Websites and platforms change all the time, and even a small update on their end can quietly break the data flowing through our APIs. A page layout shifts, a parameter stops being read correctly, or an internal process hiccups, and suddenly the results you're pulling aren't what they should be. Our job is to catch these breaks quickly, figure out exactly what changed, and fix them before they slow you down or cost you time debugging on your side. That way, everything keeps working the way you expect, without you having to lift a finger.
This week we shipped four fixes across four different APIs, covering eBay search, Google's AI-powered search results, Amazon product data, and Google Hotels. None of these require any resolution on your end and all of them are already live. Here's everything that happened, explained in plain language, along with why each change matters for the work you're doing.
1. eBay Search API Response Consistency Improved
What was happening:
Some search requests made through the eBay Search API were coming back completely empty because of an internal issue, even though the searches themselves were perfectly valid. This was intermittent, meaning the same query might work fine one moment and return nothing the next, which made it harder to catch and harder to plan around if you were pulling data at scale.
What we fixed:
We tracked the issue back to its root cause inside our systems and resolved it directly, rather than patching around the symptom. The API now consistently returns accurate search results for every valid request, with no random gaps.
Why it matters
If you rely on eBay search data for price tracking, listing monitoring, or market research, you'll now see far fewer gaps in your results. Response reliability and consistency are both improved, so you can trust the data coming back without needing to add retry logic or extra checks on your side.
2. AI Overview & AI Mode Reference Extraction Improved
What was happening
A parsing issue in our processing pipeline caused the AI Overview API and AI Mode API to return fewer reference links than they should have. Google's AI-generated answers typically cite several source pages behind each summary, and our pipeline was silently dropping some of those references before they reached you, even though the underlying page had more links available.
What we fixed
We went through the parsing logic behind both APIs and corrected how reference links are identified and extracted during processing.
Why it matters
Both APIs now consistently extract and return a more complete set of references. This means better data accuracy and completeness for anyone using this data for SEO research, content analysis, citation tracking, or building tools on top of AI search results.
3. Amazon Product Scraper Purchase Options Fixed
What was happening
Amazon changed the underlying structure of its product pages, and that change caused the purchase options field in our Amazon Product Scraper API to return incorrect and incomplete data. If you were pulling details like different sellers, quantity discounts, or subscription pricing, some of that information was either missing or not matching what was actually shown on the page.
What we fixed
Our team rebuilt the parser for this field to support Amazon's new page layout, testing it across a wide range of product types to make sure it holds up consistently.
Why it matters
The API now returns accurate and complete purchase option information consistently, so if your workflow depends on pricing, seller, or buying-option data from Amazon, you can trust it's up to date and matches what's live on the page.
4. Google Hotels API Vacation Rentals Parameter Fixed
What was happening
Requests to the Google Hotels API that included the vacation rentals parameter were returning errors instead of results, due to an internal parsing issue on our side. This meant anyone specifically trying to pull vacation rental listings alongside standard hotel data through this parameter was getting blocked entirely.
What we fixed
We identified and resolved the underlying parsing problem that was causing requests with this parameter to fail, and tested it across a range of locations and query types.
Why it matters
The API now correctly processes requests that include the vacation_rentals parameter, returning reliable and consistent results every time. If vacation rental data is part of your hotel search workflow, you can now use this parameter without running into errors.
All of these fixes are already live, no action is needed from you. If you run into any issues or have feedback, feel free to reach out to our support team. We'll be back next week with more updates. Happy scraping!
