01The organization
Picture a booking platform selling hotel stays and flight-plus-hotel packages in a dozen markets. Availability and rates come from partner APIs, prices are shown in the visitor's currency, and loyalty members see member rates after they log in.
A pricing bug here costs money twice: once in refunds and again in trust. The team ships several times a week and checks pricing by hand in a few markets.
Travel platforms depend on data they don't control. Partner rates, availability and formats change without notice, and every market adds a language and a currency. Testing has to cover what the platform does with partner data, not just the platform's own screens.
02Business objectives
- Test search, pricing and booking in every market each release
- Make sure the price at checkout matches the price in search
- Test cancellation and refund rules
- Spot slow search before customers do
03Testing challenges
Prices from partners
Rates come from partner APIs that change format and behavior without notice.
Currency and rounding
Totals in each currency have to match between search, the rate page and checkout.
Rules by rate
Refundable, non-refundable and member rates follow different cancellation rules.
Speed matters
Slow search loses bookings, but performance is checked only after complaints.
04Before and after
How the testing work would change once the journeys run in Robonito.
- Markets
- Today:Pricing spot-checked in a few markets by hand.
- With Robonito:The same journeys run in every market and currency.
- Partner data
- Today:Partner API changes found through refund tickets.
- With Robonito:API checks and price checks run together on every deploy.
- Cancellations
- Today:Refund rules tested when a complaint comes in.
- With Robonito:Refundable and member-rate rules run as named tests.
- Speed
- Today:Slow search noticed after customers complain.
- With Robonito:Response times tracked per release, with regressions flagged.
05How Robonito would test it
- 1
Journeys per market
QA writes each journey once in plain language and runs it with each market's language and currency.
- 2
API plus UI
Tests check partner API responses and the prices the site shows from them in the same run.
- 3
Response times per release
Performance testing tracks search and checkout response times per release and flags regressions.
- 4
On every deploy
GitHub Actions runs the suite on each deploy, in parallel across markets.
06Test scenarios, in plain language
Tests you'd write in Robonito on day one. Each one is typed as written; the agent works through the app and saves the steps.
- SearchSearch Lisbon hotels for three nights for two adults and a child and confirm every result shows a room that fits three guests.
- CurrencySwitch the currency to EUR and confirm the checkout total matches the price shown in search.
- CancellationBook a refundable rate, cancel inside the free cancellation window, and confirm a full refund is shown.
- LoyaltyLog in as a loyalty member and confirm member rates replace public rates on the results page.
- PackagesAdd a flight to a hotel stay and confirm the package price is lower than the two booked separately.
- PerformanceTrack the response time of a city search on every release and flag a slowdown against the last release.
07What the pilot measures
No made-up numbers: a pilot measures these against your current baseline, on your own app.
Markets covered
How many markets and currencies are tested on every release.
Pricing defects
Price mismatches caught before release, against today's refund tickets.
Response-time regressions
Slowdowns flagged by release, before customers report them.
Release frequency
Whether testing keeps up with several deploys a week.
08Rolling it out: the first month
A realistic plan for a team starting from zero, with the baseline comparison built in.
- Week 1
Cover search to booking
QA writes search, rate selection and booking journeys for the largest market.
- Week 2
Add markets and currencies
The journeys run with each market's language and currency, checking that totals match from search to checkout.
- Week 3
Add rules and speed
Cancellation, loyalty and package rules are added, and response times are tracked for search and checkout.
- Week 4
Run on every deploy
GitHub Actions runs the suite on each deploy, and the team compares pricing defects with the previous month.
09Frequently asked questions
Can Robonito test pricing in several currencies?
Yes. Write a journey once and run it with each market's currency and language, checking that the total in search, on the rate page and at checkout match.
Can Robonito test partner APIs?
Yes. Robonito tests REST APIs and can check a partner's response and the price the site shows from it in the same run. API testing is on the Pro and Custom plans.
Does Robonito track performance?
Robonito's performance testing tracks response times per release and flags regressions, so a slower search shows up against the previous release before customers notice.
Can tests run in parallel across markets?
Yes. Runs execute in parallel in isolated cloud workers. The Pro plan includes 5 parallel runs, and the Custom plan sets parallel runs to fit your needs.