First, avoid the term "soundproof" as you will never obtain it. You can also probably ignore suggestions for absorptive materials and references to home studios, the first is generally an ineffective approach to sound isolation and latter is often not applicable to commercial projects.
I find there are generally two ways to approach such projects. The first approach is a more analytical approach where you define a desired result, establish a baseline of where you are relative to that and then try to develop solutions that
address the differences between the two. The second approach is a more "try it and see" approach where you look at what is
practical and might offer the most 'bang for the buck' and apply that, then see if the result is acceptable and if not move on to the next step. Note that with either approach you may find yourself at a
point where the investment or physical changes required are greater than the Owner wishes to accept, the difference is often in whether that is determined after some investment in design and planning or in actual physical changes.
The first step in an analytical approach is typically assessing what sound isolation is required, which entails first determining what represents an acceptable result. If the local codes have a numeric noise ordinance that may be easy, it may say you have to be XdBA at the property
line or something like that, some specific number that can be used to define an acceptable result and thus be used as a basis for determining what amount of additional sound isolation is required. If you have no noise ordinance or have a 'nuisance and annoyance' type ordinance you may have a much greater challenge right from the start in establishing what will be an accepted result.
If you don't have a numeric noise code then trying to get an agreement to what is an acceptable result may mean working with the parties affected to develop some more objective criteria. That may mean meeting with them and adjusting the sound
system level and response until they deem it acceptable, documenting the resulting levels and response at various locations and getting them to 'sign off' on that being acceptable. If you can't establish an agreement upfront defining an acceptable result then in my experience you not only have rather limited basis for determining what to do or its potential value, but you may also find yourself fighting a battle that really has nothing to do with the noise levels. I tend to avoid those situations.
If you can establish a definitive goal then you probably next want to establish where you are relative to that and what reduction in the levels is required at what frequencies. At that
point you can start looking at what may be required to achieve that improvement and the associated financial and
practical factors. But without being able to first establish the goals an analytical approach may not be
practical.
The risk with the 'bang for the buck' approach is that you can end up investing more than required or, and perhaps more likely, you can end up investing a lot without ever achieving an acceptable result. You may also run the risk of doing something hoping it works and then finding yourself having to change or undo what you just did in a later step.
For both approaches you will likely want to consider how the
venue use and operations may factor in. For example, a common weak
point in performance venues sound isolation is entries, not just in the doors, windows,
etc. themselves but in what happens when they are operated. It doesn't matter how good of a door or window you use if it is open. You might also have to consider the impact of the mechanical systems in both the noise the create and the paths they provide to the exterior.
In general the sound isolation may come down to big things and little things. General wall and ceiling (don't forget to consider the possibility of sound going up and out) constructions and any penetrations of them. Space planning and wall sealing details. Masonry and framing with backer rod and resilient sealant. In many cases, if you can't or don't sweat the details then you can significantly limit the benefit of the big dollar items.