DynamoDB is fast, secure and easy to use.

DynamoDB has the high scalability that is expected from the start, because it's a DB that has been created with seamless scalability (so that you can scale both up and down without having any kind of performance impact or downtime).
This important feature allows for lower costs for workloads with predictable traffic patterns.
As any other AWS service, the auto-scaling is controlled through policies and is monitored with the aid of CloudWatch.

In our experience, the perceived performance for DynamoDB is always fast and consistent under different scenarios. IT has always been consistent on performance and responsiveness without generating any kind of complaint from our users.
Of course, we are always putting this as our edge DB use cases that are near our customers with the support of other databases for permanent persistence under an structured relational more complex model.
We have used it on limited regions where our main clients have their business, so we have not experienced any region with different behavior.

Best suited for fast interactions, with no structure on the data. The advantage of scaling it up like many other services in AWS helps on using this as a basic stone for your architecture. Since pricing is based on consumption and throughput you can play with these variables to better control your costs.

Not recommended for deploying it without carefully planning for its usage and costs, as it may become easily unpredictable. Because the indexing of the data comes with its own cost, any non-planned implementation may lead to high costs and indexes are not something that can be easily removed from an architecture once in place. Also, if your needs require a more complex implementation of a query language or multiple ways of querying for your data, then this may not be the database for you.