Ricci Street < Port 80 < Lighthouse < Internet Searching || search | sitemap | help plaza | theater | bistro

|

Research other pages Database Searches (The Deep Web) Reference Desk (info hubs) | News Desk (traditional media sites) Global News Desk Beyond the Search Engines | The Future of Searching search tips this page surveys | online surveys secondary information | customer characteristics

Who are our customers?
Who aren't our customers?
How do both groups behave online?

best bet

Webmonkey's Market Research on the Web
by Joshua Grossnickle and Oliver Raskin

primary information

expensive, time-consuming, valid, and relevant

Many companies specialize in this type of research. Much of it involves qualitative sampling and statistical analysis. By the time you get it, of course, it's secondary research.

Intelliquest has a series of whitepapers in PDF format that will give you an idea of the in-depth knowledge about brand, product, media, and channel that you can get from industrial-strength market research. They also have an erratically published newsletter called OnTrack that condenses some of their research. Their subsidiary Zona Research is also very useful.

Research Instruments include mechanical or electronic devices, although the survey questionnaire is the most common instrument for marketing research.

Observations of actual behavior may not help understand why people act as they do. Ideally, you're a fly on the wall in the mind of your customer while he or she decides to buy. It's not ideal, but you an learn a lot from watching people use a web site. Keith Instone's Usable Web site has a couple dozen links to user testing information that will get you started.

Another form of observational primary research is gleaned from the logging built into online activity and from the measurability inherent to technical information. At its most basic level the Internet is just a string of ones and zeros and computers are terrific counting machines. Learn more about these metrics.

Experiments establish cause and effect relationships. However, controlling for variables such as the 4 P's is difficult in real-world situations. We're reluctant to experiment on people. When we add avatars and telepresence, will we remain that reluctant?

Not everything that can be counted counts, and not everything that counts can be counted.
--Albert Einstein

Surveys

They help people describe reasons for their behavior, but surveys can be plagued by problems in completion. They presume self-knowledge and a willingness to reveal it.

Your sampling plan must address who (sampling unit), how many (sample size), and how to choose decisions of drawing a sample (probability or nonprobability).

Mapping Customer Attitudes
by Cliff Allen
ClickZ, October 17, 2000

Marketers have a lot of data at their disposal. They've got demographic data, purchase history data, and web tracking data. And that's just for starters. But many marketers could benefit from a map of customer attitudes. Using a perceptual map, marketers can create more compelling ads, refine media selection, and tailor content on web sites and in email messages.

mail questionnaires

They provide excellent control over interviewer effects. Since no interviewer is present, no interpersonal influences are possible. Mail questionnaires may elicit more honest and in-depth information, but the inability to distinguish respondents from non-respondents makes inference to a general population difficult.

email questionnaires

They suffer from the same problems. In comparison to snail mail, they're quicker, cheaper, easier to change, easier to process (no manual data entry), and can be linked to other online marketing information.

telephone interviews

Thy enable sample control and fast data collection. Many marketing contexts, such as political campaigns, need almost overnight information on the effectiveness of the latest promotions. Telephones reach diverse geographic markets, and can be linked to computers for easy data analysis. Problems with respondent cooperation may become increasingly important over time.

online interviews

They haven't quite integrated telephony yet. Until we gain an audio voice to supplement our written voice, online interviewing can't replace telephone interviewing. When it does, this form of data collection will retain all of its advantages without accruing any disadvantages. A combined audio and video connection will merge this form of data collection with the personal interview.

personal interviews

They are very flexible. They can provide a rich, deep volume of subjective, anecdotal, and impressionistic data for the researcher. Also, personal interviewers can follow up unexpected or unusual responses that other collection methods are not prepared to handle. Using groups in personal interviews can also reveal social influences that may be important in consumer decision making. However, we haven't begun to think about the potentials and challenges of personal interviewing in cyberspace.

