Lack of clarity is often due to one of the following:

1. Imprecision. The goals are documented and discussed at too high a level—and that vagueness leaves too much room for interpretation during the research design, analysis and reporting phases.

Example: A team agrees on an objective of creating a market segmentation model. Sounds good, right? Not really. There are many ways to do a segmentation model, and many initiatives that segmentation can support. Which ones are relevant? A better objective would be, “To develop a segmentation model that will support near-term sales strategy development,” or “To develop a segmentation model that will help us understand emerging customer opportunities.”  In the first example, the study would likely focus on purchase plans, budget/spend behaviors, brand preferences, and might be primarily quantitative. In the second, the study would likely include a rich discovery phase. Two segmentation studies, different true objectives, and different methods.

2. Unrealistic scope. If a project is defined with too many objectives, the scope becomes too broad to execute well.  In quantitative projects, this often is evidenced by a questionnaire that is so long and onerous, that the resulting data is weak.

In most cases the root cause of an unrealistic scope is conflicting agendas. Sure, everyone on the team may agree that they need to better understand customer needs, so they agree to do a project. But the team members may have their own spin on how they want to look at needs.

Here is an example. Without some discipline, a study about “customer needs” could easily end up with 5 or 6 objectives, such as:

1. Understand relevant product category needs
2. Understand customer service needs
3. Discover customer perceptions of how to address current needs
4. Measure potential value of addressing emerging needs (for pricing implications)
5. Identify which brands have permission to address different needs

Without an agreed upon, and small, set of precise objectives, your project is at serious risk. None of the objectives will be met with excellence. You won’t have enough data specific to any one objective in order to generate any clarity or insights. And chances are, none of the stakeholders will be particularly satisfied, nor will they be likely to have enough conviction about the results to take action. And that is the real disaster.

[Want help planning a market research project? Please check out the Research Rockstar on-site workshop offerings here: LINK]

We see articles all of the time lamenting lack of innovation in market research (see just about any market research-related blog or LinkedIn discussion group). What I have observed is that many client-side market research managers are so swamped with project management that they have little time for the truly strategic work of planning for innovations. I believe we need to raise awareness among corporate executives that the market research department can only innovate if it has the time and resources to do more, much more, than manage projects. My article details my suggestions on how this can be made a reality; you may not agree with my suggestions, and I know many of you won’t, but it’s a start.

So why are market research departments heroes? Because they are on the front line, fighting the battle to do more—yet they are constrained by out of date “marching orders.”

Alert magazine does require MRA membership: Link to Article.

[Do you already subscribe to the Research Rockstar newsletter? If not, sign up here: SIGNUP]

RFPs range widely in terms of their level of sophistication and complexity. If your need is for a simple project, do yourself a favor: keep your RFP simple. It will save you a lot of time when the proposals come back and you actually have to read them.

At minimum, a market research RFP states parameters and requests responses to the basics:

• Statement of objectives
• Target population (B2B, B2C), perhaps with quotas
• Statement of geographic scope
• Sample source: Are you providing it, or is the agency?
• Statement of preference for qualitative or quantitative methodologies, if any
• Deliverables required at project conclusion
• Timeline requirements
• Qualifications required

Consider the above list your minimum RFP content. Of course, you can add many more items, depending on your project’s goals. Here are some more examples of things you can request in an RFP:

• Please provide information about your firm’s relevant industry experience.
• Provide biographies for the key team members you would assign to this project.
• Describe your methodology recommendation and how it will address our objectives.
• Provide details of your quality assurance (QA) practices.
• How will you ensure sample qualifications?
• (For international studies) Please describe your experience collecting data in the target countries.
• Please list the critical success factors and risks you see for this project.
• Please describe the expected timeline for this project, including key milestones.
• Given our objectives, what screening criteria and quotas do you recommend?

Market Research RFP Resources

For more tips on how to write a great RFP: RFP eBook

Free members and VIP members can access actual RFP templates on their member access pages

For on-site help, check out our workshop offerings: workshops

Or check out the book, available in Amazon: Amazon Link

Segmentation models vary from basic to complex, and the approaches to developing and applying them is a topic for an entire book itself. But here are three examples of the very real benefits a segmentation model delivers:

1. Plan product roadmaps. It’s a lot easier to craft a roadmap when you have internal agreement about the profiles—including needs and behaviors—of the most attractive customer segments. If decision makers share a common view of what the most attractive segments are, and their emerging needs, planning new products and product improvements becomes a far more time and resource efficient  process. Not to overlook the obvious benefit that customer-inspired product ideas are more likely to succeed.
2. Embrace niche marketing. Few companies can afford to compete with the big fish in the big pond. Instead, a segmentation model shows you how to be a big fish in a little pond—or several little ponds.  A segmentation project can be designed to identify, profile and select profitable niches.
3. Setting your budget. Why let a marketing budget be set by something arbitrary, such as percent of sales? A segmentation model can provide a more meaningful, analytical approach. Given the size of attractive segments, and objectives for each segment (awareness/preference/market share, etc.), a strategic budget aligned with actual goals can be built. When you take this approach for the first time, you will likely find yourself investing less in some marketing programs, and more in others.

