[Company Logo Image] Methodology   Home Publications Services Order News Notes Methodology


 

 Methodology


It is very difficult in a limited discussion like this to adequately describe the Methodology used by The Data Institute. The systems and software in use have been for the past 20 years, and even more so today, by far the most advanced in the world and thus The Data Institute are at the apex of the application of artificial intelligence and computer technology.

The first important aspect to remember about The Data Institute reports is that they are entirely computerized - from initial survey interviewing to the final printout of the report and this means that The Data Institute clients are assured of a uniform standard and quality with all The Data Institute products.

The computerization, development of the methodology and the programs took over ten years and The Data Institute believe that the generation of their databases and reports are the most advanced in the world. No other company has such advanced techniques or such thorough methodology.

The fact that the The Data Institute interview methodology is so automated means that their units costs per question or per survey are much lower than that of our competitors, thus they can offer clients very competitive prices for both published reports and CD-Roms as well as for information products. There are extensive security checks built into The Data Institute survey methodology to ensure accuracy and confidentiality.


ORIGINAL RESEARCH


The Data Institute reports are based on wholly original research. There is little confidence in company accounts, credit reports, government or trade statistics and thus one has to approach the problem of getting corporate intelligence and market data from a number of standpoints.


DATA SOURCES


Data on the target company, its suppliers, internal operations, products, competitors and markets is examined at four levels of commercial activity:-

The Input Level:
An analysis of the Input materials, supplies and services bought or consumed by the target company, the type of goods or services bought, value products bought, quantities sourced, prices paid, scheduling of purchases, et cetera.

The Process Level:
An exploration of the internal processes at the target company which combined the Input products with the procedures that manipulate the inner components (Financial + Capital resources, Management, Labor, Equipment usage, Physical processes, Product production, Premises & Locations, Distribution, Marketing, etcetera) to produce the Output product - whether that be physical products or services.

The Competitor Level:
An investigation into the major competitors of the target company and the company's relative performance, its product offerings, its service, its quality, et cetera, in relation and relative to the other competitors.

The Market Level:
An appraisal of the interaction of the target company with the various market elements and the problems and opportunities for the target company within the markets in which the company operates and within market in which the company may potentially operate in the future.

For example, the market consumption data provided in the databases will be based on three independent sources:-

i. A statistically accurate survey of End Users, i.e. the market.

ii. A statistically accurate survey of the Distribution System.

iii. A survey of all major Competitors, i.e. the industry base.

By gaining information from these levels of market activity one can produce a very accurate picture of the market.

The same thorough methodology prevails throughout the entire study and for this reason the reports are of such a high quality.


EFFICIENT PROCESSING OF DATA


In addition to the national on-line searches and the gathering of other public and private information, the results of survey interviews are entered by the The Data Institute interviewers directly into a micro-computer or intelligent terminal and after each interview session the results are directly transferred to one of the main The Data Institute computers.


EXCELLENT FORECASTING MODELS


The Data Institute forecasts are produced using an extremely comprehensive matrix of forecasting models which take the best features from a wide range of forecasting models and blended these into an interactive matrix which ensures a high level of certainty in forecasts produced.


HIGHLY SOPHISTICATED REPORT GENERATION


The databases and subsequent reports are entirely composed by computer using the base data provided by the various surveys and from other sources. The computer's programs construct the report which is then printed out by laser printer or output on CD-Rom and it is these originals which are supplied to the client who thus receives the most up-to-date information.


AUTOMATIC UP-DATING


The Data Institute reports are automatically up-dated by the computer when new surveys are completed or company / market / competitor conditions change and thus a new forecast is needed. This is done as often as may be necessary (and especially with Consumer, Seasonal or High Technology markets where up-dates are done weekly). The majority of reports need only to be updated monthly.


INPUT FROM CONSULTANTS + SPECIALISTS

meth_consult.jpg (58743 bytes)


SURVEY METHODOLOGY


The Data Institute believe that the only reliable way to analyze and evaluate the market-place is by conducting statistically accurate analyses of the Suppliers, Bankers, Financiers, Service Providers of the target company and then the Distribution Channels and End Users of the company's products and services. Also surveys are necessary for the Competitors of the company and the overall Industry and Market environment. By correlating the results of these analyses one can provide highly accurate data and analyses on the target company, its Competitors, the Markets, the Products, the Marketing, the Industry, the Distribution and the Customers.


