From 30 minutes to 30 seconds: a tour operator's automation story

I've been in travel for 20 years. For almost all of it I've seen the same scene every day: a tour operator with Excel open, email open, the hotel pricing sheet open, the customer brief on one side — putting together a quote. 50 line items: hotels, transfers, tours, guides, museums, meals. A single quote: 2-3 hours. When prices change, everything gets redone.

I started with my own pain

We run Funny Tourism inside DYF Turizm. A Türkiye-based tour operator, B2C + B2B. We felt the same pain every day. A single package entry took 30 minutes — the biggest brake on growing a 25-destination, 33-city SEO inventory.

I asked myself: 80% of this is data entry, 20% is judgement. Can the data entry be automated? Yes.

First internal, then external

We added AI-assisted package entry to the Funny Tourism admin: drop a partner PDF, the form auto-fills. Manual entry: 30 min → 30 sec. 30+ hours back per operator each month. Then we took the engine one step further: TravelQuoteBot. Multi-tenant SaaS. Every operator gets their own supplier catalog, templates, customers.

Customer brief + catalog → personalized itinerary and pricing in seconds. Branded PDF output. We opened it to our own operator first, then to external customers. The anchor tenant tested the engine under real pressure.

What the numbers say

TravelQuoteBot today: 30 organizations, 5 paid tenants, 27 trials. Aggregated inventory: 2,018 hotels, 1,708 suppliers, 544 tours, 331 guides, 309 vehicles across 29 countries. 110 AI itineraries generated. The Q4 2025 launch wave added 11 new operators in a single month.

Anchor tenant confirmed: in-house Funny Tourism is the most active user with 107 events. The top two operators carry 54% of activity. Trial-to-paid conversion sits at 16.7% — improving that is the next move.

Three lessons I take away

First: a product born from your own pain transfers to external customers. We say it constantly — 'it worked for us first'. That's the foundation of our agency model. Second: any job that's 80% data entry is one move away from AI; act before the sector notices. Third: launching a SaaS without an anchor tenant is hard. Make your own operator the first customer; test the product under real pressure.

Which 30-minute job in your company could go to 30 seconds? A 30-minute intro call is enough to find out.