The performance improvement process

The whole method, in twelve minutes.

Two doors, one method

Performance problems arrive in two forms. The form decides your first move. It decides nothing after that.

In the first, a result is moving the wrong way. Deliveries are late, defects are up, repeat orders are down, an audit came back badly. No solution has been named. The person who owns that number wants the cause and the fix.

In the second, the solution is already chosen. It arrives as a sentence: we need a training in de-escalation, a leadership program, a culture change. A result is moving the wrong way behind that sentence too, but it has been translated into a solution before anyone established the problem. That translation is the misunderstanding this series exists to take apart, and it is the subject of Kostbaar misverstand (Arets & Heijnen, 2008), which much of the thinking here comes from.

That last row is the business case for one method rather than two. A director and a learning advisor are protecting the organization from different mistakes. They need the same evidence to act, and once they have it they are looking at the same page.

And what happens when nobody establishes the cause

The case for diagnosis is usually made as savings, which is the weakest version of it. The stronger version is about what an organization can and cannot know afterwards.

Something almost always gets bought. A program, a system, a restructure, a campaign, a new supervisor. What happens next falls into four patterns, and an organization that did not establish the cause beforehand cannot tell which of the four it is in.

The fourth is the one worth sitting with. An organization that improves a result without knowing why has not learned anything it can use again, and the next region, the next plant, the next department starts from the same place the first one did. Undiagnosed success and undiagnosed failure are the same condition wearing different faces.

Three analyses, five questions

The workflow has three phases and five questions. It is not a study, a survey or an audit. Days, not months, and each question narrows the next.

Business analysis: question 1

Which problem is worth solving? A result the organization already counts, owns and reports. Not “engagement is low” but “on-time delivery fell from 96 to 78 percent.” What that costs is a separate question with its own arithmetic, and number 13 does it. Where no such number exists, establishing one is the first output, and it usually explains why the problem has run for two years.

Performance analysis: questions 2 and 3

How far is performance from where it needs to be? Two values on one measure: current performance and desired performance. Desired is not perfection, and it is not a number somebody picked in a meeting. The strongest version of it is a level that people in this organization already reach, which is what the next question establishes. The distance between the two is the gap, and it is what any solution has to close.

Who already achieves it? In almost every organization we have analyzed, somebody gets the result under the same conditions with the same tools. What they do differently, and what they have that others lack, is evidence the organization already owns, and the first place a sound analysis looks.

Cause analysis: questions 4 and 5

Where does the cause sit? At the level of the organization, the process that crosses departments, or the person doing the work. The most expensive causes live in the handoffs: the white space on the organization chart, where work stops between two jobs and nobody owns it (Rummler & Brache, 2013).

What kind of cause is it? Six conditions make performance possible, and four of them sit in the work environment rather than in the person. The condition that is binding decides which solutions can work.

What this is, and what it is not

This is not a proprietary framework and it is not a learning method. It is a professional discipline with a name, a definition, a certifying body and more than sixty years of published work behind it.

The name is Human Performance Technology. Its society (the International Society for Performance Improvement, founded in 1962) describes it as a systematic approach to improving productivity and competence that “uses a set of methods and procedures, and a strategy for solving problems, for realizing opportunities related to the performance of people” (ISPI, 2004). The society certifies practitioners against ten published standards, and the order of those standards is the method: be about the result, establish the need, establish the cause, and only then design, implement and evaluate.

It was assembled by people working on different problems in different industries and different decades: Harless on analysis before design, Gilbert on the accomplishment rather than the behavior, Deming on results as a property of the system, Rummler on the white space between departments. They converged, and that convergence is the reason to trust the order of the questions rather than any single author. Number 02 takes the discipline, its history, its four traditions and the ten standards in full; this page is here so that you know what kind of thing you are reading about.

Note what the standards do not contain. No solution type is named anywhere in them: not training, not coaching, not a system, not a reorganization. They describe an obligation to find out and an obligation to prove, and they are silent on what the answer will turn out to be, which is the only honest position for a discipline that claims not to know in advance.

Six conditions, and the order to ask them in

The workflow puts the person at the end of the cause analysis rather than the beginning. That ordering is not a courtesy. It is where the measured evidence points, and number 03 is where that evidence is set out.

The order does the work. Asked in this sequence, each question either finds a cause that is cheaper and faster to fix than the next one, or rules it out and hands the problem on. Asked in reverse, which is how a request usually arrives, the sixth condition absorbs a problem that belonged to one of the first five, and the money is spent before anyone has looked.