END USER SURVEYS


The most important part of the base data collection are the END USERS SURVEYS, (i.e. the customers of the target company and the various competitors), as these reveal the true nature of the market-place.

The END USER SURVEYS conducted by The Data Institute use the following formula:-

1. SELECTION OF INTERVIEW PANELS. The prospective interviewee panel is selected from lists which are maintained by named respondent in each of the research areas.

2. MAILSHOT TO POTENTIAL INTERVIEWEES  A letter is sent to the potential (Industrial or Commercial) interviewees explaining the nature of the survey and the products and markets they wish to cover and asking if the respondent would help.    (Industrial and Commercial markets only)

3. PILOT SURVEY (5% OF SAMPLE)   A Pilot Survey is conducted with 5% of the sample to evaluate and modify the survey and isolate potential response problems.

4. MAIN SURVEY (90% OF SAMPLE)   The Main Survey is conducted amongst 90% of the sample using the modified questionnaire.

5. CHECK SURVEY (5% OF SAMPLE)    A Check Survey is conducted with the remaining 5% of the sample to check outstanding points and verify any problem areas.

6. CONSUMER / END USER PERSONAL SURVEYS     Where telephone surveys are not appropriate, personal or face-to-face surveys are conducted with respondents.


SAMPLE STRUCTURE


It is a standard technique of the surveys of End Users, the Distribution Channels and Suppliers to interview three levels of respondents:-

1. the majority of the sample is taken from respondents CURRENTLY involved with the company and/or the product,

2. a sample is taken of respondents FORMERLY involved with the company and/or the product, and

3. a sample is taken from respondents who are POTENTIALLY likely to be involved with the product or company in the future.

This method gives a very accurate picture of the development of the product and market over a period of time.

Thus it is possible to evaluate and analyze the reasons:-

i.    why respondents currently supply, distribute or consume the products

ii.   why respondents have ceased (for reasons of technical or technological developments, product obsolescence or substitution, et cetera) to be involved with the product, and

iii.   lastly why respondents are planning to become involved in the product (for reasons of new product development, new production facilities, acquisition, diversification, et cetera).

TOTAL SURVEY SAMPLE

Product Flow

Formerly
involved

Currently involved
in the Market

Potentially
involved

±

=

±

Suppliers
or Competitor

±

=

±

Distribution
Channels

±

=

±

End Users
& Buyers

>> Product or Market over TIME >>


The samples of the Distribution Channels and End Users are composed of a stratified random sample. The stratification of the sample reflects analysis and modelling of frequency variables (e.g. level & value of product flow), sample dispersion (e.g. type of respondents in the universe), distribution skewness, product correlations, regression, probability and significance, time projection and trends.


SAMPLE SIZES


The Data Institute guarantee a minimum VALID sample size for each of the databases. Sample sizes vary from market to market, however the following examples show the parameters that apply:-


INTERVIEWING METHOD:

Sampling is carried out by personal and/or telephone interview. The exact interviewing method used depends on the complexity of the product or market as well as the depth of information sought.

Country

Agriculture, Consumer, Travel, Catering

Electronics, Industrial, Consumables

Engineering, Motor Transport, Components

Capital Goods, Property, Financial, State Enterprises

Average Sample Size

U.S.A.

3500

2000

2500

1000

Austria

500

500

300

100

Belgium/Lux.

800

800

500

300

Denmark

500

500

300

100

Eire

500

500

200

100

Finland

800

500

300

100

France

2000

1000

1500

600

Germany

3000

1000

1500

800

Italy

1500

800

1000

500

Netherlands

900

800

500

600

Norway

800

500

400

500

Portugal

600

400

300

500

Spain

1000

600

800

900

Sweden

900

600

600

900

Switzerland

800

500

500

600

United Kingdom

1500

1000

1000

700


SURVEY ANALYSIS:

Personal interviews are recorded on data sheets which are then encoded and processed by the The Data Institute computers. Telephone interviews are conducted by interviewers who place the information received directly into a micro-computer or intelligent terminal. The data is then transferred directly by modem to the main computers for batch processing.