These are just 3 examples—there are many more. Indeed, there are many ways that a segmentation model can help your organization make money and save money. That said, actually creating a segmentation model is fairly complex, and success requires a very thoughtful planning process. Before you start, I strongly urge you to get assistance from someone who has experience with these types of studies—either a  consultant or a market research agency.

For those interested in learning more about market segmentation, I have three offers for  you:

First, please check out my new white paper, “9 Ways Market Segmentation Improves Business Success.” GET WHITE PAPER

Second, check out the preview of the new, 53-minute online introduction to market segmentation, “Divide & Conquer: Practical Steps to Market Segmentation Success.” CLASS PREVIEW [Research Rockstar VIP members: please note that this class is currently available in the members’ area—just log-in, click, and view!]

Third, for those interested in on-site training, I now have a more in-depth, 2-hour version of the online class, which includes interactive exercises for teams planning to create a market segmentation model. For a class outline, email me at KKorostoff@ResearchRockstar.com. Thanks!

But working with any type of consulting agency requires some level of project management on your part.  There are many risks when conducting market research; know them to mitigate them.

There are several points in a market research project where you, as the client-side manager, need to be  particularly cautious. Here are two of the big ones:

1. Risk: Ignoring sample quality. If your project is one where the agency is providing sample sources, you need to know what sources and something about their quality.  Ask how many sources the agency intends to use, what they are, and their experience with each source’s quality. A professional agency will be able to give you a clear, credible response. If you get a vague, confusing response, beware.
2. Risk: Not previewing final deliverables.  Ah the big day! The research agency is sending a sharp, knowledgeable professional to present your findings. The mistake? Not looking at their presentation at least 3 or 4 business days ahead of time.  It is totally appropriate for you to have a chance to preview and approve the presentation.  Don’t do it and you risk these very real scenarios:

• The presentation includes some conclusions that your colleagues, as experts in their market, know can’t possibly be correct. There goes the entire project’s credibility, all because of 1 or 2 poor slides.
• The presentation uses language that you just know will antagonize people. I am not saying you hide bad results; I am saying that you know your company culture and know how to present “bad news” in a productive way.
• The slides contain obvious errors.  Alas, it happens. The agency was swamped with deadlines the week your report was due, and insufficient time was spent checking your slides.  I know a research director from a Fortune 500 firm who bumped into a presenter in the parking lot 30 minutes before an executive-presentation; the speaker was still creating slides in the rental car!

Want more tips? Please check out the book on Amazon, “How to Hire & Manage Market Research Agencies.” OR, access a free chapter here: FREE.

Cathy’s point, “Customer satisfaction surveys are for measuring, not intervening.”

And Jeffrey’s, “Follow up with every dissatisfied customer who takes a survey.”

To be frank, my opinion on this topic has changed in just the past year or two. Before then, I was an ardent believer that all research must by anonymous—no matter what. I felt that any direct follow-up would show research participants that their survey responses could result in unexpected communications—and even if “helpful”, this experience could still impact future willingness to participate in research.

But in the past couple of years, two things have happened:

• First, I have been working with many clients who need to show that market research is not an academic exercise. Who need to demonstrate that research can directly, immediately, have positive outcomes. Many client-side market researchers have to negotiate for budget with non-researchers, who often view such studies as nice, but not necessarily actionable. Imposing anonymity on customer feedback reduces the research’s potential for clear, measurable usefulness.
• Second, I have seen raw data from several studies where it was obvious that participants expected follow-up. Indeed, anyone who has done a customer satisfaction survey knows that open-ended questions will often return entries such as, “The last software upgrade didn’t work—can you please fix it?” or “I have called your customer service number twice and can’t reach a live human being!”  You can bet that if they take the time to type that into a survey and you don’t follow-up, the damage will be irreparable.

Anonymity in Market Research

Yes, most surveys should be anonymous. But customer satisfaction surveys are an exception. Make it clear at the beginning or end of the survey that respondents can opt out (or opt in, if you prefer) of follow-up. Provide a phone number, web site or email address that can be used for any questions about how responses will be used.  The reality is that most customers expect follow-up.

What do you think? Do you agree? Have a different perspective? Please add your comment here or call the blog comments line at 508.691.6004 ext 702.

Want to learn more about customer satisfaction research? Check out the Research Rockstar class here: ClassList.

Why?

Because big projects are risky, slow and yield too much data for anyone to digest. Most of the data gets used ineffectively, if at all.

