Grew from Lawrence D. Miles's value analysis at General Electric (Techniques of Value Analysis and Engineering, 1961), rebuilt as function analysis by the TRIZ school of the 1980s, and set out by Darrell Mann (Hands-On Systematic Innovation, 2002).
Maps a product, service, or process as components joined by the functions they perform, tagging each function useful, harmful, too weak, or excessive to show where value leaks.
Eight functions of an office coffee machine, tagged by whether they are wanted and how strong they are
Each tag is one thing a part does to another, written as a verb and an object. Select one, then the box that fits: wanted or not, and too weak, right, or too much. Harmful and excessive functions are the first redesign targets. Show the sort to compare yours. Illustrative.
- Not placedPlacedToo weakUsefulExcessiveHarmful
- Not placedPlacedToo weakUsefulExcessiveHarmful
- Not placedPlacedToo weakUsefulExcessiveHarmful
- Not placedPlacedToo weakUsefulExcessiveHarmful
- Not placedPlacedToo weakUsefulExcessiveHarmful
- Not placedPlacedToo weakUsefulExcessiveHarmful
- Not placedPlacedToo weakUsefulExcessiveHarmful
- Not placedPlacedToo weakUsefulExcessiveHarmful
Is it wanted?
How strong the effect is
Your sort matched 0 of 8.Your sort matched 1 of 8.Your sort matched 2 of 8.Your sort matched 3 of 8.Your sort matched 4 of 8.Your sort matched 5 of 8.Your sort matched 6 of 8.Your sort matched 7 of 8.Your sort matched 8 of 8.
- Too weak
- A function the machine needs, delivered short. Find what is holding it back before adding power.
- Useful
- Wanted and the right size. Leave it alone, and protect it during the redesign.
- Excessive
- Wanted, but more than anyone needs. Money and energy spent past the point of use: shrink it.
- Harmful
- Nobody wants it, at any strength. Remove it, block it, or turn it into something useful. A top redesign target.
The sort
- Hard water coats heater: Harmful. No one designed it, and it insulates the heater, which is why the coffee comes out lukewarm. Tagged harmful, it leads straight to the weak function it causes.
- Heater heats the water: Too weak. Needed, and short by ten degrees. A bigger heater would only feed the scale; clearing the scale restores it.
- Pump pushes water: Useful. Wanted and the right size. Keep it as it is.
- Grinder grinds beans: Useful. Wanted and right. No complaints trace to it.
- Filter softens water: Too weak. Wanted, and spent most of the time, which is how the scale builds. A larger cartridge with a reminder fixes it.
- Hot plate heats the jug: Excessive. Wanted warmth, far past what anyone needs: it burns the coffee and runs all day. A thermal jug does the job without power.
- Screen offers 14 drinks: Excessive. More than people use, and the large screen is the machine's most expensive part to replace.
- Spills rust the frame: Harmful. Unwanted at any strength, and the second most common reason for a service call.
An example
A company that leases coffee machines to offices was losing money on its largest model: service calls cost more than the lease brought in. The usual cost review had looked at the parts list and found nothing to cut, since every part seemed necessary.
The product team ran a function analysis instead. They listed each component and what it does to another, as a verb and an object, then tagged each function. Two functions nobody had designed turned out to drive most of the calls: hard water coating the heater with scale, and an overflowing drip tray rusting the frame. The scale explained the most common complaint, lukewarm coffee, because it weakened the heater, and the filter meant to prevent it was spent most of the time. Two functions were overdone: a hot plate that burned coffee all day, and a screen offering fourteen drinks to offices that used three. The redesign added a larger filter with a change reminder, a deeper drip tray, a thermal jug in place of the hot plate, and a smaller screen.
Why it works
A parts list hides the problem, because every part looks necessary. A list of functions shows what each part does to another, and once each function is written as a verb and an object it can be judged: is this wanted, and is it the right size? Most cost and most complaints trace to a few functions that are harmful, too weak, or overdone, and they stay invisible until someone writes them down.
The four tags also order the redesign. A harmful function is a target in itself, and it often explains a weak one nearby, as scale on a heater explains lukewarm coffee. An excessive function is money spent past the point of use. A function that is wanted and the right size is left alone, which keeps a cost exercise from cutting what works.
The thinking behind it
Lawrence D. Miles Lawrence D. Miles, an engineer at General Electric, devised value analysis in 1947, the origin of value engineering, and set it out in Techniques of Value Analysis and Engineering (McGraw-Hill, 1961). Its core move is to describe each part of a product by the function it performs, in a verb and a noun, and then to look for the lowest-cost way to perform that function reliably. Writing the job a part does in place of the part's name turns a parts list into a list of jobs, each of which can be priced and questioned.
Darrell Mann Darrell Mann, an engineer and TRIZ practitioner, set out function analysis in Hands-On Systematic Innovation (CREAX Press, 2002), the handbook that updates Genrich Altshuller's theory of inventive problem solving with function analysis, trimming, the ideal final result, and the separation principles. The TRIZ form of the method, rebuilt from value analysis in the 1980s, tags each function useful, harmful, too weak, or excessive, and draws them all on one diagram so the problems show where they sit.
How to run it
- List every component. Include the parts, what they handle, and outside things that act on them, such as hard water or the user.
- Write each function as verb and object. Component, verb, target: the pump pushes water, scale coats the heater.
- Tag each function. Useful and the right size, too weak, excessive, or harmful.
- Draw them on one diagram. Components as boxes and functions as arrows, each tag in its own line style.
- Find the chains. Look for a harmful function behind a weak one; fixing the first often fixes the second.
- Target harmful and excessive functions first. Remove, block, or shrink them before adding anything new.
- Redesign and redraw. Draw the diagram again after the change to confirm no new harmful function appeared.
Where it goes wrong
The first failure is listing parts and calling it function analysis. A component list with costs is a bill of materials, and without verbs and targets nothing on it can be judged. Spot it when the diagram has boxes but no labeled arrows.
The second is leaving out the harmful functions nobody designed, such as scale, heat, vibration, or a confused user, because they appear on no drawing. Ask what acts on each component that no one intended. The third is cutting by tag alone. A function marked excessive for one customer may be exactly right for another, so check each tag with the people who use the machine before removing anything.
When to use something else
- TRIZ Trimming. When the diagram is done and the aim is to remove whole components by handing their functions to parts that stay.
- TRIZ Inventive Principles. When a harmful function cannot be removed without losing a useful one, and the conflict needs an inventive way out.
- Value Stream Mapping. When the system is a process of steps and waits, and the waste lies in how the work flows.
Where it sits in SOLVE eX
Go to the source
- Techniques of Value Analysis and Engineering, McGraw-Hill, 1961
- Hands-On Systematic Innovation, CREAX Press, 2002