DISTRIBUTION CHANNEL SURVEYS:


The Distribution Channel Surveys use the same methodology, Interview Method and Survey Analysis as do the End User Surveys.

Country

Agriculture, Consumer, Travel, Catering

Electronics, Industrial, Consumables

Engineering, Motor Transport, Components

Capital Goods, Property, Financial, State Enterprises

Average Sample Size as a % of the Total Universe

U.S.A.

10

15

15

15

Austria

10

25

10

20

Belgium/Lux.

10

20

15

25

Denmark

10

25

10

25

Eire

10

25

25

25

Finland

10

20

20

20

France

10

20

15

15

Germany

15

20

20

20

Italy

10

20

15

15

Netherlands

10

25

25

25

Norway

10

20

20

15

Portugal

10

25

20

25

Spain

10

25

25

25

Sweden

10

20

20

25

Switzerland

10

20

20

25

United Kingdom

10

30

20

25

SAMPLE SIZE: Sample sizes for the Distribution Channel Surveys are based on a percentage of the total number of companies distributing the product/s concerned. The percentage interviewed depends on the industry in question and the complexity of the product. In general however the above average sample size are used.


SUPPLIERS SURVEYS:

The Surveys of Suppliers and Service Providers use the same methodology as the surveys above.


SUPPLIER SAMPLES:

The sample of Suppliers represents at least 70 to 75% of the total universe. The surveys of Suppliers yields confirmation of supplier output to the target company, product information, financial data and future plans. Suppliers are regarded as being a reliable source of information on the target company as they tend to be keen to discuss their selling prowess, are informed about the procurement of the company and are knowledgeable about the products purchased by the target company and thus their internal processes.


COMPETITOR SURVEYS:

The Surveys of Competitors the same methodology as the surveys above.


COMPETITOR SAMPLES:

The sample of Competitors represents at least 70 to 75% of the total universe. The surveys of Competitors yields verification of product output & specifications, financial data and future product and market plans. Competitors are considered a reasonable source of information on markets as they are inclined to promote their future plans, marketing activities and product performance as well as their expertise in market and competitive conditions.


TELEPHONE SURVEY METHODOLOGY

EXAMPLE:
If an interviewer is surveying a truck operator then the computer will format the questionnaire according to the makes and models of trucks operated by the respondent. The computer also imposes range checks and input screening according to particular makes and type of trucks; therefore, if the interviewer asks questions about the engine life or gearbox life the computer will only accept answers within certain ranges. These ranges will be based on the known average life of engines or gearboxes of particular truck makes. The computer monitors answers and will amend the known standard deviation of response ranges. As answers change (due to technological or product development, changes in buying patterns, et cetera) the ranges and input screens are automatically amended.


The local interviewers receive the survey questionnaires and the names and telephone numbers of the survey respondents and the questionnaires and contact lists are stored on computer. The reverse method is used by the interviewers to transmit the survey results to The Data Institute at the end of each interview session.

The interviewers use their micro-computers when interviewing. The questionnaires are displayed on screen and the interviewers asks the questions which appear and type in the respondent's answers.

The computer firstly displays the respondent's name and telephone number. A percentage of the telephone numbers displayed are "check" numbers and the interviewers have to identify these; this assures that all the interviews are being conducted and therefore guarantees the integrity of the survey.

The questionnaire displayed on the screen is interactive and intelligent and will not allow the interviewer to input answers which do not comply to certain range checks, input screening, and correlation with preceding answers.

The interviewer's computer also :-

1. Monitors the number of interviews done for each questionnaire and thereby ensures that the correct sample size is achieved.

2. Records spoiled questionnaires which have been rejected because of insufficient data received or incorrect input ranges. These are then analyzed to establish why answers did not correspond to the range checks.

3. Records changes of contact names or job functions.

4. Monitors and ensures that the required number of "check" respondents have been identified and thereby ensures the integrity of the sample.

5. Provides notes and explanations for interviewers and respondents which facilitate the understanding of the questions asked.