Small projects have less risk. They get done quickly. Results get shared while they are still fresh. And conclusions can be communicated within the attention span of a busy recipient, so they actually get used.

In market research, small can be beautiful.

What do you think? Do you agree? Have a different perspective? Please add your comment here or call the blog comments line at 508.691.6004 ext 702.

[Want more information about Research Rockstar classes or services? Use this easy form to request more info: InfoReq. Thanks!]

• Sample quality? Hmmm…. a huge problem, but no.
• Inappropriate use of methodologies? Always a challenge, but no.
• Poor survey design? Another good one, but I think it’s part of a bigger issue.
• Over-hyping of social media-based methods? A definite issue, but not the biggest.

The biggest problem we have is that too much research gets ignored. Too many studies lose momentum. Too many important findings get disregarded.  Even if end-clients sit politely in a final presentation, not enough of them use the research to make real decisions and take real actions.  And that’s a problem because it leads to negative perceptions of research itself.  Too many people perceive research as academic or theoretical; they don’t see it as something that truly can impact business success. Indeed, many non-researchers see market research as having a poor ROI.

If we want people to really use research, it needs to be more credible to them. And that means 3 things:

1. Let them see it, so they can believe it. When end clients are removed from the process, they are less likely to believe the findings. Especially any that contradict their personal opinions.  Jon Last, former MRA president and president of Sports and Leisure Research Group., talks about data proximity. He’s right. http://www.researchrockstar.com/why-internal-clients-ignore-market-research-results/
2. Let them have a say in it’s design and execution. Involve them in the process, start to finish. And I mean real involvement. Cross-functional teams sounds like a nice way of keeping people involved, and can work in some company cultures. But in reality? In most cases I have seen, the team members from outside the research function have neither real responsibility nor authority; they come to the meetings as clients or even just semi-interested observers. Research isn’t in their job description, its not what they get evaluated on, and they are easily intimidated by the “experts.”
3. Train them. Market research isn’t a secret society. A lot of the stuff we do is teachable. A little education will overcome many of the poor assumptions that lead to perceived credibility issues.

So here’s the point: to really address these three items, we need to re-think what market research functions should be centralized versus decentralized.

My proposal is that for many (not all) organizations, decentralizing market research can address these 3 items very well. Let different functional areas have their own mini-market research teams. Better still, have some  people in existing functional areas take on some market research responsibility for their groups; they will have a unique blend of expertise areas, which will help ensure actionability.

Objection: Research takes too much Skill, You can’t just Train People

My, what egos we researchers have! Do we really think that all market research projects are so hard, that other business professionals couldn’t possibly learn how to manage them?

In my 25 years’ of market research experience, I have seen that at least 70% of market research projects are pretty simple. They don’t require multivariate analysis. They don’t require lengthy questionnaires. They don’t require complex skipping or branching programming. Many don’t even require weighting.

Of course, 100% of research projects do require certain quality standards. Length. Objectivity. Scale choices. Knowledge of basic statistics.  All of which are teachable to any reasonably motivated professional.

One Vision For The Future of Market Research Departments

A centralized research function will remain an important resource, and support the decentralized teams.  In this scenario, the centralized research function of the future will focus on four functions:

1. Policy creation and enforcement. For example, how many times a year customers can be invited to research events, and what types of incentives are permissible. Heck, you can even have a policy that says, “Any questionnaires over 20 items long must be approved by (name of the organization’s market research director).” This is something, BTW, that many market research groups never have enough time to do!  If they can delegate some of the smaller projects, perhaps they will have more time for this type of important, strategic role.
2. Resources. Centralized access to research tools, secondary reports, in-house research results, standardized questionnaire templates, sample sources, in-house panels and communities.
3. Sophisticated studies and Trackers. Management of high-end, organization-wide studies (those requiring advanced analytics and longitudinal tracking)
4. Training. For example, either producing directly, or through partners, ongoing training, which will include “Market Research 101” as well as functional area specific content (such as “Product concept testing” for product development groups, and “Message testing” for marketing teams).

In this scenario, the decentralized functions will be specialists that serve the unique research needs of specific areas. They will have day-to-day relationships with the people who will be using the research. They will speak their language, and make research accessible and credible to them.  They will have the skills to do basic research, and have access to experts as needed.

Bottom Line: Improving Market Research ROI

A decentralization approach has the potential to boost research credibility AND also address the issue of rogue, unsanctioned, DIY research.  We all know there are plenty of bad questionnaires going out these days (though many come from “professionals,” too).  Clearly, more non-researchers WANT to do research. They want fresh insights. They want involvement in the process. So let them! With some intelligent policies, access to resources, and training, we can have the best of both worlds: quality research and greater research ROI.

