Me personally i would do waht @Vlad Iakimov says. Start small with come phrases and then check your search terms every day and keep adding ones that are not relevant at all. If you have several campaigns set up i would think about doing a shared budget for them all if plausible. Since its just you doing it going into google ads every day or every other day and spending 5 minutes excluding negative keywords is fast and easy. I find with small business and small budgets it might be better that was you can monitor and distribute money to what is driving your conversions. Then you can really focusing in on the winning category first and get that going well then focus on the next best campaign. Each time taking the one campaign that is more ahead and giving it its own budget and letting the smaller campaigns prioritize themselves. I wouldnt do any google ads at all unless you website and landing pages are in peak form though. Google will want to crawl and get info from them to power the searches and if you start out on a week footing then its an uphill battle. Making sure all landing pages have similar text and images to the ad and search terms that are being used. If you 4 different landing pages because services are slightly different than do it. This part is super important. Feel free to post landing pages and we can take a look at them.

I generally would not switch keyword match types. Were you getting over 30 conversion per month? I would guess that google might broadly look at keywords during the initial days of the campaigns to help navigate to get to your phrase match goals but i am not sure. Do you have enough negative keywords to filter down any excessive searches that the broad match might bring? If you are getting 30 plus conversions now you can do an experiment and use broad match keywords and see what happens.

@Amanda Coleman ugh i would want to be you. First turn off those auto apply things. They generally suck.

@Todd Creager i usually wait 4-6 weeks with new campaigns. All of them

You could set up campaigns for Best sellers, average sellers and zombie products and have a campaign for each with the correct strategies. This way you could allocate the spend accordingly. Do you have a solid negative keyword list for the shopping campaign. On one of my websites getting quality brand searches is a task as we just sell small fraction of what the brand does so i have to constantly add negative keywords otherwise we get hit with everything.

I do mostly ecommerce. In responsive ads i have some short and some longer length descriptions. I cant seem to see any significant differences in conversions when looking at them both. I think the real test could be A/B them and see if there is any difference. While i would like to try that i dont think i have enough time in the day to do it unless i thought it work make a drastic increase and i have other things that i have to improve on now.

@Vlad Iakimov I run all tests for 30 days. after 2-3 weeks if there is some definitive results i might shut it off. If its a all brand new you need to give everything at least 30 days before you start running any test against it. I usually only run test on my older campaigns for when i am going to do bid strategy changes or something major. I will usually update headlines and descriptions once i get some data on it. Once i am comfortable on the amount of clicks i am getting in total I can compare the low performing assets to the high performing ones and try to make appropriate changes.