Using experimental data from the northern Namibia pearl millet based subsistence sector the gross margins (GM) and savings on food expenditure (SFE) were computed and compared. The value to cost ratios from the SFE approach were higher than those from the GM approach implying that investment into the selected technologies would be more attractive under subsistence production than under commercial production. This underlines the importance of change agents not to change the farmers' target for farming. Furthermore, it helps explain why subsistence farmers continue to farm when all GM calculations show that they are always loosing money.