How to Reduce Business Costs

Tool: Variance Analysis · Time: 1 to 2 weeks

Start with last year's own numbers. Variance Analysis compares what was actually spent with what was planned, and splits each difference into named causes, such as price, quantity, mix and timing. One overrun becomes several, each with its own owner.

Each piece is labeled favorable or unfavorable. Those words only say which way the number moved against plan, not whether anything good happened. A favorable price from cheaper material often causes the unfavorable usage next to it.

Look for the large pieces, and ask who could change each one. A cost driven by a market price needs a different answer from one driven by how the work is done. Overruns caused by the business doing more, such as higher sales volume, aren't overruns at all; split them out first, so they don't hide the real ones.

Tool: Benchmarking · Time: 2 to 4 weeks

Your own plan only tells you where you overspent against yourself. Benchmarking compares your costs with a reference point outside, to show which are high by any standard. It comes in four types, by how far you look.

The four types of benchmarking, by distance

1Internal

One plant or team against another in the same company

2Competitive

Against direct competitors, often through industry surveys

3Functional

One function against the same function in another industry

4Generic

A process against whoever does it best, in any industry

Compare like with like: cost per unit of output, not total spend, and businesses of similar size and type. For cost work, competitive figures from an industry association are usually the most useful start. Check how each figure is defined before comparing it; the same word can cover different costs in different companies.

Not every gap is waste, and this is the step most often skipped. A cost that is high because of a choice, such as faster delivery that customers pay for, is a strength. Mark those, and leave them out of the cuts.

Tool: Activity-Based Costing · Time: 4 to 6 weeks

Benchmarking shows which costs are high. Activity-Based Costing shows who causes them. Most accounts spread overhead by volume, so a small, fiddly order looks as cheap to serve as a large, simple one. ABC assigns overhead in two stages instead.

Activity-Based Costing's two stages

1First stage

Overhead is gathered into a cost pool for each activity, such as machine setups

2Second stage

Each pool flows to products and customers by a cost driver, such as the number of setups

A cost driver is a measure of how much demand something places on an activity. Divide each pool by its driver to get a rate, such as EUR 500 per setup, and charge each product or customer for what it actually uses.

The result usually points the same way: small and complex orders cost far more to serve than anyone thought, and large standard ones far less.

The driver data usually exists already, in the planning system, the shipping log or the engineering change records. A year of it is enough. Precision matters less than direction.

Tool: Zero-Based Budgeting · Time: 6 to 8 weeks, once a year

Now rebuild the budget. Zero-Based Budgeting justifies every activity from nothing, instead of starting from last year's figure. Each activity is written up as a decision package: what it does, what it costs, and what happens at a lower or higher level of funding.

Managers then rank every package against the others, and money is given out down the list until it runs out. What sits below the line isn't funded. Use stages 1 to 3 to rank: packages that cut scrap, air freight or loss-making work rise, whatever their size.

Note that spending can go up. The planning team's improved package costs more and saves far more, so it ranked near the top. Zero-based does not mean lower; it means every amount has a reason.