Rate Parity After the DMA: How EU Hotels Test Direct-Rate Differentials Without Losing Ranking

The EU Digital Markets Act forced Booking.com to drop contractual rate-parity clauses in July 2024. For the first time in a decade, European hotels can legally offer lower rates on their direct channel than Booking.com displays — with no contract clause to punish them.

What hoteliers underestimated is that the algorithm kept enforcing parity after the contract stopped. Properties that lowered their direct rate by 10% saw Booking.com impressions drop by 30–60% within 2 weeks. The contractual stick is gone; the algorithmic stick is still swinging.

Here's the playbook that works in 2026 — not in theory, but in the 14-day tests our cohort has actually run.

01What the DMA actually changed

The DMA designated Booking.com a gatekeeper in May 2024. Gatekeepers must unblock merchants from selling at lower prices on other channels. In practice:

• You can offer a lower rate on your direct website, your call center, and your walk-ins.
• You can offer a lower rate on a competing OTA (Expedia, Agoda, Hotels.com, etc.).
• You can run closed-group loyalty rates, member rates, and corporate-portal rates below the Booking.com public rate without contractual penalty.
• Booking.com cannot remove you from the platform, downgrade your Preferred Partner status, or issue a formal warning specifically for undercutting their price.

02What the algorithm still does

Booking.com's ranking algorithm is a black box. What we know from controlled tests across our cohort:

• Rate-match weighting appears to still be present. Properties whose Booking.com rate is the same as or cheaper than their direct rate consistently rank higher than properties where Booking.com is more expensive.
• The effect is not binary — it's a ranking-score multiplier. A 2–5% differential shows a small ranking impact; a 10%+ differential shows a large one.
• The impact plays out over 10–21 days after the differential is introduced, not immediately. This is important: short tests don't show the effect.
• Review response rate, review score, and cancellation rate are independent ranking signals. Doing those well can offset some of the rate-match drag.

03The test-and-measure protocol

Phase 1 — choose your test surface

Don't test on your flexible rate plan. That's your hero rate for direct conversion and your primary Booking.com rate. Test on a non-refundable or advance-purchase plan where the business risk of a small ranking dip is smaller.

Reason: if Booking.com demotes you, the cost of demotion is incremental — you lose some bookings on the non-refundable plan, which is a smaller share of your total than the flexible plan.

Phase 2 — set the differential

Start at 5% direct-below-Booking.com. Lower than that, the signal is hard to measure; higher than that, you risk triggering the algorithm's visible enforcement (not a ban, but a measurable ranking drop).

Phase 3 — baseline measurement

For 14 days before the test, record: daily Booking.com impressions, conversion rate from impression, rank position on a fixed search (e.g., 'hotels in Portland, Oregon, July 15–17'), and net direct bookings for the test rate plan.

Phase 4 — run the test for 14 days

Introduce the 5% differential on the non-refundable plan. Direct channel sells that rate 5% below Booking.com's publicly-visible rate for the same plan. Keep all other variables constant (content, offers, promotions).

Record the same four metrics daily. Expect noise — individual days vary ±15%. Focus on 7-day rolling averages.

Phase 5 — decide

04Closed-group rates — the safer alternative

If the open-rate test is too aggressive, run a closed-group differential instead. Members-only, email-verified guests, or corporate-portal bookings can see a 10–15% discount below Booking.com without appearing in search results that the algorithm scrapes. Legal under DMA; contractually permitted by most OTA agreements; low ranking risk because the scraped rate remains at parity.

This is how Hilton's Price Match Guarantee works in practice — the lower rate exists but is walled off from public scrapers, so OTA algorithms don't register it.

05Metrics that matter, metrics that don't

• Matter: net direct revenue, Booking.com impressions on a fixed search, direct-channel conversion rate, total room revenue across all channels.
• Don't matter: Booking.com commission rate (it didn't change), direct-channel traffic (changes too slowly), Booking.com review score (independent).