First step is to
The first step is to confirm that your measurement system is accurate. In this case, your static lines, airspeed indicator, and so forth. Prior to that, do a 3-way GPS run, full throttle, at 8500 ft or so. Record outside air temp.Fly one gps direction with autopilot on, record gps and indicated, change gps direction 120deg and once stabilized record gps speed and indicated, and change gps direction another 120deg and once stabilized record gps speed and indicated. Land and average the gps speed. Average the indicated speed. Calculate TAS from the indicated and then compare with gps speed. Note the error if any. Repeat except this time do the procedure at say 80kts or so. Is your indicated error greater at slow speed than cruise speed? Or vice versa? Is your Calculated TAS average the same as GPS average or is it different? What is the difference? Report back grasshopper and we can better advise you on what steps you need to do.