In the past, internal data was the starting point in any research (see Figure 3.2). The idea being that there may be no need to look outside the organisation when required data could be found within. The next step was to look a little further, outside the confi nes of the organisation: this meant inspecting external secondary data. However, information technology has revolutionised the retrieval of data and the idea of looking at internal data fi rst, because of ease of access, has less relevance today. It is perfectly possible that it is easier to
fi nd government data on the purchase behaviour of 2000 people than it is to fi nd customer invoices from an accounts offi ce in the same building (see Table 3.2).
It is worth noting that large corporations have integrated their own databases into ‘intranets’ and ‘extranets’. This organisation of data makes it easier for information to be made available widely and quickly inside the fi rm. Popular search engines and information professionals have adapted their services to allow user-friendly searches of desktop computers and related networks. One such tool is Research Reporter software (see http://www.researchreporter. com), which allows large users of research to consolidate their research, so that databases of surveys are linked and are cross-searchable. Please refer back to Chapter 1 for more details on marketing decision support systems.
External secondary data
The dictionary and thesaurus are good starting points for any search of external data, not least
to provide synonyms for subsequent searching. Many reference books are available, but it is worth mentioning that the UK standard dictionary is the Oxford English and the standard for
Part 2 Data collection
74
the USA is Webster’s. Similarly, any encyclopaedia can provide essential data; these are now
easily available on the Internet and give knowledge along with terminology. They include the
Encyclopaedia Britannica (http://www.britannica.com), and Wikipedia (http://www.wikipedia.
org), available in several languages (see Figure 3.3).
The Internet
In 1965, there were fi nal agreements, based on work by the Pentagon’s Advance Research
Projects Agency, to link the computers of three American universities: the Stanford Research
Institute, Utah University, and the University of California. This was the start of the Internet.
In the early 1970s, computer developments promised a fast and fl exible way to track down
documents. The 1970s also saw the quiet arrival of fi rms that later took the title of ‘aggregators’.
These organisations combined information from various sources and made it available to
customers; three such big names are now Dialog, LexisNexis, and Factiva, which were to play a
big role two decades later when the Internet fully emerged.
In the 1980s, a great deal of knowledge was being shared on the Internet: for example, full-
text access to newspapers, magazines, and journals started to appear. There was clearly a need
to fi nd information in an effi cient way, so this was about to be created. We will look at search
engines, but fi rst, to set these into context, we will look at the Internet itself.
– Complaints – Service records Sales
Inventories
Distribution
costs
Adspend
Prices Internal
secondary data
Sales team records
Figure 3.2 Internal data resources
Chapter 3 Secondary data
75
In 1989, the World Wide Web became an integral part of the Internet as a simple method of publishing and presenting information. The web was fi rst devised by Tim Berners-Lee, working
at CERN, Europe’s particle physics research centre; it was a way of linking documents. A web page is similar to an encyclopaedia page, which includes cross-references. In a book, to follow
up a cross-reference, you need to fi nd another volume, then the page number, and then the exact paragraph. The web is the same, except that it will fl ash up thousands of references within seconds—even if the relevant documents are on another continent. Browser software
is needed to gain access to the web, and the best-known browsers are Internet Explorer and Firefox.
By 1996, over 30 million users were linked together, and the Internet, or the ‘Information Superhighway’, had established itself as a cheap method of making information available to the masses. The Internet performs a wide range of tasks including fi nding a job, fi nding a car, locating friends, fi nding book references, buying or selling any product or service, fi nding Index
Guides Abstracts
Annual reports
Trade press
Professional bodies Govt
sources
Commercial sources
Academic sources
Others
Industry sources
Tr Internet p
Census
Others
External secondary data
Figure 3.3 External data resources
Part 2 Data collection
76
news, answering questions received, downloading software, searching for information, fi nding
wanted criminals, etc. It off ers the researcher a window on the world, a new avenue, a new
way to access people and facts.
The very fi rst tool used for searching on the Internet was called ‘Archie’. It was created in
1990 by Alan Emtage. The program downloaded the directory listings of FTP (fi le transfer
protocol) sites, creating a database of fi lenames that could be searched. The next tool used
for searching the Internet was ‘Veronica’. This program downloaded menu titles from Gopher
servers, creating an index of plain text documents that could be searched. In 1993, Matthew
Gray created the World Wide Web Wanderer. Excite was also introduced in 1993, and this
engine used statistical analysis of word relationships to aid in the search process. Jerry Yang
and David Filo created Yahoo in 1994. Unlike the Wanderer, which only listed each URL,
Yahoo featured a description of the page content. Also in 1994, WebCrawler was introduced
as the fi rst full-text search engine; the entire text of each page was indexed for the fi rst time.
