Incredibly robust software for an enterprise organization to plug into their application. If you have a full development resource team at your disposal, this is great software and I highly recommend it. Largely, however, you won't be able to use this prior to the enterprise level. It's just too complicated and cumbersome of a product.
The Google Search Appliance is well suited to most site search needs and it is possible to customize the front end seen by website visitors to produce a satisfactory basic interface with basic branding applied to it. However, in some situations it may be necessary to instead make API calls from a web page to retrieve search results and the format in which those results are returned might be a little more difficult to work with.
Azure Search provides a fully-managed service for loading, indexing, and querying content.
Azure Search has an easy C# SDK that allows you to implement loading and retrieving data from the service very easy. Any developer with some Microsoft experience should feel immediate familiarity.
Azure Search has a robust set of abilities around slicing and presenting the data during a search, such as narrowing by geospatial data and providing an auto-complete capabilities via "Suggesters".
Azure Search has one-of-a-kind "Cognitive Search" capabilities that enable running AI algorithms over data to enrich it before it is stored into the service. For example, one could automatically do a sentiment analysis when ingesting the data and store that as one of the searchable fields on the content.
It's an enterprise level product so you need to have the budget for it.
Challenging-to-impossible for a non-technical administrator to implement.
It further locks you into Microsoft's ecosystem and doesn't play well with non-Microsoft software. Depending on your point of view, this can be a pro or a con.
As I've mentioned, the biggest competitor to Azure Search is actually Azure SQL Database. It doesn't have as many features, but it's more economical and most .Net applications will have one already. As long as you can arrive at a schema and ranking strategy, it's a "good enough" solution. There are a variety of search technologies (Lucene, Solr, Elasticsearch) that implement a search service. Some of them are even open source, though I would only say "free" if you do not value your time. They most likely need to be hosted via Container (or VM if you're old school), so you're incurring DevOps costs to not only set them up but monitor and maintain them yourself.
If you're already on AWS, there is almost no reason to use Azure Search. Unless you're already multi-cloud, desperately need the cognitive abilities, and don't mind a potential performance hit from looking across datacenters (hey, it could happen), you should probably just use Amazon CloudSearch.
There has also been a cost associated with our need to switch to another site search solution in the near future, due to Google ending support for the sppliance.
That said, we have so much support content that we owed it to our customers to offer a comprehensive search tool to help them find the most relevant content - and of course Google are the masters of this. However, it's worth noting that the alogorithms used on the GSA are far older than those used on google.com.
