Marketing Research Blog

The Great Proposal Wasteland

How would one develop a branding model using text data and deep learning methods?

Corporations that use outside research agencies to conduct research tend to be wedded to a proposal-bidding process. It typically goes something like this. Sally, the consumer insights director at Big Corporation talks with her internal clients and develops a Request For Proposal (RFP). Let’s say it’s a Market Structure study. A carefully designed, well-written RFP would take Sally several days to put together. This RFP is then sent to five or ten (or more) research firms for proposals. Each research agency puts in two or three days of effort in crafting a “winning” proposal. Big Corporation, it is thought, will get the best thinking from these different research firms and lower prices because of the competition among the firms. The proposals come in, and Sally spends two or three days reading and analyzing the proposals and re-contacting the research firms for clarifications and additional details. Sally then presents the proposals to her internal clients and discusses with them which research firm should be selected.

At this point, if it’s a large project, Sally’s purchasing department might become involved to hammer down prices among the “finalist” proposals. After much back and forth and consultations with internal clients, Purchasing, and Legal, and more pricing discussions with research agencies, a final choice is made, and the research is awarded to one of the supplicants. The losing research firms request follow-up time with Sally to better understand why they lost, so there goes another half day of Sally’s time.

Sally and Purchasing choose a new research agency because of its very nice, highly visual proposal at a highly attractive price. However, it is a new research supplier, so she spends another day or two of her time educating the new firm, re-writing its questionnaires, and explaining her industry and her company to the new firm. It’s now 45 days since Sally started the process, but the project has not yet actually launched. Sally explains to the new research firm that speed is of the essence and insists that the new research firm must squeeze time out of its schedule and complete the new project in 15 days instead of the proposed 26 days.

The total elapsed time from Sally’s initial internal discussions to project completion is 60 days (45 days plus 15 days). The cost for Sally, the internal clients, Purchasing, and Legal to write the RFP, get bids, make a choice, and supervise the project is about a month of staff time (approximate internal cost to Big Corporation is $15,000). Big Corporation paid $48,750 for the project and saved $2,850 through the proposal process by choosing the lowest-priced bidder. Big Corporation also saved 11 days in executing the project. Big Corporation is happy.

Not every corporation, however, pursues the proposal-bidding model followed by Big Corporation. Just down the street in the same industry is Nimble Corporation. Nimble completed a Market Structure study similar to Big Corporation’s project, with a total lapsed time of 30 days, instead of 60. Nimble paid $45,000 for the project, quite a bit less than Big Corporation paid. The total cost of project administration (supplier selection and supplier management) was $2,500 (not $15,000). Nimble actually achieved higher quality on the project because its research agency had previous experience working with Nimble and completed the project on a regular schedule (30 days instead of the 15-day rush schedule).

How in the world did Nimble Corporation save money, save time, get higher quality results, and spend almost no time or money administering the project? Betty, the research director at Nimble Corporation, had talked her boss into allowing her to choose a very limited number of research partners to do all of Nimble’s research. She asked each research partner to provide standard costs for different types of studies, assuming that Nimble would give the partner a substantial amount of business each year.

So, when the need for the Market Structure study appeared, she chose the partner research firm best qualified to conduct the Market Structure study. She knew what the cost would be in advance. When she talked to her internal clients about the project, her partner research agency participated in the discussions. There was no need for a RFP or proposal, but she did ask the partner research firm to prepare a statement of work (SOW) and to confirm the price. The project was designed and launched in 8 days. Topline results were available 22 days after the launch.

Let’s summarize the schedules and cost for these two different models:

• Proposal/RFP/Research Agency Selection Process
  • RFP model = 45 days
  • Standard-cost model = < 1 day
• Project Design and Execution
  • RFP model = 15 days (a rush schedule)
  • Standard-cost model = 26 days (a normal schedule)
• Total Elapsed Time From Start to Completion of Project
  • RFP model = 60 days
  • Standard-cost model = 27 days
• Total Cost of Project (Internal Costs Plus Research Project Costs)
  • RFP model = $63,750
  • Standard-cost model = $47,500

These two approaches (proposal-bidding model versus partner-standard-costs model) present stark contrasts in efficiency and time, yet the proposal-bidding model is far more common in practice, at least in the U.S. With the growing role of purchasing departments in the vendor-selection process, even more time and effort is being invested into the proposal-bidding model by most large corporations (after all, Purchasing must justify its existence by the money saved during the bidding process).

The RFP model, however, can have a role to play. The RFP model can be used to screen candidate research agencies (and their standard costs) to select those firms that will become a corporation’s research partners, but there is a caveat. If research agencies are chosen based on a RFP “dog and pony show,” the corporation will end up with research agencies who are extremely good and “dog and pony” shows, but maybe not so good at the day-to-day tedium of turning out solid, consistent, accurate research.

An even better way to select research partners is to give the likely candidates a series of research projects to see how they perform “under battle conditions.” Those that consistently perform at a high level can then be asked to provide standard costs for an array of different study types, assuming a substantial amount of business per year. This is the safest approach.

It’s generally best for a corporation to choose a small set of research agencies, based on performance history and standard costs—and then use those research partners over and over again so that the research firms become experts on the corporation’s brands and categories. By replacing the proposal-bidding process with standard-cost partners, project execution times are reduced, costs are reduced, risks are minimized, and the quality of the research improves. These battle-tested research partners will make your life easier and your corporation more profitable.

The RFP model is good for choosing long-term suppliers of a wide range of materials, components, and equipment, but it’s a poor model for choosing research agencies.

Jerry W. Thomas

Chief Executive Officer

Jerry founded Decision Analyst in September 1978. The firm has grown over the years and is now one of the largest privately held, employee-owned research agencies in North America. The firm prides itself on mastery of advanced analytics, predictive modeling, and choice modeling to optimize marketing decisions, and is deeply involved in the development of leading-edge analytic software. Jerry plays a key role in the development of Decision Analyst’s proprietary research services and related mathematical models.

Jerry graduated from the University of Texas at Arlington, earned his MBA at the University of Texas at Austin, and studied graduate economics at SMU.

• 
•