OTA Commission vs Building a Website: A Five-Year Comparison

“A website costs too much” is a reasonable objection when the website is priced against nothing. Priced against five years of commission on the bookings it would have captured, the comparison usually reverses — but only if the model is honest about how much revenue actually shifts.

This article builds that model with deliberately conservative assumptions, in dirhams, for three real Moroccan business shapes. The arithmetic is shown so you can substitute your own numbers.

The structural difference

Commission is a percentage of revenue, forever, rising as you grow. A website is a fixed cost with ongoing maintenance, flat as you grow. That asymmetry is the whole argument: the more successful you become, the more the commission model costs you and the less the website model does, per booking.

The assumptions

Kept conservative on purpose:

• Blended platform take-rate: 21% — derived using the method in the real cost of OTA commission, not the 15% headline.
• Direct-share shift: 10 percentage points over five years. Modest. Well-executed properties do considerably better; this keeps the model defensible.
• Website cost: a custom build plus five years of hosting, maintenance and content work.
• Direct bookings are not free: payment processing at 2% is charged against direct revenue.
• No revenue growth assumed. Flat revenue across five years, which understates the benefit.

Scenario one: a twelve-room Marrakech riad

Annual revenue 3,000,000 MAD, of which 85% currently arrives through platforms.

• Platform revenue: 2,550,000 MAD × 21% = 535,500 MAD/year
• Five-year commission at current mix: 2,677,500 MAD
• After a 10-point shift: platform revenue 2,250,000 MAD × 21% = 472,500 MAD/year. Direct payment fees on the extra 300,000 MAD = 6,000 MAD.
• Five-year cost after shift: 2,392,500 MAD
• Gross saving: 285,000 MAD
• Website build + 5 years maintenance: roughly 120,000 MAD
• Net five-year benefit: about 165,000 MAD, with break-even around month 26.

Scenario two: a desert camp near Merzouga

Annual revenue 1,400,000 MAD, 90% through Viator, GetYourGuide and agencies at a blended 26% — experience platforms take more than accommodation platforms.

• Platform revenue: 1,260,000 MAD × 26% = 327,600 MAD/year
• Five-year commission: 1,638,000 MAD
• After a 10-point shift: 1,120,000 MAD × 26% = 291,200 MAD/year, plus 2,800 MAD payment fees
• Five-year cost after shift: 1,470,000 MAD
• Gross saving: 168,000 MAD
• Website build + maintenance: roughly 90,000 MAD
• Net five-year benefit: about 78,000 MAD, break-even around month 33.

Slower than the riad, because the camp’s revenue is smaller — but the higher experience-platform commission means each shifted dirham is worth more. The allocation logic is in Viator vs GetYourGuide.

Scenario three: a Marrakech tour operator

Annual revenue 4,500,000 MAD, 75% through platforms at 25%, with a substantial share of high-value private and multi-day circuits.

• Platform revenue: 3,375,000 MAD × 25% = 843,750 MAD/year
• Five-year commission: 4,218,750 MAD
• After a 10-point shift: 2,925,000 MAD × 25% = 731,250 MAD/year, plus 9,000 MAD payment fees
• Five-year cost after shift: 3,701,250 MAD
• Gross saving: 517,500 MAD
• Website build + maintenance: roughly 150,000 MAD
• Net five-year benefit: about 367,500 MAD, break-even around month 18.

The operator benefits most, for a structural reason: custom multi-day circuits are exactly the product platforms template badly, so the direct shift is easiest to achieve there.

What the model deliberately leaves out

Every omission below makes the website look worse than it is, which is the point:

• Owned guest data. Five years of consented contacts producing repeat and referral bookings at near-zero acquisition cost — see who owns your guests.
• Products platforms cannot list. Whole-property rentals, weddings, retreats, corporate and MICE work.
• Risk reduction. The value of not being one account suspension away from zero revenue — reducing OTA dependency.
• Pricing freedom on direct inventory.

When the answer is genuinely no

If your annual revenue is under roughly 400,000 MAD, or you have no capacity to answer direct enquiries promptly, or your occupancy is already near capacity through platforms with no seasonality problem, a custom build may not clear its cost. In that case a small, fast, honest site focused purely on branded search — the billboard effect problem — captures most of the benefit at a fraction of the price.

Run the model with your own figures. The three inputs that matter are your real blended take-rate, your platform revenue, and how many points of direct share you can honestly expect to move.