Buying by pipe diameter alone. Comparing accuracy percentages with different bases. Treating a maximum/overload flow as continuous capacity. Assuming a hot-water filename proves the temperature rating. Using a clean-water mechanical meter for unspecified sewage. Choosing a mag meter without conductivity data. Buying an ordinary full-pipe instrument for a partly full drain. Assuming RS485 means Modbus. Forgetting the pulse pickup or gateway. Accepting battery life without an operating profile. Calling daily uploads “real time.” Using an old price list as a current offer. Combining the best specifications from several incompatible variants into one imaginary product.
The remedy is the same: identify one exact configuration and validate it against one complete application specification.