Information boards, indicative systems and management

Several times I had to deal with the tasks formulated something like this: "We need to build an information board." These were different organizations, they planned to use their boards for various purposes; they were united only by the fact that all these goals were in the control plane. The use of the term "information board" does not quite fit with the management; it is closer to the “reference information” topic, such as here ; therefore, I will further use the term “indicative system”. But this is not the essence of the article. It was interesting to summarize the main points of building automated systems of information indicators for the management of the enterprise .

As it should be, let's start with the appointment of indicators. In one way or another, they serve as informational support for making management decisions.

Indicator classification

First class of indicators

based on planned (or standardized) performance indicators. For trading it is, for example, the amount of sales , margin , stock in days of sale , etc. In fact, there are much more indicators (there are dozens of them), but now we are interested in those of them for which they assign planned values. In advanced offices, they increasingly talk not about “plans”, but about “KPI target values” (within the framework of the BI-management approach, which is briefly and clearly written here ). For myself, I divided this way: “target KPI values” are used in the development mode, and the “plan” is used in the operation mode of enterprises. But in both contexts, the use of a plan or a target value immanently implies a plan-fact analysis (PFA).

The PFA provides information on the mismatch between actual and planned values, but, as a rule, it is still not enough to make a management decision: well, we see that the sales plan is overwhelming; What to do, how to fix it? A thoughtful manager follows a path that the wise Japanese have summarized in the principle of “5 why”, and gets to such levels of detail of the problem, when it is clear by what actions you can influence the situation. And this is where the indicative system is very relevant . Formed several typical tracks analysis of the situation on the basis of the PFA, which are surrounded by indicators. Running along such paths becomes much faster .

In width, the situation is divided into several functional blocks, and then, for example, total sales are represented as
number of buyers × average amount of goods in check × average price of goods .
')
Already it is easier, activity forms are already peeking. If the number of buyers decreases, probably, marketers with advertisers have won over or have activated a competitor to blink. If the quantity of goods in the receipt is junk, then you need to look for in the calculation - the buyer cannot find the right one or the product matrix is ​​broken - they are separated from the needs of the buyers. With the average price in the check, too, their reasons are connected, for our purposes, no matter what; while the number of indicators is growing .
And this set of indicators still needs to be looked at in depth: in the context of stores, departments, product groups, category managers, suppliers, etc. Here, “at depth”, comparative analysis works very well and, through indicators, iron out problems with a specific store, manager, supplier.

So we have assembled a part of the indicative system that serves the PFA ( PFA indicators ). How does it differ from just the “information system”, which in the form of reports can give out the values ​​of all the specified indicators? The difference is that the “indicative system” is closely related to business processes (BP), and therefore special requirements apply to it:

• typification of a set of considered indicators;
• their direct connection with the activity (the value of the indicator can be influenced by such and such definite actions);
• unification of forms of their representation;
• the efficiency of their provision.

Such an indicative system of the class PFA fits well with the operational forms of management: in the "weekly briefs" and "daily five minutes". Managers quickly walk along typical PFA tracks and quickly find typical action plans (PD). The key word here is “fast.”

Of course, “the world is more interesting than it seems to us,” and we do not find all the answers on the PFA model tracks. In unusual cases, we are helped by immersion in a variety of data collected by information systems, intuition and other techniques of thinking people.

The second class of indicators

closer to the "land". These are indicators that are not just “closely related” to the BP, but are their direct part: on the principle of “light bulb -> action”. For example, if the return limit agreed upon with the supplier has not yet been chosen, then the bracker issues a return of the defective goods; if you have already chosen - then draws up a cancellation.
Such BP-indicators are designed together with the BP itself. They are intended for line personnel (operator level). In contrast to the PFA indicators, they differ in the unambiguity of the actions that entail. The more processes of the company will be covered by such BP-indicators, the more predictable will be the behavior of lower-level personnel; the more operations can be brought down from managers to operators.

The third class of indicators

The third class of indicators is related to business development. If a manager wants to estimate some perspective, test a hypothesis, or just keep abreast of his ideas, then he covers up the idea with indicators and monitors them, making his “big conclusions”.

Yes, he may require the assistant to form him a weekly report from the IP, but advanced top managers automate this on his “information board”

(for the P-indicators, such a term is already quite appropriate).

Stages of design of indicative system

They are fairly obvious:

1. create a set of indicators (indicators);
2. build mechanisms for collecting (calculating) these indicators;
3. to construct mechanisms for the derivation of these indicators with regard to ease of use and separation of access rights

Formation of a set of indicators

At first glance, this task lies outside the IT field. IT specialists cannot determine which parameters should be used for business management. But here's the catch. If this set of indicators was chosen speculatively, directively, ill-conceived, without reference to real activity , then it is highly likely that further efforts to collect and provide these indicators will be lost in vain .

