Hey @Kasim21, welcome to the Community!
It appears you’ve also reached out to our Support Team and they’re helping you with this. If you have any further questions you can reply directly on that email thread.
Would you mind updating this thread with the answer once the issue has been resolved? We’d really appreciated it! That way anyone running into similar issues with the Search Mention trigger can benefit from your experience here. :)
Hey Kasim,
Thanks again for writing into our Support team! Wanted to post of summary of our findings on this inquiry for broader awareness / future discussion.
Your search query of "paper OR article OR study filter:follows filter:links" looks to be set a valid query for Twitter. Providing the URL encoded query is not required in our trigger step, as the Zap will properly encode that when the Zap runs ️
I belive the answer in short is that you may see that your search feature within a browser or the Twitter app yields some different results than our Search Mention trigger step will yield.
Our integration uses v1.1 of the Twitter API. While I would expect the Search Mention trigger to behave very similarly to the ‘latest’ Tweets section, I cannot say for sure that Twitter’s search logic when you search in your browser is not built differently than Twitter's API that is used by developers (and used in our Twitter integration).
One note on how we take the non-URL encoded text during setup and URL-encode it to get search results: When Zapier encodes your search query, it does add a few more search parameters to optimize your search results for this Zap. I reviewed the following parameters that get automatically added to the search query when the step runs:
These four parameters, as described by the Twitters docs should do the following:
- result_type=recent - lists search results in 'latest' order, instead of 'Top' tweets or a mix of Top/latest.
- count=100 - lists up to 100 Tweets per page on results (e.g., i think you get the 100 latest tweets in your results)
- include_entities=false - for the Tweets in your results, this setting will excludes some Tweet fields related to metadata/metrics for the results (but I don't think it excludes any Tweets when setting this to false)
- tweet_mode=extended - this should ensure that Tweets over the standard 140 characters still get returned in full, according to this forum
Potential Workarounds
If you were wanting to get really specific with the type of search that is being made, to see if you can get the search results matching what you are seeing in your browser results, you can try using a Webhooks by Zapier > Retrieve Poll step, which will in theory act like the Twitter > Search Mention trigger step, only it will be fully customizable in how you set up the search request to Twitter. This does come with some technical setup, but if interested, you can head here for instructions: https://zapier.com/help/doc/how-get-started-webhooks-zapier
It's possible that since our Twitter integration is using v1.1 of Twitter's API and not V2.0, that the results that Twitter's API is giving Zapier when your Zap runs is a little bit different.
In short, none of these fields seem to be the issue on their own, but the differences between Twitter’s APIs and their built in search function could be the reason for the difference in results.
Sorry I couldn’t resolve this issue in full Kasim, but I hope that clarifies why you are getting the search results seen in your Zap!
If any others have added findings on how their search results have differed and want to add to this discussion, I’d be happy to here, and contribute where I can :)