The computer screen cannot replace the physical presence of two human beings in the same room. However, to what extent does the computer screen resemble what the theater folks call the fourth wall? As we gain the bandwidth to emerge as actors, we will be represented by avatars (digital puppets) and be able to assume a range of personae not possible in a room. In a 3D virtual reality, the market researcher's avatar will be interviewing the customer's avatar. Our psychiatric hospitals may have to plan on expanding.

Steps in the Marketing Research Process Including Defining the Problem and Research Objectives

Phase	Step
Figure out what to research	1 Establish Need	Is there a real need? Cost/benefit analysis
	2 Define Problem	Gap analysis
	3 Determine Research Objectives	What the researchers must do
Design the mechanics of the research	4 Determine Design	Exploratory? Descriptive? Causal?
	5 Information Source	Primary: information collected specifically for the problem at hand Secondary: information already collected
	6 Data Collection Method	Have a person ask questions Use computer assisted or direct questioning Allow respondents to answer questions themselves without computer assistance
	7 Questionnaire Design	Questionnaire must be worded objectively, clearly, and without bias Observation forms are used to record observed behavior
	8 Sample Size and Plan	Who? How many? How chosen?
Gather data from respondents	9 Collect Data	Data collection errors
Generate findings and interpret them	10 Analyze Data
	11 Write and Present Report	Closes the loop; finishes the project

Step 2: Define the Problem is the most important. If the problem is incorrectly defined, all else is wasted effort.

Working with the client to define the problem

Two sources of problems: Gap analysis

a gap between what was supposed to happen and what did happen, i.e., failure to meet an objective. That's a problem.

a gap between what did happen and what could have happened. That's an opportunity.

Both problems and opportunities can profit from marketing research.

Recognizing the Problem

The problem gaps are easier to recognize, often because the boss is unhappy. The opportunity gaps are harder to recognize.

Symptoms or Problems?

“We have a problem … we are losing money.”

Managers must be careful to avoid confusing symptoms with problems.

Symptoms are changes in the level of some key monitor that measures the achievement of an objective.
The role of the symptom is to alert management to a problem; there is a gap between what should be happening and what is happening.

The standard performance measures include:
Sales level
Market share
Profit
ROI (return on investment)
Satisfaction index
Complaints

The Role of the Researcher in Problem Definition

Researchers should ensure managers are defining the problem correctly. This is particularly true when the manager has already defined the problem in very specific terms.
Researchers sometimes take additional investigations, known as a “situation analysis,” to ensure the problem is adequately defined.

Impediments to Problem Definition

Failure to change behavior for problem definition situations
Managers deal with outside suppliers efficiently with little interaction. Marketing research requires a great deal of interaction and communication.
There are differences between managers’ and researchers’ backgrounds.
Traditionally researchers were technicians and managers were trained in general decision making.
Today managers are much more aware of technical software such as using your XL Data Analyst.

Assess the Background and the Manager’s Situation

The researcher should first understand the industry, the competitors, and the company.
The researcher must understand the manager’s unique situation…
Does the manager have a particular objective?
What constraints is the manager operating under?

Clarify the Symptoms of the Problem

The researcher must understand the control system and determine what symptoms are being identified by the system.

Pinpoint Suspected Causes of the Symptom

There’s always some cause or causes for a change.
It is important to determine all possible causes.
Researchers should narrow possible causes to a small set of probable causes.

Specify Possible Solutions that May Alleviate the Symptoms

Possible solutions include any marketing action that the marketing manager thinks may solve the problem, such as price changes, product modification, etc.

Speculate on Anticipated Consequences of the Solutions

“What if” questions should be made regarding possible consequences of each marketing action being considered.

Identify the Manager’s Assumptions About the Consequences of the Solutions

Assumptions are assertions that certain conditions exist or certain reactions will take place if considered solutions are implemented.
If the manager is completely certain of assumptions there is no need for research.
For those uncertain assumptions, research will eliminate a manager’s uncertainty and therefore aid in decision making.

The Marketing Research Proposal

State the problem.

List the specific research objectives.

Detail the research method.