For example, in one retail chain, in order to manage the activities of sellers in retail stores, they decided to introduce marginal indicators. We formulated the task for the IT department, which performed it successfully (from the point of view of the task itself): information on margins is formed with the required detail, with the required efficiency, at the required workplaces. But sellers do not know what to do with this information. Of course, they were instructed that there are marginal goods, and there are non-marginal (so-called “locomotives”, “indicators”, “magnets”, etc.) and, accordingly, the focus should be on sales first, not second. However, the real activity dispersed with a benchmark: the buyer - with his needs, and the seller - with his margin (apparently not the sellers have to manage the margin). Therefore, in the best case, this indicator is ignored - “Be it as it will be” (as a result, just lost time in vain on the setting). And there is also a worse option: they are trying to somehow optimize this system, bring it to mind. They suggested, for example, that marginality be calculated with an accuracy of a delivery batch so that sellers could “predict” it “more accurately” - wait for the expensive consignment to go away and pounce on a batch cheaper with higher margins. Thank God, they got out, they did not do it.
If IT does not participate in the formation of a set of indicators, then at least it should be aware of the possible rakes. And since IT managers are mostly a model of rationality, they can be quite successfully attracted to such work, at least as experts. It would be very successful to consolidate such a rule in the company as the basis of an indicative system.

Building a mechanism for collecting (calculating) indicators

This is an IT task itself, and there are three paths here.

Road One: Revision of Transactional IPs - Embedding the Required Indicator Set

For me, this is quite an acceptable way . Maximum reliability, good speed. As a rule, the operational level configurations are far from “typical”, therefore adding new functionality will not “spoil” them anymore. But the biggest advantage that such a path gives is the possible integration of the indicative system with the BP control system (up to their complete merging): well, the IP will not give you a refund if the limit is chosen or the manager has not allowed a special action.

Of the minuses . Not all integrated circuits allow their functionality to be easily modified, not all of them are “designers” - many are “closed”. Sometimes this closeness is justified and purposeful; sometimes these are platform limitations and archaism. But now it’s not about that: the fact that not an indicative IC can be embedded in every transactional IP.
Another minus is the need for its programmers or the costs of external support.
The additional computational burden on the IC is also a minus, for example, when it is tempting to see the sales margins when drawing up checks.

Road Two: Upload to Excel

Dear you are our Excel, how are you good: flexible, fast, affordable . Today agreed on a set of indicators; a day later we saw a prototype; a week later - a powerful analysis tool. Everything is wonderfully consolidating or, according to your wish, it turns around to the store, to the goods, to the manager, to the supplier, to everything that was laid. And universal filters and graphic diagrams - you can see everything at once.

But how do you depend on the human factor, how are you ruined by multi-stage unloading-loading, preliminary processing of data.
I had a chance to see a very developed and powerful (breathtaking) indicative system built on Excel. But practically at every planning meeting (the basis of which the system was laid), the responsible person declared that such indicators were wrong, they needed clarification, verification, etc. And how many inaccuracies he did not announce? ..

In general, Excel is useful for testing the design of an indicative system. In the working mode - only if you trust the person responsible for it, well, and you have external mechanisms for checking the values ​​of the indicators.

Road Three: Business Intelligence Systems

They are also called BI-systems. They connect to transactional integrated circuits (directly to databases or via uploads), upload data into themselves, stack it into their OLAP-cubes and give users convenient tools to view them.

These are such information eaters. They swallow an incredible amount of checks and other transactions, and they give out indicators that are neatly arranged in the necessary racks-cuts. You can connect them to almost any, incl. most "closed" IP. Very convenient is that one BI can simultaneously collect data from various and very heterogeneous databases, of which even a medium-sized company can do quite a lot. These are all pluses.
There is another plus sign. Some of them have such simple mechanisms for customizing indicators based on the data they upload, that not only the local programmer can figure them out, but the advanced manager himself will master them - then he can complete the task of forming the P-indicators of his own.

Among the minuses . Bulkiness: always a separate server, separate data loading procedures, separate procedures for accessing indicators (you work in the transactional information system, and you need to see the indicators in BI).
If the work is built through the unloading of data from the transactional information system, then we catch all the limitations of this unloading - primarily in terms of the composition of the data being uploaded (but this is rather a limitation of the transactional information systems themselves).
If the work is built through a connection to the database of transactional information systems, then an interesting task grows to understand, and where, in which tables and fields the data we need is stored. The interest in this task is given by the absence of a description of the database structures, which is not such a rarity. Outsourcers for connecting to our databases request such indiscreet rewards, that we thought it appropriate to hire another programmer to figure it out.

Constructing mechanisms for displaying indicators

From the point of view of the presentation of indicators, Excel and BI have an advantage over transactional IP: there are more available types of charts and graphs.

From the point of view of differentiation of access rights, the leaders in terms of convenience and capabilities are transactional and BI systems. In Excel, this problem is usually solved through the formation of individual files that are laid out in folders with different rights. This adds Excel one more minus - so many related files are obtained.

So what am i talking about

The indicative system is part of the IS, which is very closely connected with the BP, with operational activities. Therefore, special requirements are imposed on indicators for the typing of the presentation form and the speed of their formation.

There are three main classes of indicators:

• PFA - indicators that follow from the plan-actual analysis and allow during the PFA to quickly find typical causes of deviations and the appropriate actions to eliminate them.
• BP-indicators, which are directly integrated into the business processes on the principle of “light bulb-action” and allow, thus, to take many actions from managers to operators.
• P-indicators are development indicators with which top-level managers test their hypotheses and predictions.

There are three ways to implement the indicative system:

• integration of indicators into transactional IP,
• using Excel,
• using business intelligence systems (BI-systems).

Happy New Year! Such is the indicative Christmas tree .