Crawlera is a proxy rotater which helps in evading bans from the target website by adjusting delays, handling cookies and managing IPs. When Crawlera is added it can lead to slow down of jobs as it will adjust the delays to ensure crawl is not too fast, not overburden target server (thereby increasing the risk of recognition and bans).
But with Crawlera we can control Concurrency that is Number of requests that can be made in parallel. This concurrency is based on the subscribed Crawlera Plan. To increase the Crawl speed you can increase the concurrency. Concurrency is cumulative of requests made to all domains from all Crawlera accounts in the Organization.
For Scrapy Cloud, Autothrottle is enabled by default which limits maximum number of concurrent requests sent to the same host domain. To increase the concurrency you would need to disable Autothrottle and set Concurrency settings in settings.py through UI as given in this article on Customizing Scrapy settings in Scrapy Cloud.
Please note each Crawlera plan has limits on Concurrency and exceeding it would lead to errors with code 429.