PureSpectrum - Schedule A Demo
Qualtrics: Here to Help

Step 1: Get Out of the Survey Business

Why should you move away from surveys? Your interviewees are increasingly fatigued, and your clients are now your competitors. There are no longer any technical barriers to entry, and the technology itself will move faster than your ability to remain competitive.


By Jason Anderson

A couple of weeks ago, I wrote what turned out to be a somewhat controversial opinion piece about things I would do if I were a market research company. The writing process was not a “fair and balanced” editorial; it reflected my current opinions and beliefs based on what I believe are long-term trends in how research is conducted and acted upon in my corner of the world. I listed eight changes I would pursue, if it were my decision to make. Most of these changes were very substantial, but didn’t include much depth as to why or when or how.

The first one, and I think perhaps the one that stirred up the loudest disagreement, was to “get out of the survey business.” This is obviously not an overnight change. Rather, it’s a call to action to prepare for a future model where the design, programming, fielding, collection, and data processing of surveys has greatly diminished in value. The arguments in favor of this are most obvious in the online space, of course, where we’re bombarded by “opportunities” to be surveyed at every turn. To me, the long-term impact is no different from the challenges faced by advertising: new generations grow up desensitized to media (and surveys), and grow increasingly adept at escaping or manipulating our efforts to interview them.

In other words: not only will surveys be of less value to researchers, but to the interview subjects as well. The days of compensating consumers with virtual points at an exchange rate of $0.50 per hour won’t last forever.

Ignoring the effectiveness issue, there’s also the simple matter of pricing. A cost-conscious research manager and a couple of skilled employees can already generate output faster, cheaper, and of comparable quality to that of a research vendor (assuming you hire good people). Do-it-yourself technologies have (as I’ve said before) removed most of the technical barriers to conducting a quality online survey project. If we simplify the value proposition of a research vendor to needs assessment, methodology and design, execution, analysis and synthesis, and measurement of any post-study actions, where do you currently spend most of your billable hours? Where will you need to be spending them instead if I don’t need you to execute the study?

“But wait!” you say. “We’re a cost-plus industry!” That’s true, most research vendors are cost-plus…today. And the costs of sample acquisition and survey execution offer you a nice buffer in your financial model. But Zoomerang and SurveyGizmo are not cost plus, and the existence of this competing pricing model has permanently changed the perceived value of surveying.

Finding true insight from piles of data, however, is hard. There’s both art and science involved; this is where your years of experience with thousands of studies offers value and competitive advantage. You don’t want to be your clients’ go-to survey supplier, you want to be your clients’ go-to analysis expert. You want me to choose you for a project not because you’re the cheapest, but because you can find things in the data that I can’t. Sometimes, that will be because I don’t have enough human capital to invest in the data; other times, it will be because you have better or specialized skills that I can’t justify year-round.

In summary:

  • Why should you move away from surveys? Your interviewees are increasingly fatigued, and your clients are now your competitors. There are no longer any technical barriers to entry, and the technology itself will move faster than your ability to remain competitive. You will still be involved in surveys, of course, but this shouldn’t be your identity (unless you’re offering a proprietary technology or methodology).
  • When should you be ready to change? Before your clients drop you and start doing things themselves.
  • How do you make the change? Sell expertise and analytics, not execution. Acknowledge your clients’ DIY efforts and help them do it better. Stop trying to compete with me for my own work. Add value before I start my DIY phase through consultation, and after I’ve collected my data with modelling and analytics that exceed what I might do on my own.

Please share...

8 responses to “Step 1: Get Out of the Survey Business

  1. Instead of getting out of the survey business, I would challenge that market research companies need to reinvent the survey. Most respondents want to give good quality answers, but as an industry we drive them to bad behavior with long, boring, and repetitive surveys. Market research companies need to work with their clients to develop engaging surveys that will provide good quality data, which will lead to sound business decisions.

    Also, from the previous posting I wanted to say I couldn’t agree more that market research companies need to recruit technologists. Once hired, we need to make sure the technologists are not working in a vacuum. They need to work with the researchers to develop tools that will help address client business questions.

  2. Great conversaion piece.

    The survey business has many faces from the simple Survey Monkey to the technology sites that spit out regression analysis reports. The former will always have a place as their srevices are very inexpensive and appeal to those that think having scores in a pie chart are the means to path a way to better a company. Unfortunately for the companies, there are too many of these type shoppers for surveys.

    Those survey results used as a consultative tool to idnetify the hidden gems are a different story. These business requires 2 competencies: a great analytical tool, PLUS a competency in in analysis, change management, alignment skills and an implementation plan that will breed success.

    This is the business that is innovative. These skills and tools must be learned, tested, adapted for clients. There may be many entrants into the survey market, but very few in the “transformation” business. Data is easy to get; change from data is not!

  3. Interesting thoughts, but maybe oversimplified?

    To your point, I agree that there are many exploratory, simple surveys that can be conducted in this manner.

    However, at least to this point in time, the free survey tools cannot handle the more complex logic and exercises required to meet objectives on many studies.

    What I have a hard time wrapping my head around (and selling to clients) is the almost astronomical drop in importance of random sampling and adherence to sampling staples such as non-response.

    So maybe not getting out of the survey business altogether, but offering analytic prowess and insight as you indicated, with the ability to sell and conduct more comprehensive research as needed.

  4. I would agree that seeing the collection of survey data as integral to your business model is a very bad idea. No question we’ll be able to replace much of what we collect today using other forms of data in the future.
    I don’t think we’ll ever fully replace the survey. Some types of data will not be available to us any other means (and I believe there will always be a way to get respondents). So understanding how surveys can fill in the blanks missing from other available data will be necessary. The synthesis of the two is what clients will value.

  5. A point on survey software and the ease of conducting a quality survey. Many research companies today cannot define a consice set of research objectives and develop an effective survey instrument.

    I think to all of the surveys conducted in-house at the manufacturing companies I have worked for and I just don’t buy that “easy” technology is going to fix the underlying problems with self-administered research…..

  6. I’m currently a client-side researcher (coming from an agency) and I have to say – I completely agree with this article. I’m continually asked to do cheaper surveys, with the same level of respondents (partly at my own insistence for good data) and as internal teams learn more about ‘SurveyMonkey’ and ‘Zoomerang’ – I’m finding myself being asked to do more surveys myself. This saves money and time.
    Several years ago when these DIY survey tools started becoming popular, I scoffed at the idea of DIY surveys. After all, they didn’t have a researcher to correctly word the questions and analyze the results. But as times have changed, copying a correctly worded questionnaire from the many research surveys that have peppered society has become a lot easier. There are plenty of templates and everyone knows the infamous agree-to-disagree scale. However, the analysis side of things hasn’t changed. Sure, a lot can give you a percentage. Fewer can decipher or know when to correctly run significance testing. Even fewer know how to use and analyze correlation. And minute percentages of people can go beyond that.
    So we researchers can add the value of analytics (qualitative or quantitative) and being able to take masses and masses of data and tell businesses ‘what it all means’. SurveyMonkey can’t do that on its own. The marketing department (unless it has a trained person) often can’t do that – researchers can.
    I don’t think the survey is dead – in fact, Zoomerang has made it immensely popular to the point where society is becoming blinded to surveys. But, as Jason noted, research value is in the consultation, analytics and making sense of the data.

Join the conversation