Now what? I say: test it. Pick a functional area to test this with, preferably one where you know there is some rogue research going on anyway. You can take advantage of the existing interest to build a satellite research team.  Provide some training and policies, and give them some time. See how they do.  Do their projects go well? Do they have impact? Is the ROI good? Are the internal clients satisfied?  You just may be surprised.

As always, all comments welcome!

Note: This blog post was inspired by the Tweetoff held at the AMA’s virtual market research conference on June 24th. Thanks to @JHenning and @VirtualMR for a GREAT event! You guys are fun to debate with!

[Hey, did you miss my new white paper on online panels? Check it out here: GET WHITEPAPER]

There are some obvious cases where you hold it tightly. Data that is specific to proprietary product ideas is a good example.  But other cases aren’t so clear-cut.

When in doubt, share it. At least some of it.

A Tale of Lost Market Research Opportunity

A few years ago, I worked with a technology company on a large survey project. Very fresh stuff.  They discovered customer needs and values that could be leveraged in totally original marketing messages. They identified unmet customer needs that could be addressed simply by repackaging existing products.  In short: we had lots of cool data on our hands.

Enthusiastic about the research’s findings, a VP at the company invited me to a meeting with one of their largest distribution partners—a huge, well-known company. He asked me to share a subset of the data with them. At the meeting, the distribution partner had 12 attendees—all people who have direct influence over their business relationships. I presented some key findings. They loved it. They even offered to reciprocate by sharing some new research they had done. We started to talk about the results, generating ideas about possible implications for business opportunities and product innovations. We planned next steps.

It was great. There was real energy.

The next day, I got an email from my client. He had returned to his office to have his hands severely slapped for sharing research results. And not too long after that, this clear-thinking, collaboration-minded VP left that company. I wasn’t surprised. I would have, too.

The Greatest Market Research Risk?

Yes, market research can yield a competitive advantage. And it costs a lot of money. So I understand the inclination to keep it secret and for inside-eyes-only.  But really, in most cases, what is the risk? The biggest risk is that you share the data with someone and they give it to a competitor. How likely is that, really? It’s a lot more likely that you will give them data, and they will ignore it.

But if you did share fresh market insights with another organization and as a result created new opportunities, wouldn’t that be great? Isn’t that a risk worth taking?

If research results can help business partners, inspire clients, encourage employees or motivate suppliers, I encourage you to consider it.

The potential upside is fantastic. And few companies are bold enough to do it.

[What do you think? I welcome comments here or contact me by email: KKorostoff@ResearchRockstar.com]

After all, you may very well have a list of customers that would be convenient for you to use for your survey project. But is that a wise choice? Maybe, maybe not.

In the quest for objectivity, we often do research with a mix of customer and non-customers. At the end of the project, we compare their responses to answer questions like:

• Do our current customers perceive our brand differently than do non-customers?
• Do our current customers value our product category differently than non-customers?
• Do the two groups differ in terms of feature requirements? Price sensitivity? Other purchase behavior?

Clearly, there is a lot of value in surveying both current customers and “potential” customers.  But realistically, it can be a lot easier and faster to survey current customers. After all, you may have their contact information so emailing them a survey invitation is a snap. They know who you are, and may even like you enough to want to help out by sharing their opinions—so response rates are good. And you know they are real—no worries about the veracity of responses.

In contrast, getting access to non-customers can be expensive, time-consuming, and risky. Sure you can buy sample from a panel provider, or seek other options (partnerships with professional associations, user groups, etc that may meet your needs). But how much will it cost? How long to arrange it? What kind of response rate will you get? How are those respondents authenticated?

So, what’s a time-constrained, budget-sensitive researcher to do?

In an ideal worked, you would examine your research objective, and then decide if it is best met by surveying current customers only, or a mix. For example, if your objective is to choose which features need to be in the next release of a software product, you may very well want both groups. Current customers, to keep them happy and maximize their likelihood to upgrade to the new release when it is available. New customers, in hopes of attracting them to the product for the first time.

But if time or budget doesn’t allow reaching non-customers?

Then you need to ask yourself two questions:

1. Are my customers an acceptable proxy for non-customers? Do I know enough about both groups to understand how different they are—to have an informed opinion of the likelihood that they have similar attitudes and behaviors relevant to the study I am planning? The answers will tell you how much of a risk you are taking, so you can deliver the research with appropriate caveats for anyone who will use it.
2. Is there an acceptable fallback? Even 10 interviews with non-customers is better than nothing. Or perhaps an outside expert with relevant experience is available (try Zintro or GLG for quick access to expert networks) for an objective sanity check.

Customer insights gathered through research deliver great value. But we always need to think about what we really mean by “customer”—current customers, or potential ones? Even if time or financial constraints prohibit you from conducting research with the ideal mix, take the time to understand the risks or consider fallback options.

[Subscribe by email to get the latest blogs as they are published. Click here:  Yes, Send me Blog posts by email!]