4 months ago, Google launched its Do-It-Yourself research tool, Google Consumer Surveys (#gsurveys). While it would be unfair to say that Google Consumer Surveys staggers forward, coughing blood, it would be equally unfair to say the launch has been a robust success.
Two problems are key:
- Not enough publishers have signed up to host the content.
- Not enough DIY researchers have signed up to generate content for the publishers.
This is the critical push-me-pull-you for Google Consumer Surveys (#gsurveys). Without enough publisher websites, it’s impossible to have enough eyeballs looking at the poll questions to get a necessary cross-section of respondents. Without enough DIY surveys to publish, it doesn’t matter how many publishers there are, there’s no research questions to feed to the waiting eyeballs.
Let’s examine two different Google Consumer surveys sessions that we studied leading up to the 4-month anniversary of GCS launch.
In session #1, we cycled through 227 questions and never once saw a new question. Every question served up to us was either a question we had seen in a previous session or a question we had seen previously in this same session. Take as an article of faith: if Google Consumer Surveys had a new question to serve us, they would have served it to us. The only possible explanation for this is: Google Consumer Surveys had no new questions that it needed to have answered. 120 days after launch, this is pitiful.
Below, we give you the full text of all 227 questions that we saw, so you can see:
- The warp and woof of the kind of questions that DIY researchers are using Google Consumer Surveys to answer.
- The way in which Google’s coders have programmed the queue of waiting questions.
- The extent to which Google is padding the queue with “canned” (aka: “stale,” “worthless”) questions, so that a respondent does not see a dry hole, and so that publishers have some monetization coming their way.
The table below keeps rough track of when we first saw a question appear in rotation on Google Consumer Surveys. Since Google claims that research fielded on GCS can be completed in hours or days, questions that are 60, 90 or 120 days stale are clearly serving no client purpose, though it’s conceivable Google is tracking the answers for some internal purpose.
Chart continues after the jump …
To see if this was typical or atypical, we started a new Google Consumer Surveys session the next day. In this second session, 27% of the questions served to us were questions we had never seen before (aka: “new” questions). 73% of the questions served to us were recycled questions. The full session is re-created below, so you can get a better feel for how Google’s programmers code the queue so that any available “new” questions pop to the top. We stopped the session after we began to see “recycled” questions that had already appeared once before in this same session.
Google has unlimited resources. For Google Consumer Surveys to succeed, Google will need to deck more of them against #gsurveys.
To see everything we have written about Google Consumer Surveys, click here.