Which paper answers which question

This is a map, so here is the route in the form a map should take: your situation, the question you are actually at, and the number to read next. If you have one afternoon and one problem, the shortest useful path is 05, then 06, then 07, which between them produce a measured gap with a defensible target and a list of conditions to examine.

The five questions, on one problem

A map is not much use without one demonstration. A building-products distributor: three orders in ten arriving outside the five-day promise, and everybody with the same explanation (the warehouse had an ownership problem, and specifically a supervisor with a reputation). Here is that problem taken through all five questions. Number 04 carries it in full.

The fix was a scheduled zone-to-zone transfer twice a shift and a timestamp on the handoff, so that the wait would appear in a system for the first time. The timestamp was a configuration change. The performance conversation did not take place.

What comes back

Run the three analyses and you do not get an opinion. You get a document with a fixed anatomy, and every part of it can be argued with. That is the point.

It comes out in more than one voice. The person who owns the money, the manager who owns the team and the analyst who owns the detail do not read the same document. The evidence stays identical; the length and the language change. That is what lets a diagnosis travel from the shop floor to the board without being rewritten by three people on the way.

The subject of the report changes most. Not what was delivered (how many people, how satisfied, how complete) but what was found, what it is costing, and what to do about it.

Why this was never routine

If the case is this strong and this old, the obvious question is why almost nobody works this way. The answer is not disagreement. It is price.

Running the workflow properly meant a skilled analyst, several weeks, and access to the person who actually owns the problem. It also meant the standing to question a manager’s own diagnosis: on every request, not on the two or three a year somebody could find time for. Most organizations had neither the analyst nor the weeks. So the request went straight to delivery, and everyone told themselves the analysis would have said the same thing.

Two constraints changed. The administrative weight of an analysis (structuring a conversation, computing what a request costs, classifying causes, writing the same finding four ways) is arithmetic and formatting, and it is now automatic. And the part that was genuinely scarce, the diagnostic reasoning, no longer has to sit in one head in one building. A system that carries the method puts the questions an experienced investigator would ask in front of whoever is holding the problem, in any region, on any day.

What it cannot do

A method that will not name its own boundaries should not be trusted with a budget. Four limits, stated plainly.

• It cannot invent data that does not exist. Where nobody measures the result the organization is unhappy about, the first output is a short list of the numbers it does not keep.
• It cannot watch the work. Standing beside someone who already gets the result, and noticing the step that appears in no procedure, is human work. The method says whom to watch and what to look for. It cannot look.
• It cannot make anyone accountable. A diagnosis produces results only when the person who owns the problem also owns the solution. Where the causes sit in another director’s process, naming them changes nothing by itself.
• It cannot carry the decision. Every cause is labeled with how far the evidence reaches. Acting on evidence that is good enough, and defending that choice, belongs to a person.

Your door, your problem, one diagnosis

Ten minutes, on the problem in front of you now. Answer these five and you have the input a diagnosis needs. Leave one blank and you have found the first thing to establish.

Two problems, one method, sixteen papers

Two models carry the method. Gilbert’s six conditions ask what kind of cause; Rummler and Brache’s three levels ask where it sits. Both are sound and this series uses both, and it leads with Gilbert, because his unit is the accomplishment, which is what an organization already counts, and because his six conditions are the categories an organization can actually change.

Where the method comes from

Two further pieces close the set and carry no number, because each points at all the numbered papers. Before the guide sits Arets L&D Business Models, the destination piece for the people who receive the requests: four models a learning function can run, and the route from HR-driven to business-driven. Last sits Solutions that cannot work, the field guide, ordered by the sentence that lands on your desk: seventeen forms a performance problem arrives in, each with an address in one of the numbered papers. Three orders in all: the destination you choose, the question you are on, and the form that landed on your desk.

Sixteen papers, six parts

Number 01 is the map and number 02 is the foundation. Every number from 05 is depth on exactly one step of the method, and no number repeats what 01 explains. The filled block is the paper you are holding; the two unnumbered pieces close the set.

Where every claim in this paper comes from

This number takes one claim from Kostbaar misverstand (Arets & Heijnen, 2008), never translated, so the book is listed below. Much of the thinking in this series comes from it. The other two books named earlier are listed in full in number 02.