[from Market Trends - A Research Company, at http://www.markettrends.com/p0000045.htm]

The following advice is written from the point of view of a client – a perspective all proposal writers should try to imagine. In this case we are dealing with Market Research – but these criteria are suitable for any survey-based research proposal.

Once you've decided to go outside to a professional research organization, you may wish to review proposals from several firms. A good research proposal submitted to you should cover, in detail, a number of elements; if specific enough, it can serve as the body of the Agreement for Services. If you're writing a request for proposal (RFP) for research, the more elements you request be covered in a proposal, the more likely you'll be able to compare apples and apples when reviewing proposals. As a minimum, the following elements should be included:

RESEARCH PURPOSE/OBJECTIVES
METHODOLOGY
SAMPLE SIZE
TARGET RESPONDENTS/QUOTA GROUPS
RELEVANT UNIVERSE OF RESPONDENTS
ANALYSIS AND REPORTING
TIME FRAME
CONFIDENTIALITY
PROJECT TEAM EXPERIENCE
REFERENCES
SOME SPECIFICS FOR QUANTITATIVE RESEARCH

Your RFP should provide some background information, as well as the decisions to be made as a result of the research. The proposals submitted to you should reflect an understanding of these needs.

METHODOLOGY

Telephone surveys vs. personal interviews vs. mail surveys vs. focus groups.

In some cases, a quantitative study might follow from focus group research.

The number of completed interviews/surveys in each area.

Example: 100 completed telephone interviews in each of the three market areas.

Who qualifies to complete a survey (such as the executive most responsible for purchasing decisions)? Also, should quotas be established to ensure enough interviews are conducted with various subgroups? Quotas might be established by region, household size, purchasing patterns, customer status (such as past, present, prospective), public vs. private sector, etc.

RELEVANT UNIVERSE OF RESPONDENTS

Invariably, when surveying customers and prospective customers, you must obtain a list of relevant firms or people from which to sample. The lists may be derived from professional association directories, area phone directories, lists arranged by SIC code, purchased mailing lists, lists already assembled by you, etc. Specify in the RFP what you already have available; if you do have lists, how many people or organizations are on the list, and do they have working phone numbers and current addresses? Proposals you receive should detail what lists will be used and any associated costs.

A list will have to be significantly larger than the intended sample size given refusals to participate, vacation and travel schedules, etc. For example, in order to complete a sample size of 100 quickly, you might need a list of 400 to 500 or more names.

ANALYSIS AND REPORTING

How will the information be analyzed, summarized, and reported? Do you or does your organization have any preferred means of reporting formats? Should graphics be provided to facilitate understanding? Should reports contain recommendations based on the findings? Should the research supplier be available to make presentations to interested parties in your organization?

TIME FRAME

By what date will you require the findings of the research? Will the research supplier require four weeks or four months to complete the work? Will top-line findings be available prior to detailed analysis and reporting?

Can the research supplier guarantee that all discussions, materials, and findings will be held in confidence? This is especially critical in cases of new product feasibility studies, particularly in the industrial sector with long product development/introduction cycles.

PROJECT TEAM EXPERIENCE

Request information on who specifically will be involved in your project. What relevant experience and credentials do they have?

REFERENCES

Relevant references provide the most valuable input in evaluating any professional service, including market research firms. Request references not only for the firm, but for the individuals to be working on your project. Ask the references about their satisfaction with both the working relationship, and with the final product.

* * *

If a research study under consideration is quantitative, you may want proposals to detail specific assumptions made that lead to the time frame and budget estimates (time implications are as important for you as money to a client!). Again, this will simplify your comparison of proposals. The following items are especially relevant to telephone surveys:

• Data Collection Facilities/Personnel
• Completed Surveys Per Interviewer Hour
• Pretesting
• Coding/Content Analysis
• Data Processing

-- Data Collection Facilities/Personnel

Will the telephone interviews be conducted from an in-house facility, or subcontracted? This has implications for quality control measures and scheduling. What related experience do the interviewers have?

The data collection arm of a research firm, in simple terms, is an information factory. The costs of data collection depend largely on the number of surveys that can be completed in a labor hour. A short, executive level survey with accurate lists might be completed at a rate of 1.0 to 1.2 completed surveys per hour (CPH). Knowing the proposer's assumptions will again allow for "apples to apples" comparisons.

-- Pretesting

Will the draft questionnaire be pretested to evaluate clarity, question wording, and its ability to generate relevant information? Will you be given the opportunity to monitor these interviews personally?

Coding is the process of reviewing open-ended responses to questions, categorizing them, and assigning them code numbers for computerization. The budgeted number of open-ended questions will explain the coding budget in the proposal. Also, how many open-ended responses will be typed and included in the report? This may involve extra typing costs, but will provide dramatic insight into the respondents' state of mind.

You may wish to request an example of raw data output to evaluate its readability and usefulness to management. Depending on the proposer's computer system, the output may be highly useful in its raw form, or it may be esoteric, requiring extensive analysis. How many breakouts (cross-tabulations) of the data are associated with the budget?