Discuss timetable and budget.

Primary Research:  Define the Problem

Two sources of problems:

A problem exists when a gap exists between what was supposed to happen and what did happen, i.e., failure to meet an objective.

An opportunity occurs when there is a gap between what did happen and what could have happened…called an opportunity.

Recognizing the Problem

To recognize a problem managers must be knowledgeable of objectives and actual performance.
To be aware of opportunities, managers must have a process for monitoring opportunities, such as opportunity identification.

Symptoms or Problems?

“We have a problem … we are losing money”

Managers must be careful to avoid confusing symptoms with problems.

Symptoms are changes in the level of some key monitor that measures the achievement of an objective.
The role of the symptom is to alert management to a problem; there is a gap between what should be happening and what is happening.

The standard performance measures include:
Sales level
Market share
Profit
ROI (return on investment)
Satisfaction index
Complaints

The Role of the Researcher in Problem Definition

Researchers should ensure managers are defining the problem correctly.
This is particularly true when the manager has already defined the problem in very specific terms.
Researchers sometimes take additional investigations, known as a “situation analysis,” to ensure the problem is adequately defined.

Pinpoint Suspected Causes of the Symptom

There’s always some cause or causes for a change.
It is important to determine all possible causes.
Researchers should narrow possible causes to a small set of probable causes.

Specify Possible Solutions that May Alleviate the Symptoms

Possible solutions include any marketing action that the marketing manager thinks may solve the problem, such as price changes, product modification, etc.

Examples

Identifying marketing opportunities and problems

Market-demand determination
Market segments identification

Generating, refining, and evaluating potential marketing actions

Proposed marketing-mix evaluation testing
New-product prototype testing
Web site usability
Advertising pretesting

Online Surveys

what they are

A quantitative research method that provides information that can be measured.

Using The Internet Responsibly For Market Research
press release, Penn State

It is possible to use the Internet to perform a wide range of useful and accurate, surveys without spamming, without being intrusive, and in a manner that facilitates communication generally.

why use them

With an online survey you eliminate the delays for printing, mailing and data keying.

Results are received in hours not days, and they are ready in a database for analysis.

Printing, keying, mailing and interviewing costs are eliminated.

Every aspect of the survey can change midstream.

Improve a question's wording, change the wording, whatever... It can be done quickly and easily.

Color, images, stylized fonts and graphics and make the experience much more pleasant. Streaming audio and video is available to further enhance the survey.

Skips, branching, data verification, dynamic or randomized question generation etc... can be use to enhance the user experience and protect the integrity of your data.

why not use them

Some drawbacks to using this feature is that it depends on the information you are looking for and the product or service you are trying to get information for. If you were looking for a survey about how the latest Ford pickup drove, then you couldn't have an online survey pop up every time a person clicked onto the Ford web site. It could be an optional survey if they did have one already but the chances of that are slim.

Web-based surveys: Changing the Survey Process
by Holly Gunn
First Monday, December 2002

The challenges and pitfalls of Web surveying, revealing what we've suspected all along: it's different online. Survey research is a very well studied science, but the Web adds several new factors that govern how people respond.

where to find them

GoGlobal

Custom web based survey development for large corporations and full service market research companies.

Clickin

One of the first market research companies to begin collecting data over the Internet. Our CyberleagueTM panel has grown to include over 50,000 members, each one affecting the way companies do business by sharing their opinions.

SurveySite

Their Pop-Up Software counts how many people come to a Web site and selects visitors at a pre-determined interval (for example, every 100th visitor). When the "Nth" visitor browses the site, a small JavaScript window pops into view asking the user if they will complete a short online survey after they finish browsing the site.

Sweepstakes Online

Internet promotions company specializing in sweepstakes development and promotion for companies and their Web sites.

NetVotes - "Build Your Own Online Voting Booth"

examples of sites that use them

Hewlett Packard - helped choose a new name for its scanning technology. For this research project, they created an online survey that presented over 4,000 respondents with five names/images of the new scanning technology. Respondents then chose the name/image that best communicated the objectives of the technology.

