The first ten
Under about ten, the count is doing you harm on two fronts. A person comparing three companies sees the short list and wonders, and the testing suggests the map treats a very thin count the same way.
Getting to ten is usually a week of asking, not a campaign. Most contractors have finished twenty jobs in the last two months and asked nobody. The customers are already there, and most of them would say something decent if anybody put the question in front of them.
What matters after ten
Once you're past the first handful, the total stops being the useful number and the flow starts. Two to four a month, forever, is the shape that shows up in the testing as the one that holds.
The same testing found rankings can slip when reviews stop coming for even a few weeks. Read that as a reason to make asking part of closing out a job rather than a project you do twice a year.
This is practitioner testing, not something Google publishes, so treat it as the best available evidence rather than as a rule. It matches what a customer does either way: they look at the dates.
Reviews with words in them
A star with no text tells nobody anything. Reviews carrying a sentence about the job did more in the testing, and they do far more for the person reading, because they name the work and the town in the customer's own words.
The way to get them is to ask a question rather than to ask for a favor. What job did we do, and how did it go, gets a paragraph. Please leave us five stars gets a star, if it gets anything.
What a high count can't do
It can't put you in a town where your website has nothing to say. Reviews sit on your listing and help Google decide between businesses it already considers a match. Being considered at all comes from the pages.
And it can't be bought safely. Google's rules prohibit paid and traded reviews, the reviews get removed when they're found, and the account that posted them takes the rest of your profile down with it often enough to make it a bad bet.