Articles

A New Era in Customer Driven Marketing - Part Two

From Casual Contacts to Relationships...

Despite the many benefits, Internet research shares a common challenge with all forms of traditional survey research - how to solicit the cooperation of survey respondents? In the world of telephone research cooperation has historically been based on appeals to the fundamental "civility" of potential respondents. The social standards that were engrained into most citizens and consumers from an early age prevented contacts from being rude to an interviewer with a pleasant voice and polite manners. But today a combination of screening technology that provides a means of avoidance and the abuse of telephone interviewing by telemarketing organizations, which provides the desire for avoidance have made consumers far more likely to respond negatively to even the most polite stranger contacting them by phone.

In the world of Internet surveys the "link" between respondent and surveyor is potentially even more tenuous. With little more than the click of a mouse, unwanted surveys can be dispatched to the trash bin. Even worse the sender can be permanently blocked so that respondents never have to deal with them again. Resolving the issue of respondent cooperation is the most important challenge facing the new and exciting world of Internet research.

The "good news" is that unlike traditional telephone surveys, where almost all of the data collection dollars are invested in hiring interviewers, renting facilities and paying telephone charges; the minimal cost of "distribution" under the Internet model allows clients and survey companies to invest greater thought and resources in how best to secure respondent cooperation. For most categories it makes sense to repeatedly seek views and input from respondents who are willing to participate This is the central concept of a panel and panels are the key to unleashing the power of Internet research.

For some academics the idea of multiple surveys with the same respondent raises the usual chorus of quibbles about whether survey measurements will be tainted by "professional respondents". But in the world of real research, panels actually improve the quality of measures for most applications. That's because individual surveys are usually shorter since it's not necessary to continue re-asking demographic and other background questions. And shorter surveys don't produce the annoyance and irritation associated with marathon survey sessions. Panels also allow for measures which track changes in attitude, perception or behaviour over time thus providing a much more accurate measure than is obtained by comparing the results of successive surveys.

Think of a panel as a type of consumer advisory board made up of current and/or potential customers. An Internet panel can be as small or as large as is necessary with total size only having a relatively marginal impact on cost. Indeed the major driver of panel size is the frequency and scope of panel usage since it's important to involve consumers once their cooperation has been assured.

PART THREE

Reprinted with permission from Angus Reid. Read the full article at http://angus-reid.com.

Posted on October 28, 2004
Permalink

A New Era in Customer Driven Marketing - Part One

Internet Research... Faster, Cheaper, Better

It's been almost forty years since I started my research career. In those four decades I've witnessed the birth and maturation of two eras. Each period saw market intelligence and consumer feedback become successively cheaper, more accessible and more powerful as tools in decision making. From the mid 1970s onwards, the telephone rapidly gained penetration across North America (and later in most other developed countries) as the principle tool for collecting information from consumers and citizens. From the mid 1980s onwards, the personal computer provided market research agencies and their clients with a cheap and easy-to-use means of analyzing the mountains of data that are produced in the field of survey research.

Today we are in the midst of a third revolution in the field of market intelligence - one that combines elements of both telecommunications and personal computers. In contrast to previous times, this new era involves the personal computers on the desks of consumers and citizens and the phone lines (and cable connections) that link them together. I'm referring of course to the Internet, a technology that is quickly transforming the marketing research world and providing end clients with unprecedented choices and options.

Internet Research: Better, Cheaper, Faster
The $7 billion (US) market research industry in North America is undergoing a fundamental transformation as the Internet is rapidly replacing the telephone as the principle means of collecting information on the experiences, attitudes and perceptions of consumers. Perhaps it's only a coincidence, but at the same time that telephone survey refusal rates consistently exceed 80% , the number of North Americans accessible via the Internet is rapidly approaching 80%.

The Internet "revolution" as applied to marketing research is about more than simply replacing the means by which surveys are conducted. It represents a fundamental shift in the entire business model of marketing research and business intelligence. The old pricing system, in which the cost of research is almost totally dependent on the number of completed telephone surveys, is being replaced by a pricing system in which the sample size becomes increasingly irrelevant as a deciding factor in the cost of doing research. The big winners in this shift are end clients who can obtain far more detailed, accurate and timely information at a fraction of the price of conventional research.