SurveySite used question/image "randomization" in this study. That is, in order to prevent bias caused by order effects, they automatically rotated the order of the names/images that were presented to the respondents.

AT&T - GoGlobal Technologies, Inc. developed an online survey.

Dawsons

ClickMusic

secondary information

cheap, quick, of questionable validity and relevance

Reference Desk: Learn more about "hubs" or "centers" or "document depositories" including demographics and information about competitors.

News Desk: Learn more about timely news sources on the Web.

Soothsayers and other fortune tellers

"By 2005, online research firms predict, 110% of the U.S. population will ... "

110%??

It's that silly if not always so obvious. Predictions and projections are based on assumptions. If the assumptions don't come true, the predictions may not, either. It's more useful to talk about the assumptions than to parade the silly projections that result from applying them.

For starters, what's the delta (the change rate)? According to Metcalfe's Law, the value of a network grows exponentially as the number of nodes grows arithmetically. That is, doubling the nodes provides four times the value.

Will the diffusion of the Internet and Internet-enabled change follow a non-uniform formula with a value of delta that is substantially greater than one? If so, what? Two? (gasp!) One and a half?

Don't under any circumstances make any important decisions based on some soothsayer's projections. Don't be misled by the hard-number look, the seeming objectivity of a chart or graph. That means you should examine the y-axis of every chart and graph you see. As soon as it gets into the future, ignore it. Yes, that means you will ignore the right-hand part of almost every chart and graph you see.

What's the psychology here?

Proverbs -- Concise, familiar, unpretentious expressions believed to contain truth or wisdom. Guidance, often contradictory, to explain the unexplainable.

Memes -- Ideas, regardless of validity, that multiply like a virus by spreading through the population and grasping hold of the imagination.

Learn more about psychometrics. Beware!

Worldwide, upwards of a hundred academic journals around some marketing topic are published in English. Many of them come from conglomerates like MCB University Press in the U.K. and Elsevier Press's Science Direct in the Netherlands. Both make the full text of many journals available online -- but only to those subscribing to the print versions. They and others also make articles' abstracts available online.

Here I've listed several journals that have at least some full texts of articles online for free. If you know of others, please let me know.

Academy of Marketing Science Review (formerly the Journal of Consumer and Market Research)
Journal of Empirical Generalisations in Marketing Science
Marketing Research On-Line

Where to do Research -- Major Think Tanks & Policy Groups

management models

Perhaps not quite as misleading as projections are models used by management to predict the future. Use them with great care.

You're on firmer ground when looking at brand-new models based on current primary research.

Consumer Behavior Model (B2C)

Consumer Purchasing Decision-Making Model

Organizational Buyer Behavior Model (B2B)

These models work best as a descriptive tool. They can help you recognize threats and opportunities. They don't explain much. And I'd be very careful about using them to predicting the future, to plan strategic responses to threats and opportunities, and especially to provide guidance for managing change (accelerating or retarding it). The future will fool you.

diffusion of innovation

Customer Characteristics

learning styles

"Learning" to shop in a physical store is something we do at an early age. Learning to shop in an online store puts the computer between buyer and seller. Smart online sellers will take learning styles into account when designing marketing webs.

Student Learning Styles

Personal Learning Style Inventory

Learning Modalities, Styles and Strategies

Index of Learning Styles Questionnaire

usability

How easy is a web site to use? That question begs another: for whom to use? The CEO? The marketing director? Or your target buyer?

The Usable Web

Jakob Nielsen's UseIt.com

profile targeting

The Versatility of Profile Targeting
by Tom Hespos

Ever been to a web site that personalizes its offerings for users? Typically, such a site offers to target relevant information to users who tell the site a little bit about themselves. Not surprisingly, this information facilitates the delivery of targeted advertising as well as content. Executed properly, profile targeting can go a long way toward providing a win-win-win situation for users, sites and advertisers.