Ref:
This policy letter has been revised to fully clarify the correct method of reading stat trends under the sections "Backlogs" and “The Dangerous Graph" and to reference the main policy letters containing data on reading stats and stat trends.
The subject of making up statistics is probably well known. How one draws one. But the subject of what they mean after they are drawn is another subject and one which executives should know well.
Things are not always what they seem in statistics.
A backlog caught up gives one a high soaring statistic which promptly slumps. To call the soar affluence and the slump emergency is an executive error.
When you see a leaping and diving pattern on something that can be backlogged, you can be very sure it has been.
This activity is working in fits and starts, usually only occasionally manned.
For a long time nothing is done or counted; then suddenly a month’s worth is all counted in one week.
So when you see one of these, realize that the one surge in stats is averaged out with the smaller peaks and the depressions. You have to visually average the peaks and valleys and note the trend the entire stat is taking.
In any set of statistics of several kinds or activities, you can always find one or more that are not “by luck” but can be directly caused by the org or a part of it.
An example is the “letters out” and “completions.” Gross divisional statistics.
Whatever else is happening, the org itself can improve these as they depend only on the org, not on “fate.”
So if you see the gross divisional statistics generally down or going down for the last couple or three weeks and yet see no beginning upsurge in the current week in “letters out” and “completions,” you know that the org’s management is probably inactive and asking to be removed. For if they saw all stats going down they should have piled in on “letters out” and “completions” amongst other things as the least they could do. They can push those up.
So amongst any set of statistics are those which can be pushed up regardless of the rest, and if these aren’t, then you know the worst — no management.
If you see a statistic going up in “completions” and see a falling “enrollment” statistic, you know at once the body repeat sign-up line is out.
People who graduate are not being handed their certs and awards by a Registrar but are being given them by Certs and Awards or in mass meetings, or in some way repeat sign-up is not being procured.
Thus the 40% to 60% repeat sign-up business is being lost.
This also means, if continued over a long period of time, that bad technology is present as poor word-of-mouth advertising is going around.
Look in such a case at a third statistic — Qual collections. If this is poor or very, very high, you can be sure that lack of enrollments is caused by bad tech.
A very high Qual collections statistic and a low enrollment statistic is a terrible condemnation of the Tech Division. Gross income will soon after collapse as tech service just isn’t good.
Thus you get the idea. Statistics are read against each other.
A statistic is a difference between two or more periods in time so is always comparative.
Also, two different statistics are comparative, such as in examples above.
You can predict what is going to happen far in advance of the occurrence, using statistics.
High book sales mean eventual prosperity. Low book sales mean eventual emergency all along the line.
High gross income and low completions mean eventual trouble as the org isn’t delivering but is “backlogging” students and pcs simply by not getting results. Carried on long enough this means eventual civic and legal trouble.
Low FSM commissions may only mean no FSM program. But if there is an FSM program, then it may mean bad tech. So a low completion and low Qual will mean an eventual collapsed FSM statistic also, as the FSM’s own area is being muddied up by failed cases.
High book sales, high letters out, high Tech and high Qual statistics mean the gross income statistic will soon rise. If these are low, then gross income will fall.
Bills owed and cash in hand are read by the distance between the two lines. If it is narrowing, things are improving; if widening, things are getting worse. If they are far apart and have not closed for a long while, with the cash graph below, the management is dangerous and not at all alert.
When all statistics on one set of graphs show a sinking TREND line, it is a dangerous situation.
TREND means an inclination or tendency toward a general course or direction. Thus to get the trend one would look at several weeks worth of stats.
To read the stat trend, one needs to visually average the peaks and valleys over a specific time period on the graph. It is done with the eye; there is no internal system of lines that can be drawn to assist this. One sits back and looks at the pattern as a whole and there is a definite pitch or slant that one can determine by this. That is the stat trend.
If all of these stat trends or most of them are down, the management is inactive.
