Lets forget about luck, at least for now.
So forget about the people that visit via search engines, referrals and other channels over which you have limited control.
If you need to set targets, then do it based upon the activities you control.
If you invite 5 people a day to join the community, it's not unreasonable (assuming your invites are personalized and highly targeted) to expect 3 – 4 of the recipients to join.
Of these, 2 might still be active after a few months. Over time, you should get better at this and increase your ratios, but lets not focus on that for now.
Based upon these assumptions, you can make a prediction for the early growth of your community based upon your own actions.
So if you spend 5 hours inviting 25 people a week, you can reasonably expect your community to grow by 40 members a month. That sounds small, but most communities develop this way.
Now if you spend 10 hours a week and invite 50 people, that figure doubles (this is where organizations that have a big budget for community development should invest in staff to do this work and not the platform).
Better still, you can measure this. After the first few weeks, you can identify how many people you invited actually joined. Then you can set benchmarks and future predictions.
Later you will get referral growth. Members will mention the community to their friends, they will share discussions via social networks, and a few might drift in from search engines. When this happens you can measure this and benchmark this and set new predictions.
