Please help!

I am struggling with this problem.

I work in a call centre and I'm looking how I can optimise our quota management techniques.

I would like to minimize the number of dials made, required to achieve our targets, by effective quota management.

My constraints are as follows:

(gender)
Male >= 441
Female >=459

(age)
18-29 >=147
30-44 >=297
45-59 >=243
60-74 >=171

(social economic class)
A/B >=207
C1/C2 >=396
DE >=231

total >=1000

I know how many dials are made and know how the interlocked (male 18-29 AB) quotas fall out.

I made my objective function to be the number of dials made divided by the number of interviews completed per quota (quotas listed below)

Men 18-29 ABMen 30-44 ABMen 45-59 ABMen 60-74 ABMen 18-29 C1C2Men 30-44 C1C2Men 45-59 C1C2Men 60-74 C1C2Men 18-29 DEMen 30-44 DEMen 45-59 DEMen 60-74 DEFem 18-29 ABFem 30-44 ABFem 45-59 ABFem 60-74 ABFem 18-29 C1C2Fem 30-44 C1C2Fem 45-59 C1C2Fem 60-74 C1C2Fem 18-29 DEFem 30-44 DEFem 45-59 DEFem 60-74 DE

I realise that the coefficients to the variables in the objective function will fluctuate. There are 24 variables, I have 368 completes.

When I use Solver in excel to minimise the objective function, it comes up with a solution and the expected minimising of dials is expected to be 10 times greater than what we currently do.

I can't see how the small variable chanage would impact so greatly on the optimal solution.

Can anyone help as to how I can go about tweaking it to make it more efficient?