. These all connect: the customer sadcat506 complained about constant delays and zero heads-up on schedule changes, while Leonidas waited 3 days for support only to hit a dead end. Essentially, when things go wrong, Ferryhopper's response system makes it worse.
Phase 2 - Rewrite
For each problem area, departing from the "before" (screenshot or paraphrase of original), I wrote a proposed "after" in both English and Greek in the form of a side-by-side bilingual comparison.
The principles that guided my rewrites, specifically for this context, include the following:
1) The use of plain numbers and concrete conditions rather than adjectives ("Cancellable up to 48 hours before departure" instead of "Flexible cancellation")
2) The making of the fee explicit before the user commits ("A €5 service fee applies — the remaining amount is refunded"), and
3) The front-loading of the limitation in error states ("This ticket can no longer be changed—departure is in less than 48 hours").
For the Greek versions, it is very important to consider the nuances of the Greek grammatical register, which is appropriate for a travel booking context. In line with this, and given the friendly and direct tone of Ferryhopper's website, I made the distinction between formal vs. informal second person (εσείς vs. εσύ), so εσύ is right, but below I explain my conscious and documented choices.
Examples of rewritten texts with rationale (ENGLISH)
Examples of rewritten error messages
Examples of rewritten Frequently Asked Questions (FAQs)
Examples of rewritten texts with rationale (GREEK/ΕΛΛΗΝΙΚΑ)
Examples of rewritten error messages
Examples of rewritten Frequently Asked Questions (FAQs)
Localization annotation
Based on this project, I have created a localization annotation layer, which maps seven EN→GR copy rewrites across three dimensions: UI impact, translation choices, and language-manager flags. On length, Greek strings run 7–16% longer than English on average — manageable in body text but worth testing in constrained UI elements like inline error boxes and FAQ preview teasers. On idiom, several EN phrases had no clean Greek equivalent ("up in the air", "before you sail", "reach you"), requiring sense-for-sense rewrites that preserve tone over literal accuracy. The single highest-priority flag across the full set is register consistency: all Greek strings default to informal singular (εσύ), but if Ferryhopper's style guide mandates formal plural (εσείς), every imperative verb and possessive pronoun needs updating across both error messages and FAQ strings simultaneously.
What I'd measure
To validate these rewrites, I'd track support ticket volume tagged to cancellation and booking changes, the top complaint themes, looking for a measurable drop post-launch. For the checkout error messages, I'd monitor abandonment rate at the payment step. For the Flexi/Flexi Cancel copy, conversion rate on the add-on and post-purchase comprehension scores from usability testing would show whether clearer conditions actually reduce expectation mismatch and the disputes that follow.