In this way relatively non-technical interviewers can be used to survey technical and specialist products and markets without the risk of the accuracy of the input data being compromised.

  method_tele1.jpg (39228 bytes)


TELEPHONE INTERVIEWERS USING AUTOMATED SYSTEMS

method_tele2.jpg (3295 bytes)


EXAMPLE:

If an interviewer is surveying electronic components buyers then the computer will format the questionnaire according to the equipment being manufactured at the respondent's factory or plant. Therefore, if the respondent's factory is manufacturing televisions, the respondent will be questioned specifically about components used in television manufacture. The range checks imposed by the computer will also be critical to the product quantities manufactured; therefore, the computer will only accept input data (on passive components, sub assemblies, et cetera) which complies to known parameters. Thus, for a given number of television sets manufactured, the computer will expect to input certain numbers of each component or sub-assembly which is critical to both the products and the quantity manufactured.


DATABASE STRUCTURE


The Data Institute databases are composed of individual company and product databases which are then aggregated to form report databases.

The actual report will be composed of a number of database levels which reflect the company and their individual products being investigated. The Data Institute databases are held at a number of levels and thus editors can choose exactly how detailed the information they require need be.


FOR EXAMPLE,


If one were to analyze a target company mainly producing
ACTIVE ELECTRONIC COMPONENTS
the corporate database structure would be as follows:-

1. Isolate those Suppliers which are providing the Capital, Input materials and Services which is used by the target company.

2. Analyze the Processes of the target company which take the Capital Resources, the Input materials and the Services provided, merge them with the Internal activities of the target company and thereby Output the product:-

3. Review all the competitors of the target company and determine how these competitors will impinge in market terms.

4. Investigate the Markets, actual and potential, for the target company.

datastructure.jpg (117349 bytes)


DATABASE MATRIX AND INTERACTIONS

FOR EXAMPLE,

If one were to forecast revenue and sales for a target company which produces

VEHICLE ENGINES & COMPONENTS,

the database matrix and interaction would probably be as shown:

The same database management methodology is used to produce all The Data Institute company databases and their concomitant reports.


It is necessary when forecasting corporate, product and market trends to use a matrix of related databases in order to evaluate and analyze the movement of products and markets due to supply and demand, product and market forces, product life cycles, product obsolescence and substitution, new product developments, technical and technological innovations, market saturation, new markets, market and product segmentation, et cetera.

datamatrix.jpg (112818 bytes)


This database management methodology ensures that The Data Institute reports are extremely accurate as they use the most discrete and detailed company operation, product and market sector as the foundation, and then, builds up the database step by step.


THE FORECAST/ PLANNING MODELS

In order to accurately analyze and project the above areas, a battery of forecasting models are used.

These models are interactive and simultaneous and draw from a common database which is designed to be critical to the report concerned.

The complexity of the market / product / company and the desired accuracy of the prediction dictates how many of the battery models are used. The more complicated the product, market or company the greater the need for accuracy and thus to ensure this more battery models are used.


The success of a Forecast/Planning model, for periods in excess of two years, depends on the ability of that model to analyze and evaluate a series of interrelated levels of corporate, economic and commercial activity. Each level tends to be equally critical to the forecasting method and thus accuracy must be maintained at all levels. The Forecast/Planning Model used by The Data Institute is complied from a number of programs and seeks to interact the various levels of corporate / economic / commercial activity, not only through a series of forecasting routines, but also via a number of refining procedures: the distillate of which represents the Forecast.