When a Continental Org includes its own org on its combined graphs for area orgs, it can have a very false picture.
Its own org’s stats obscure those of the area orgs which may be dying.
Thus if you include a big function with a lot of small ones on a combined graph, you can get a very false idea.
Thus, graph big functions as themselves and keep them out of small functions of the same kind.
The Continental Org should not be part of a Continental Exec Div’s statistics. Similarly, SH stats should not be part of WW’s.
A combined statistic is, of course, where you take the same stats from several functions and add them up to one line. A very large function added into a combined graph can therefore obscure bad situations. It can also obscure a totally inactive senior management as the big function under its own management may be wholly alert and competent, but the senior management is masked from view by this one going concern, whereas all its other points except the big one may be collapsing.
The one big god-awful mistake an executive can make in reading and managing by graph is being reasonable about graphs. This is called JUSTIFYING A STATISTIC. This is the single biggest error in graph interpretation by executives and the one thing that will clobber an org.
One sees a graph down and says, “Oh well, of course, that’s. . .,” and at that moment you’ve had it.
I have seen a whole org tolerate a collapsed completions graph for literally months because they all “knew the new type process wasn’t working well.” The Tech Sec had JUSTIFIED his graph. The org bought it. None thought to question it. When it was pointed out that with the same processes the preceding Tech Sec had a continual high graph, and a suppressive was looked for, it turned out to be the Tech Sec!
Never JUSTIFY why a graph continues to be down and never be reasonable about it. A down graph is simply a down graph and somebody is goofing. The only explanation that is valid at all is, “What was changed just before it fell? Good. Unchange it fast!” If a graph is down it can and must go up. How it is going to go up is the only interest. “What did we do each time the last few times just before it went up? Good. Do it!”
Justifying a graph is saying, “Well, graphs are always down in December due to Christmas.” That doesn’t get it up or even really say why it’s down!
And don’t think you know why a graph is up or down without thorough investigation. If it doesn’t stay up or continues down then one didn’t know. It takes very close study on the ground where the work is done to find why a graph suddenly rose or why it fell.
This pretended knowledge can be very dangerous. “The graph stays high because we send out the XY Info Packet,” as a snap judgment, may result in changing the Dissem Sec who was the real reason with his questionnaires. And the graphs fall suddenly even though no info packet change occurred.
Graphs don’t fall or rise for tiny, obscure, hard-to-find reasons. As in auditing, the errors are always BIG.
Book sales fall. People design new flyers for books, appropriate display money, go mad trying to get it up. And then at long last one discovers the real reason. The bookstore is always shut.
A big reason graphs fall is there’s nobody there. Either the executive is double- hatted and is too busy on the other hat, or he just doesn’t come to work.
Bad graphs which resist all efforts to improve them are made. They don’t just happen.
A sticky graph is one that won’t rise no matter what one does.
Such a graph is made. It is not a matter of omission. It is a matter of action.
If one is putting heavy effort into pushing a graph up and it won’t go up, then there must be a hidden counter-effort to keep it down.
You can normally find this counter-effort by locating your biggest area of non- compliance with orders. That person is working hard to keep graphs down.
In this case it isn’t laziness that’s at fault. It’s counter-action.
I have never seen an org or a division or a section that had a sticky graph that was not actively pushing the graph down.
Such areas are not idle. They are not doing their jobs. They are always doing something else. And that something else may suddenly hit you in the teeth.
So beware of a sticky graph. Find the area of noncompliance and reorganize the personnel or you, as an executive, will soon be in real hot water from that quarter.
Those things which suddenly reared up out of your in-basket, all claws, happened after a long period of sticky graphs in that area.
Today’s grief was visible months ago on your stats.
The simple ups and downs of graphs mean little when not watched over a period of time or compared to other graphs in the same activity.
One should know how to read stats and what they mean and why they behave that way so that one can take action in ample time.
Never get reasonable about a graph. The only reason it or its trend is down is that it is down. The thing to do is get it up.