Under the traditional research model feedback from consumers is collected via telephone interviews (or more rarely face to face interviews) where the cost per interview ranges from $20 to $100 (or more) depending on the nature of the target audience (e.g. business travelers represent a small segment of the entire population and are therefore more difficult and costly to interview than, say, pet owners).

This cost model has historically presented many clients with less than adequate information for some of their most pressing business issues. For example, a multi location retailer with 100 shops across the country would ideally wish to have consumer feedback which allows for an analysis of "store level" data. The problem is that a minimum sample size of 300 customers per location is required for this type of analysis thereby yielding the need to conduct 30,000 surveys to produce data for this type of analysis. In the world of telephone surveys this type of survey would cost between $500,000 and $1 million. For most retailers this cost is well beyond their means.

Internet research removes much of this cost from the research equation. The charges for sending and receiving surveys are negligible. And unlike conventional research where a survey of 30,000 consumers would take a few months to complete, a survey of this size in the Internet space can be deployed and completed in several days.

Large Internet surveys, unlike their telephone counterparts, are completed by respondents on their own schedule and not one imposed by the telephone interviewer. Rather than being interrupted by a telephone interviewer in the middle of dinner, respondents can choose to complete surveys on their own timeframe - and often in an environment which allows for more considered thought on the subject matters being investigated.

Finally Internet based research permits those conducting studies to present visually engaging stimuli - pictures and video - rather than simply the spoken word. For example, it's far more realistic to have respondents in a survey of fast food companies give their impression based on a picture of a McDonald's location than to respond to the word "McDonald's'' spoken over the telephone line. After all it's the visual that consumers actually interact with in the real world.

These benefits are part of the reason that Internet research is mushrooming across the developed world. In the US it's estimated that 20% of all research is now conducted over the Internet - up from less than 5% five years ago. The major Internet research suppliers such as Harris Interactive, MarketTools, and Vision Critical report skyrocketing revenues while those in parts of the industry still with their heads in the sand, plying telephone surveys, are flat to down.

PART TWO

Reprinted with permission from Angus Reid. Read the full article at http://angus-reid.com.

Posted on September 23, 2004
Permalink

The Best & the Brightest - Tamara Pritchard

We couldn't keep this place running without them, so we thought we should share them with you. This bi-monthly feature is about the people behind the scenes that keep this ship afloat.

Name:	Tamara Pritchard
Title:	Senior Project Manager
Age:	405 full moons
Height:	37200 pixels, in high resolution
Nickname as a child:	TJ
Likes:	Ice cream, Mountain biking, and the colour orange
Dislikes:	Poor grammar, doing laundry, and eggplant

How long have you been at Vision Critical?
I was here when they opened the doors, and they still can't get rid of me. This is the fourth move I've survived, and hopefully we'll stay here long enough for me to unpack my pen collection.

What role do you play at VC?
Every time we've printed new business cards my title has changed. I'm not sure if that's a good thing or bad, but at least it's always evolving. I think my first card read "everything girl"; we've actually gotten real titles since then, but I do still do a little bit of everything. Currently I'm positioned as a project manager, but my day could consist of everything from writing code, to training clients on our software. I get to mix it up, which keeps me from any possibility of getting bored.

What is your educational background and how does it benefit VC clients?
I have a degree in Environmental Design/Architecture, a certificate in New Media, and several programming courses under my belt. I have such a diverse background, and I've worn so many different hats since the company started that I know at least something about almost every aspect of the company. This benefits clients by allowing me to quickly turn around solutions that fit their needs.

What major improvements in products or service delivery have you seen in your lengthy tenure at Vision Critical?
Ha! The improvements are enormous; we now HAVE products! The amount of growth and increase in focus over just the past year is amazing. We started off doing just about anything, for pretty much anybody, just to pay the bills. Learning new technologies every other week is educational, but not a good way to grow a company. We ARE still here, however, so we must have done something right. Refocusing our energies into the market research/software development fields has really given VC some centre, and helped us prove that we really do know what we're talking about.