The levels of economic / commercial activity are as follows:

1. The Trade Cell (e.g. The EU or NAFTA or OECD countries)
2. The National Cell (e.g. The United States)
3. The Industry
4. The Suppliers
5. The Company
6. The Product
7. The Competitors
8. The Consumer

Level

Model

OECD
Econometric
Model

PIMS
type
model

Business
Environment
model

Consumer
Values

Product
Life
Cycles

BPI
+
QI

Advertising
Efficiency
model

Trade Cell

KSIM

National Cell

Industry

TIA

Company

Product

CIA

Consumer


I. OECD ECONOMETRIC MODEL:

The model program used is the standard OECD Econometric program which has been developed by the OECD Department of Economics and Statistics in Paris. The program, as used by The Data Institute utilizes a number of databases beginning in the 1970s with the basic econometric databases, including these:

a) The International Financial Statistics supplied by the I M F in Washington. These are: IMF.1/FMB, IMF.2/FGNSTAT,
IMF.3/FNDACCT, IMF.5/FOOTMST, IMF.6/FBOPMST, IMF.7 - et cetera.

b) OECD databases on wages & prices developments, consumption, investment, trade, et cetera.

c) In addition the following databases are also available for specific applications: UNSO.1, UNSO.2/NAMAST, UNSO.3/ICPDATA,
UNSO.4/WORLDENERGY, UNSO.5/STAPC, UNSO.6, UNSO.7/UNIDUST, ECE.11059A, ECA Series, ILO.1, FAO.1, GATT.1/TTDF - et cetera.

d) National governmental databases, including Customs' databases, Sales Tax / Value Added Tax databases, Internal Revenue Tax databases, Social Security databases, Judicial databases, et cetera.

e) Since that time the regressional databases have expanded enormously to include most of the major multi-national econometric databases.

II. PIMS TYPE MODEL: PIMS
uses an industry critical Multiple Regressional Analysis program which identifies those variables affecting industries, companies / products, their market share and profitability.

III. BUSINESS ENVIRONMENT FORECAST MODELS:
Three models are used to forecast business environment and these are specific to certain levels of economic / commercial activity:

Cross Impact Simulation (KSIM): This is a program for interactive variables and is used to forecast business environments in an international and national context and provides a representation of changes in a system of variables over time; especially trends, events and structural relationships. The program is widely used in the U S and users of the U S Army, U S Navy, NASA, National Science Foundation and many multinational companies.

Trends Impact Analysis (TIA): TIA is used for problems involving changes in trends at Industry and Company level and produces a representation of changes in trend/s and event/s. A further representation of probabilities is also produced. The program is widely used by U S multinational companies.

Cross Impact Analysis (CIA): This is a program designed for use in solving project level problems, including new product developments, and deals with a representation of cumulative probabilities of an event over time. The model investigates events, structural relationships and probabilities and is successfully used by many U S multinational companies.

IV. CONSUMER VALUES MODEL:
The use of Consumer Values models have two purposes. Firstly because it extends and correlates market behavior beyond such external forces as Disposable Income or Demographics and secondly, it enhances the predictability of market behavior by providing insight into why consumers act as they do. The Data Institute have trend projections for 14 years for all Consumer Values; some 34 to 40 major cells. These values affect consumer spending on definable items and markets. In the case of Capital Equipment one additional value is used, namely, the monthly Business Confidence Monitor, which seeks to quantify purchasing trends amongst buyers.

V. PRODUCT LIFE CYCLE MODEL:
The Product Life Cycle Model uses the normal innovation / diffusion Gompertz Function with the added advantage of a matrix of industry critical correlation.

VI. BPI & QI MODELS:
It is essential to effectively measure the relative buying power of various market segments and geographic markets, this is done with a Buying Power Index and a Quality Index. BPI is also useful for determining how effectively a company uses its marketing effort. QI is used to analyse the ability of a market sector or geographic market to purchase unessential goods with a high discretionary index.

Additionally, a number of other programs may have to be used, including:

i. Market Saturation Index;

ii. Inventory costs;

iii. Recorder points (Trade);

iv. MNR:MC ratios

VII. ADVERTISING EFFICIENCY MODELS:
These models evaluate the effect of revenue / advertising relationships over time and is designed to investigate those functions of advertising efficiency which affect revenue and to provide predictions on the actions of those factors over time.

The main functions of the model are:
1. The Sales Decay Constant, which measures advertising effect over time.

2. The Market Saturation Level, which investigates the effectiveness of the media mix and the asymptotic values of the media expenditure.

3. The Sales Response Constant, which gives the ratio between revenue and advertising.

The above models provide the Coefficient of Advertising Efficiency at a given revenue / advertising expenditure function.


Home Publications Services Order News Notes Methodology