Tim Taylor and Agronomist gave you some good advice.  Let me add to it:

• Lime is the simplest product to VRA, and as Tim says, may give you the most short term results.
• Your greatest response will be based on addressing the GREATEST VARIABILITY that is CORRECTIBLE.  This may be P, K, or Lime, or other things....
  
• It could even be sandier dirt that may have adequate to good nutrient levels, but suffer lower production because of HIGHER NEMATODE numbers, COMPACTION, or perhaps more NITROGEN LEACHING.
  
• If you go with 2.5 acre grids, be sure that the average size is NO MORE than 2.5 acres. This is the easiest way to do precision ag, the least time consuming, and probably the most economical.
  
• Consider pulling samples in zones, dividing fields into sub-fields.  We base this on lots of things, including, but not limited to, aerial imagery, soil type maps, yield monitors, experience and knowledge of the field.  We have found pH and fertility levels in low production areas of the field sometimes to be VERY GOOD, and attribute this from a poor yields removing fewer nutrients.  Here in our coastal plain low CEC sandy soils, we have to watch calcium and magnesium levels very closely, and the  tendency is to watch pH, K, and P. 
  
• Nematode samples are expensive.  We typically will send in two samples per field, sample one being a consolidation of dirt from two poor yielding zones, and sample two being a consolidation of dirt from high yielding zones.  This gives us a benchmark.  Be careful of nematode numbers, though, as it is getting late in the season for good nematode sampling, and they also will have false low numbers if the dirt is dry.  As for nematodes, I put more faith in growing season analysis of the crop, its performance in the fruit maturity period, and root galling.
• Hire a qualifed independent agronomist whose salary is not subsidized through the sale of fertilzer.  Go to the NAICC web site and find one that works in your region, http://www.naicc.org.