From flat spend
to targeted
budget.

Five numbers that summarise the whole story — the rest of the page explains how they were achieved.

The "1,213" figure covers all bookings recorded between August 2023 and June 2025. The 31.58% conversion rate is calculated across all 16 months with enquiry data.

The client is a young travel startup operating for approximately two years, specialising in a narrow, well-defined customer segment. The existing marketing model allocated budget equally across every month — a flat spend that ignored seasonal demand patterns, destination performance, and customer behaviour.

I analysed 18 months of transactional and enquiry data to build an analytical pipeline that revealed how customers actually behave — which destinations they book, how far in advance they plan, which months convert best, and how age segments differ. The result is a data-driven allocation model that replaces guesswork with measurable evidence.

The company had no substantiated view of audience behaviour, seasonal demand fluctuations, or the effectiveness of individual destinations. The marketing budget was distributed in equal monthly instalments — wasting resources during low-demand periods and failing to capture potential customers during peak decision windows.

The question: could 18 months of transactional and enquiry data reveal a spending pattern that actually matches customer behaviour?

The study is based on 18 months of operational data from two internal sources: sales reports (transaction records, departure dates, prices, commissions, group sizes) and enquiry forms (submission dates, selected destinations, demographic data).

The dashboard was built around five analytical blocks, each answering a specific business question about marketing effectiveness and budget optimisation.

04.1 — Destinations

Turkey dominates with 268 bookings — more than 2x the runner-up Thailand (113). The top 3 destinations account for nearly half of all bookings, while the bottom 5 generate only around 10% combined.

04.2 — Booking window

This agency's clients are classic "planners" — 808 sales (67%) occur more than 121 days in advance. Last-minute bookings (0-7 days) represent just 0.25% of sales.

Two groups dominate the customer base: 35-44 (89 people, 38.7%) and 45-54 (78 people, 33.9%), together forming approximately 70%. This is critical information for positioning marketing messaging and visual design.

The enquiry-to-conversion relationship reveals a clear inverse pattern: when enquiries peak in summer, conversion drops. The highest enquiry volumes occur in February (80) and April (72), with conversion swinging from 55% in June down to 12% in August.

Spring (Jan-Apr) is the optimal window for both volume and conversion (58-73%). Summer conversion drops from 55% (June) to 12% (August).

Conversion efficiency measures how many enquiries turn into actual bookings for each destination. A high score means every marketing euro spent returns bookings. A low score means you're paying for leads that never convert.

Data Analysis · Business Intelligence · Marketing Strategy · Local AI Infrastructure

• Collected and cleaned 18 months of operational data from two internal sources
• Anonymised all personal identifiers before processing
• Built SQL aggregation pipeline using Apache Superset
• Designed and implemented Tableau visualisation dashboard
• Analysed 5 analytical blocks: destinations, booking window, demographics, seasonality, efficiency
• Identified seasonal patterns and conversion trends
• Delivered data-driven budget allocation recommendations
• Maintained full data confidentiality via local AI infrastructure
• Presented findings and strategic recommendations to the client

• Volume is not value. A destination with many enquiries is not necessarily profitable — Egypt had high interest but the lowest conversion. Always measure efficiency, not just volume.
• Long lead times change everything. 67% of bookings were made 121+ days in advance. This means marketing must start months before the season — last-minute campaigns are a waste of budget for this segment.
• Interest does not equal efficiency. High enquiry counts do not guarantee bookings. The Egypt paradox — high interest, low conversion — showed that a destination can look popular on paper while failing to deliver returns.
• Privacy is a design constraint. Running all analysis on local infrastructure was not optional — it was a core requirement. This shaped the entire approach: local SQL aggregation, local visualisation, no external data transfer.

How I would approach a similar data-driven marketing transformation:

1. 01 — AuditMap current spend and existing data sources.
2. 02 — CollectGather transactional and enquiry data from internal systems.
3. 03 — Clean & anonymiseRemove technical noise, protect personal data.
4. 04 — AnalyseRun SQL aggregation, identify patterns and inefficiencies.
5. 05 — VisualiseBuild dashboards that make the data actionable for decision-makers.
6. 06 — RecommendTranslate findings into concrete budget allocation decisions.

And this is where I work best — at the intersection of business needs and data-driven evidence.

The "1,213" figure covers all bookings recorded between August 2023 and June 2025. The 31.58% conversion rate is calculated across all 16 months with enquiry data.

A flat marketing budget transformed into a measurable, data-driven allocation model.

Have a process, a problem or an idea worth analysing? Write to me and we will turn it into a concrete next step.

Client confidentiality: client identity and commercially sensitive information have been anonymised. This case study presents the methodology, analytical approach and aggregated metrics without exposing confidential client information.