And again in 1994, Michael Mauldin created Lycos, a large search engine with relevance
retrieval, prefi x matching, and word proximity.
Alta Vista began in 1995 as the fi rst search engine to allow natural language inquiries and
advanced searching techniques using Boolean logic; in the same year, Metacrawler was
developed by Eric Selburg. This was a novel idea whereby the search query was forwarded
to several major search engines simultaneously, clearly taking advantage of the strengths and
weaknesses of each. This has since been called a ‘metasearch engine’. Ask Jeeves and Northern
Light were both launched in 1997, as was Google. Google was created by Sergey Brin and
Larry Page as a no-nonsense, quick engine, delivering very relevant answers to queries. In
1998, MSN Search and the Open Directory were also started. The Open Directory is a human-
edited directory of the web, using volunteer editors.
It is fortunate that these search tools were developed because more information was soon
to become available as a direct result of market demand and legislation. Market demand
meant that aggregators became more important, delivering material from many sources to
users, often free of charge.
We have seen much change in how materials are stored, located, retrieved, and used. From
the very rigidly organised storage of documents according to the Dewey classifi cation or other
hierarchy, we are moving towards an age where data is stored as it is captured, in an apparently
unorganised way, with little or no fi ling. Electronic searching and data capture makes sense of
this and delivers the relevant information to the user.
Sources of secondary data
Government sources
Most governments in the world have statistical departments and these provide a list of
publications available. Typically, such offi ces have general population census records and
specialist censuses, for example, on agriculture, industrial output, housing, and construction.
The European Union (EU), as an institution, has extremely good resources: it has a
statistical offi ce called Eurostat, which collates fi gures from Member States. The idea was
Chapter 3 Secondary data
77
that decision-making and plans for European Community members should be based on reliable and comparable statistics and a coordinator was required. Therefore, in 1953, the European Community created Eurostat (see http://epp.eurostat.ec.europa.eu). Eurostat serves community institutions with information that helps them to design, manage, and evaluate EU policies. This information is also made available to public entities in Member States, to educational establishments, to companies, the media, and others. The information comes from the Member States, the United Nations, The World Bank, and the Organisation for Economic Co-operation and Development (OECD, http://www.oecd.org). Information
is summarised in the Eurostat Yearbook, published every 12 months. It also comes from specially commissioned studies such as the Eurobarometer series. This is a continuous monitor of opinions conducted in the EU using samples of around 1000 people per country.
The UK has a statistical offi ce and two publications: the Annual Abstract of Statistics and The
Guide to Offi cial Statistics provide excellent pointers to information. Frequently used research
agencies, notably Ipsos MORI, also publish reports to access a wider readership (see http:// www.ipsos-mori.com/researchpublications.aspx). The BOPCRIS service identifi es relevant British offi cial publications from the period 1688–1995 with its web-based bibliographic database (at http://www.bopcris.ac.uk). For Scotland, Scottish Executive Statistics are available (at http://www.scotland.gov.uk). Similarly, France has a statistical offi ce called INSEE, and the Italian government offi ce is ISTAT, and the same pattern repeats itself throughout all EU member nations. The USA has a major resource created by the Central Intelligence Agency (CIA). It is an online book that examines all countries of the world (http://www.cia.gov/cia/ publications/factbook/). It provides statistics including for each population, economic overviews, transportation, government, and maps. The US Department of Commerce site at http://www.commerce.gov also off ers such information, organised by industry sector.
The census and other major surveys are shown in Table 3.3. These clearly off er useful secondary data sources since the date started; this date indicates the extent of longitudinal data available for analysis. Survey sizes show the comprehensive nature of such work, but
do not always show the full numbers of interviews: some studies interview all adults in a household; others interview only one. Size usually refers to that for the full cycle of the study or for a 12-month period. The survey sizes in Table 3.3 are based on recent results and have been rounded up or down to the nearest thousand.
Academic sources
The academic world also works continuously to produce new knowledge, and usually this
is given free of charge and readily. In all cases, this knowledge must be organised in some way. We will see that these ways diff er, but it is important to be aware of the Dewey system. The Dewey decimal classifi cation system (DDC) is commonly used in educational libraries throughout the world. It is made up of ten categories. These are further subdivided. When seeking a book, one must be aware that class marks have more numbers after a decimal point, showing the degree of specialisation within the given subject area. Books are physically placed
on bookshelves in decimal order, but because more than one book will carry the same class mark, the fi rst three letters of the author’s name is added to the end; for example: 658.0072 BRY, 658.1599 BUC, 658.3 CAR, 658.408 RES, 658.802854678 MIC, etc.
