There are a couple of different approaches, which are best combined.
First of all, makes sure the proper "Protector" with gas discharge tubes is installed correctly at the building entrance. This is primarily for safety, not for equipoment protection, but it is a necessary first step. You should also get a ground tester and verify the ground is good. Also, the telecom ground MUST be bonded to the power ground. This is required by NEC for good reason - if sufficient current is conducted to drive the ground to a higher voltage (and this does happen with the massive currents involved in a strike), you want all grounds to float up by the same amount. You want to avoid differences in ground potential.
In places like Florida a second stage of protection is advisable (this is also true in other places, depeding somewhat on the cost and vulnerability of the equipment to be protected). The goal hear is to prevent major voltage differences between the phone lines and the protected equipment. Again, if both rise together temporarily that is good, versus the phone line rising and the equipment laggign behind thereby permitting a difference.
In this second case the proetection is typicall MOV-based and should be installed as close as possible (less than 12 inches) to the equipment to be protected. For simple gear with modular plugs this is easy, insert it in series with a short 6" cable to the equipment. Then AND THIS IS CRITICAL to success, connect the surge protector ground wire to the grounded chassis of the equipment (again keep it short). This will serve to keep the two at the same potential.
A radio station Cheif Engineer freind has had very good luck in FL dramatically reducing problems using this approach. He typically uses the Polyphasor products, but he has proven again and again that this technique of referencing the surge protector to the chassis is the key to success.
Another technique that is very easy is the lightning choke. Simply wind 10 turns of wire around in a circles and wore tie. Even placing 3 or 4 knots in the cable has been proven to help. This creates and inductor. Inductors don't like to pass sudden changes in current, and since lightning is a very rapid change (e.g. a spike) this simple technique works well, is low cost, and is ideal to combine with the above.
In FL nothing will completely eliminate problems, but expect a dramatic reductions using these techniques.