Public Interest in Space, Google-Style
Avid Google users may know that the big G offers a limited / free view of their historical search data called “Google Insights“. Insights provides search data back to 2004, and allows neat tricks like geographic subsetting (down to the city level), overlays of major news events, and more. For kicks, I’ve started tracking the Google search interest in a couple of space-related keywords, to see how the public has reacted to different administrations’ plans to go (and then to not go) (and then to maybe go) to the Moon, Mars, and beyond.
First, a short tutorial on what Google allows you to glean from their Insights data. Insights reports the relative search interest in a topic, not the absolute interest (presumably, Google would like to sell rather than give away the latter). So Google Insights provides a numerical score which shows the relative likelihood that people are searching for a given topic rather than something else. To make it even more confusing, the score is normalized to its maximum value during the search period.
A picture is worth a thousand words, so here are some actual data:
. Google’s record begins in January 2004. Coincidentally, the second week of January 2004 was the week that President Bush delivered his speech on the Vision for Space Exploration (a year after the Columbia accident), a plan which triggered alot of online searching (the most in the 2004-present sample, actually, hence its relative score of 100). The next highest peak is in July 2005, and if one drills into the data, it’s easy to show that this corresponds to the first Shuttle Return to Flight after Columbia, which triggered alot of interest. Its score of 60 denotes that the public was 60% as likely to be online searching for “NASA” then, relative to the week of Bush’s speech. The Google data can be drilled all the way down to week-by-week levels, yielding a very rich dataset. (The effects of every Shuttle mission are clearly visible in the weekly data.)
To further illustrate the “relative” scoring, visiting the full Insight page for this search reveals that the highest regional (state) score is found in Alabama. This isn’t because more people in Alabama are searching for NASA, it’s because people in Alabama (Huntsville, specifically) are much more likely to search for NASA rather than something else. (To me, this makes perfect sense. There’s alot more going in other cities that host NASA centers, such as Houston, Norfolk, DC, Cleveland, etc; in Huntsville, NASA has much greater relative ‘market penetration’ into the public consciousness. This is probably not unrelated to the fact that Huntsville engages and employs the 4th highest STEM-related workforce in the country, as shown in Table 3-8 here.)
Keen viewers will also spot an overall decline in public interest in NASA, perhaps even beyond the “expected” drop after the news of President Bush’s speech, and RTF, wore off. To probe a little deeper, I reassembled the data into annual averages:
Since the data are sensitive to significantly enhanced public interest during shuttle launches/missions, and the number of those per year varies alot for both technical and operational reasons, I also computed annual averages excluding the data from the weeks of, and the weeks following, a shuttle launch (the yellow curve above), as a crude control sample. The results are the same, and not surprising: NASA has its work cut out for it in re-engaging public interest after the two years of extended uncertainty attached to cancellation of the Constellation program. To date, it seems that the public hasn’t been tuning in to the new plans (the data suggest the opposite).
Of course, NASA does not exist to promote NASA, it should be spurring interest in space development and exploration overall. Tracking Google data for “space exploration” yields a very interesting conclusion: this topic appears to be overwhelmingly driven by education interest, as shown by the clear “summer vacation” in every year’s data:
(here’s the full search page link). Lest the power of teachers be forgotten, the data clearly show that searching for “space exploration” is twice as likely during the school year, as during the summer.
As importantly, interest during the school year itself isn’t monolithic. It appears that teachers use space exploration as a “late semester” topic, with peaks clearly evident in November and April/May. The differences are again notable: in the spring, search interest in May is 15-20% greater than in February/March, and the April/May peak is a full 35-40% higher than in September. Recalculating (and displaying) the averages based on actual school years (rather than calendar years) drives the point home even further:
This is not at all surprising, but one would like to think that NASA has aligned its education programs and supporting online content to the way that teachers actually use the topic of “space”, providing fresh content on a schedule and in a way that takes maximum advantage of the November and May space “sweeps weeks”.
This topic is of keen interest to me as my day job partially involves public communications and outreach, and it is my take that NASA has not yet moved to a measurement (or outcome) driven view of communications, beyond periodic purchases of consultant-driven survey and focus group studies, which (quite frankly) keep telling us the same thing. In a “Mad Men” marketing mentality era when this was the only means by which to gain situational awareness, this was perhaps understandable. However, rich web analytics and quantitative assessments of online brand penetration are becoming de facto standards in the “real” world. If even the imperfect Google data shown above were shared in a “real” board room, my hunch is that the marching orders the next day would be extremely clear and unambiguous.