Part 2 Data collection
78
The Economic and Social Research Council (ESRC) is the UK’s leading research funding and
training agency, addressing economic and social concerns. Its intention is to provide high-
quality research on issues of importance to business, the public sector, and government. There
are several conditions that must be met before the ESRC provides money for research. One
condition is that raw data sets are deposited with the archive and the second is that there must
be some tangible output in the form of published papers, conference presentations, book
Major UK social surveys
Research studies Frequency Started Respondent Survey size
Census of population Decennial 1801 Adults All households
Agricultural and
Horticultural Census
Annual 1866 Farmers 110,000 farms
International Passenger
Survey
Continuous 1970 Individuals at
ports
258,000 travellers
Annual Survey of Hours
and Earnings (formerly
New Earnings Survey)
Annual 1998 Employees 223,000
records
Annual Business
Inquiry (ABI)
Annual 1998 Company
directors
78,000 businesses
Labour Force Survey Quarterly 1979 Adults 60,000
addresses British Crime Survey Annual 1982 Adults 50,000 adults
Health Survey for
England
Continuous 1991 Adults and
children
16,000 adults
General Household
Survey
Continuous 1971 Adults 13,300 adults
Expenditure and Food
Survey (formerly the
Family Expenditure
Survey and the National
Food Survey)
Continuous 1957 Household 7900 addresses
British Social Attitudes
Survey
Annual 1983 Adults 3600 interviews
British Household Panel
Survey
Annual 1991 Adults 5500
households
English Housing Survey Annual 1993 Household 17,000
households Various sources in 2012.
Table
3.3
Chapter 3 Secondary data
79
chapters, etc. Because of the high budget, there is an assurance that almost 5000 researchers are actively producing new knowledge at any one time. For market researchers, this provides
a massive amount of free information of high quality. Published journal articles from these projects and many more sponsors off er a great deal of information. Numerous databases contain journal articles.
Various bodies act as academic researchers. An example is the Pew Research Center, which describes itself as an independent opinion research group, serving as a forum for ideas on the media and public policy. It off ers this information resource to political leaders, journalists, and scholars. All of the current survey results are made available free of charge at its website (http:// www.people-press.org).
Industry sources
The commercial world has a regular output of trade periodicals. These can be identifi ed in BRAD (see http://www.bradinsight.com). It is useful to identify fi ve main specialist titles per sector. Once these are consulted, information will fl ow about that sector.
Company data are used for competitor analysis, identifying suppliers, or building profi les
of potential customers. Much free data is available for UK fi rms from Companies House (http://www.companieshouse.gov.uk). Details available include: location addresses, date of incorporation, country of origin, status, nature of business, activities overseas. Another source
of fi nancial data is FT Interactive Data (http://www.interactivedata.com).
Annual reports are a legal requirement: the International Reporting Standards 2005 (IFRS) requires all listed European companies to report using a single set of accounting standards. This will allow investors to compare the performance of companies across the world. Also, acquired intangible assets must appear on balance sheets. This includes brands, Internet domain names, licensing agreements, trademarks, goodwill, and even customer list details. It
is apparent that for the desk researcher, the annual report, easily obtained directly from the company, is an extremely valuable source.
All US public companies are required to fi le registration statements, periodic reports, and other forms electronically, and anyone can access and download this information for free from the US Securities and Exchange Commission at its website (at http://www.sec.gov).
Directories
Directories provide details of companies that supply or buy. They provide size in terms
of the number of employees or turnover, and areas of activity. In using them, the user must become acquainted with several classifi cations that are used for businesses. These include the Standard Industrial Classifi cation (SIC) and Statistical Classifi cation of Economic Activities in the European Community (NACE). Very important directories come from Kompass (http:// www.kompass.com) and from Dun & Bradstreet, such as Who Owns Whom and Key British
Enterprises (http://www.dnb.co.uk). Directories that look at specifi c sectors can be located
through guides such as Current British Directories (http://www.cbdresearch.com).
Telephone directories can inform up to a certain point. Every country has a Yellow Pages and White Pages (see http://www.infobel.com for links to every directory in the world).
Part 2 Data collection
80
Research in focus