The awkward part of a survey happens before the first question. You need people who fit the brief, in the right countries, willing to answer. Then somebody must make the questionnaire work, move the data, and deliver something a client can use. GMI built its business in those gaps. Its subject was human opinion. Its daily work was logistics.
- GMI combined research software, respondent panels, and survey services.
- Its buyers included research agencies and corporate insights teams.
- Kantar acquired it in 2011; the GMI name left the combined brand in 2016.
First, connect the job
Rob Monster founded Global Market Insite in 1999. The company belonged to an early generation of businesses asking what market research might look like when the Internet became its working environment. Researchers could already formulate questions. GMI offered infrastructure for getting those questions into the field and bringing the results back.
Its Net-MR software linked project administration, interviewing, panels, and reporting. A Federal Reserve Bank of Boston research guide described a suite intended to automate work across the project lifecycle. The appeal was practical: fewer separate arrangements between the initial brief and the finished analysis. A research department could use software and outside respondents without becoming a software company or a panel operator itself.
That positioning placed GMI between specialist software vendors and the agencies buying research infrastructure. There were alternatives for individual pieces: Greenfield Online supplied respondents; survey packages supplied tools. GMI’s pitch put more of those pieces together. In 2002, MSRI signed an enterprise Net-MR license for Web-based collection and reporting. The buyer was another research business.
The rough edges came early
A 2001 Quirk’s review caught the tension inside that ambition. TNS Intersearch used Net-MR to coordinate 80 field service companies on a complex consumer study. Yet its operations executive said some tools lacked production strength, particularly telephone interviewing. The reviewer also found survey routing cryptic and help material poorly organized. These were specific early product shortcomings, rather than evidence that the whole company had failed.
“The speed of innovation is unreal”
Ian Kiernan, TNS Intersearch, in 2001
The lesson for a buyer is straightforward. A connected system can remove handoffs while introducing dependence on one supplier’s weakest component. Test the complicated interview, the awkward quota, and the reporting requirement before committing a live study. Convenience becomes valuable when it survives the actual job.
The audience was another product
Software could organize fieldwork. It could not conjure the right respondents. GMI also supplied consumer and specialty panels, with audience profiles that helped researchers target particular groups. Its 2007 advertising placed IT professionals, physicians, expectant mothers, and other audiences in a bright collage. Behind that cheerful presentation sat a serious commercial question: who can reliably reach the people this study needs?
Synovate’s 2007 worldwide agreement supplies a useful answer to the cost question. Its chief executive explained that buying GMI panel access helped avoid building proprietary panels in markets used less frequently. The attraction was shared capacity. Owning every recruitment operation makes less economic sense when demand arrives only occasionally.
GMI also worked with iHispanic Marketing Group to expand Hispanic panel recruitment. That partnership illustrates a different constraint: access depends on distribution and relationships with particular audiences. A broad international footprint and a well-recruited specialist group solve related, but distinct, problems.
The commercial model had several entry points. Customers could license technology, buy sample, or commission operational work such as programming, translation, hosting, and data processing. Net-MR 5.0, announced in 2005, could integrate with existing applications, including incentive management and customer relationship management systems. That mattered to organizations with tools already in place: adopting GMI did not require starting the entire operation from scratch.
There was capital behind that breadth. Contemporary industry reporting recorded a $35 million Series C in 2006, led by Technology Crossover Ventures, with FTVentures and Voyager Capital participating. The following question is an editorial one: does adding another capability reduce the customer’s work, or merely enlarge the supplier’s catalog? GMI’s most persuasive examples involve a concrete operational need, whether that was entering an occasional market or running a multinational study.
More people. Still a sampling problem.
By 2010, GMI was selling Pinnacle, a calibrated sample-blending technology. It was designed to balance respondent sources against a benchmark using at least 60 demographic, psychographic, and behavioral measures. The product acknowledged something easily lost in a panel-size headline: different sources bring different skews.
The reasonable inference is that scale alone was an incomplete answer. Pinnacle’s announced purpose was to manage composition and consistency. Its existence does not establish that every resulting dataset represented every population. Buyers still needed an appropriate benchmark, a reachable audience, and a sound questionnaire. When those conditions fail, a polished collection process can produce a misleading answer very efficiently.
IT leaders surveyed for Cisco and Intel, across 18 industries and nine countries.
GMI’s work extended beyond consumer opinions. Cisco and Intel commissioned it to survey IT leaders about cloud computing in March and April 2013. Respondents answered anonymously, without those sponsors being identified. The eligibility criteria focused on people familiar with cloud concepts at organizations with cloud experience or immediate investment plans. A useful audience starts with a definition.
The name got smaller. The operation joined up.
GMI reported $64 million in 2010 revenue. Kantar acquired it the following September and aligned it with Lightspeed Research. GMI added panels in 13 countries to Lightspeed’s footprint. The often-repeated figure of more than four million panelists across 38 countries described the combined reach after the deal.
In 2014, the integrated business adopted the Lightspeed GMI identity and the promise “We Make Research Easy.” Its chief executive described a move from a traditional panel company toward a broader digital data collection enterprise. In 2016, Kantar announced another naming change: Lightspeed GMI became Lightspeed.
What can another business copy? Map the expensive transitions in the customer’s work. Decide which capabilities to own and which to share. Make the audience definition as deliberate as the software purchase. GMI’s history suggests a useful place to look for a business: the necessary work everybody depends on, between the visible steps.