Two engagements, both measured

Net numbers, not attributed. Both name what didn't work.

Case study one · Flights & Holidays · FY21 → FY22

OTA - 6 active markets

From UAE market leader to a multi-market growth engine

Role: Marketing lead — brand, performance, affiliate, social, retention and organic
Team: 22 people, including 6 offshore
Engagement: 12 months

Situation

32 markets, revenue concentrated in two. Grow the core and win new markets without raising acquisition cost or deepening losses.

What changed

Demand: One performance view across paid, organic and social. Market-specific segmentation.
Arrival: Localised messaging per market, not translated.
Decision: Commercial rules that protected unit economics.
Compounding: Affiliate and retention as owned channels. A repeatable market-entry sequence.

What it produced

$14M → $57MRevenue, 4.1x in twelve months

−76%Acquisition cost

2 → 6Revenue markets, of 32 operating

21% → 5%Loss as a share of revenue, with the $3M loss held flat

What didn't work

Retention underdelivered. Desktop and PWA ran as separate instances on a fragmented CMS, so lifecycle could not be integrated. Engineering fixed the infrastructure first — right call, and a year of forecast retention revenue lost.

Case study two · Online travel · 36-month programme

MyFlightSearch

From one Google Ads engine to a multi-channel growth system

Role: Marketing lead — brand, performance, affiliate, social, retention and organic
Team: 26 people
Engagement: 36 months

Situation

80 transactions a day, all from Google Ads. A 43% cancellation rate eroding net revenue, and no second channel.

What changed

Demand: Search rebuilt as multi-campaign. Bing and paid social added. Coverage extended across the region.
Arrival: Platform experience rebuilt around the booking flow.
Decision: CRO programme that halved the cancellation rate.
Compounding: CRM worked as a fourth channel, not an afterthought.

What it produced

80 → 3,700Transactions per day, 46x over 36 months

43% → 21%Cancellation rate

64xNet confirmed volume after cancellations

1 → 4Acquisition channels

What didn't work

Spanish-language acquisition never reached efficiency. Those markets were more price-sensitive than our fare margins allowed, so traffic converted only on the cheapest routes. We cut the spend rather than discount into the segment.

We have run this from the inside

Gross conversions get measured. Net ones get paid. Cancellations, failed payments and post-booking drop-off take about a fifth of apparent revenue, and almost nobody reports it.

Gross conversions — what gets measured

Net revenue — what gets paid

About a fifth — cancellations, failed payments, post-booking drop-off

Proportion as stated above. Varies by operator and route mix.

Questions people ask

Where do OTA funnels lose the most bookings?

Across five steps — search, fare, details, payment and confirmation — and each fails differently. The costliest gap is often price: the price a traveller sees is not always the price they are charged.

Why measure net-confirmed revenue instead of bookings?

Because a booking can still fail at payment or be cancelled. Net-confirmed revenue counts only what is confirmed, paid and not cancelled. That is the number that reaches the P&L, so that is what we optimise.

How do you decide which market to enter next?

We set the order: which market next, in what sequence, and which channel leads in each. Entering a new market is a sequence of decisions, not a translation of the existing site.

Start with the audit

Three weeks, fixed fee. You keep the roadmap either way.

Book the audit