Multiply your merit increase percentage times the job value (the salary divided by the Compensation Ratio). For example, 4% times ($20,000/.93) generates a 4% increase on a job value of $21,505 or $860. That would equate to a "4.3% of salary" increase for an incumbent earning $20K.
Using the same example ($21,505=100% C/R), the same 4% Performance Rating against Job Value would give a woman earning $19K the same $860 but it would be 4.5% of her base while the equally-performing peer-position guy earning $23K would find his equal $860 to be only 3.7% of his base salary. Very fair and moves each appropriately on the maturity curve.
On the other hand, what does Development Potential have to do with Performance Rating or Compensation Ratio? One is a future guess and the others are respectively a recent historical assessment and a status quo measurement.
If you consider it important that high-performers be promoted swiftly (questionable at times), you could do a logical if formua to accelerate the progression for hi-pots, but sophisticated outfits frequently assign folks for seasoning into areas where they cannot excel but merely need exposure/experience. And some who have hit their Peter-Principle level require merit cash to retain them in lieu of promotability premiums they can't win. Most upwardly mobile get the greatest dollars from promotions rather than from annual merit increments, anyway, so paying now merely for future potential may be unnecessary. You can always make the necessary adjustments to promotional increases when they are actually awarded rather than when they are merely theoretically reflected by "potential". In today's scarce-skill world, however, there can be good reason to slap golden handcuffs on both your institutional memory and your future leaders.
E. James (Jim)