This piece was brought to you by: Sunday mornings, the number 35 and the letter P.

Posted on September 23, 2004
Permalink

Determining Panel Size

When bigger isn't better...

One of the most common misconceptions in the panel building business is that a panel must be as big as possible. When initially asked, most clients will quote a desired panel size that is easily twice as large as it needs to be. Every client and circumstance is unique, but by examining a few key factors the optimal panel size, can be determined

Frequency of Contact
The first, and most important, factor you should consider is frequency of contact. Ten years of online research has taught us that while panelists do not like to be overburdened with research requests, there is also such a thing as too little contact. A high quality panel is about creating a relationship with the panelist and if they are only contacted two or three times a year it is hard to build that relationship. Our experience has shown us that panelists should be invited to participate in research, or contacted in some other way (i.e. newsletters), between 12 and 15 times per year. If you are planning on rolling out 60 studies over the course of the year, you will need to ensure that you have ample panelists available to include only a quarter of your sample in each study. Conversely, if you only have a handful of studies planned you will be able to return to the same sample group for each subsequent study.

Sample Size
The second factor to examine is the acceptable sample size per project, and your anticipated response rate. You will need to consider the level of analysis you would like to do on the results of specific research studies, which will lead to a decision about your acceptable sample size per project. From there, response rates are anticipated and factored in to determine what size of panel you require.

Target Audience
The third factor to explore is the necessity for, and composition of, desired sub-groups. Certain demographics will respond better than others (females over males for instance) and, depending on the recruitment method you use, low-incidence sub-groups can make larger panels necessary.

Refreshing Panelists
Fourth is your willingness to refresh panelist throughout the year by inviting new people to join. This is mainly a factor when longitudinal research is planned. If it is important to survey the exact same people at the beginning of the year as at the end then a larger panel size is required at the recruitment stage.

Sample Source
The fifth and final factor is access to, and cost of, sample for recruitment. This factor is most important where sample for your recruit is expensive. In this case it is a wise decision to start with a smaller panel size in order to determine more accurate response rates and attrition estimates before recruiting a larger panel.

SAMPLE SCENARIOS

Example One- Season Ticket Holder Panel
A sports team wants to build a season ticket-holder panel. They want to conduct four large studies during the year, but also plan to use the panel to poll their biggest fans about team issues as they arise- maybe one per month. Their analytic needs are pretty straight-forward, they need 300 completes per study, and they anticipate a high response rate (55%). They do want to have a number of corporate ticket holders, but can target them directly through the recruitment process. They are willing to boost the panel size throughout the year and access to the sample for the recruitment is free. Considering these factors we would suggest a panel size of 800 to 1,000 people.



Example Two- Customer Satisfaction Panel
A telecommunications company wants to build a panel to monitor the satisfaction of their various customer groups. They have six departments who are interested in using the panel as well as the management team of the company as a whole. There are about 30 research projects scheduled for the year. They do have several variables they will want to use for analysis and therefore feel they will need a minimum of 600 completes per study. They anticipate response rates to be slightly below average. The have 5 key sub-groups that must be represented, though they will be able to target them directly through the recruitment process. They are willing to boost panel size during the year and access to the sample is free. Considering these factors a panel size of 3,000 to 5,000 would be recommended.



Efficient panel size is one of the keys to a healthy and productive panel. Vision Critical's panel experts can guide you through the process of determining the best panel size that is right for you.

Posted on September 23, 2004
Permalink

Online Panels: The Evolution of an ‘Old but New’ Methodology

As consumers continue to migrate online, traditional marketing research methodologies evolve to deliver data in a "results now" business environment. 

The concept of a panel is not new to the research industry. Panels have always consisted of groups of people who have consented to take part in multiple surveys. They have always tracked the various socio-demographic characteristics of these people for the purpose of selecting sub-groups for future studies. Yet the industry can't stop talking about panels in their newest form - Internet panels.

Posted on August 24, 2004
Permalink

« First  <  15 16 17