Google’s new research offering – Google Consumer Surveys – is attracting understandable interest from the research community. Lenny provides a forthright overview on Greenbook, and I couldn’t hope to match his big-picture take. But here are seven things I’ve noticed, wondered or thought about the new Google service.
1. TRUST THE NUMBERS: Google is famously puritanical about data and this extends to its research service, which uniquely among DIY offerings doesn’t support open questions. Verbatims begone! This fits with Google’s overall selling point – they’re not just letting you ask questions, they’re taking care of the analysis too, which they couldn’t easily do with opens. Even so it’ll be interesting to see how – or if – Google helps its customers get to the “why”.
2. WHERE’S THE MAGIC? Crunching data – with the occasional hint towards meaning – is what Google do, but in general on a far larger scale than an n=1000 survey. Google products like NGram and Insights For Trends surprise and delight because of the scope and scale of their dataset. Can Google work its magic on a more intimate scale?
3. THE FORBIDDEN ZONE: The tool comes with a huge number of caveats – some rules, some guidelines – and they’re revealing about how Google expects its tools will be used. Don’t ask anything about demograpghics – Google will handle that stuff. Don’t “promote” booze, drugs, medicines et al (can you even ask about them?) Stick to English. Google is having to think hard – harder than any other research company has – about the boundaries between research and spam, since it knows that every tool it introduces gets exploited by scamsters very quickly. The danger, of course, is that spam surveys might further damage research’s reputation…
. ACROSS THE GREAT DIVIDE: And in fact, so could the surveys which DO get through – Google’s use of “promote” is very interesting. The divide between marketing and research has been worn away for a long time but still exists in principle within most industry codes (and is a bulwark against regulation). Google rule out “calls to action” questions, but their list of things they don’t want you to “promote” suggests that promoting in general isn’t a worry. In Google’s world surveys, like the adverts they’re in place of, are essentially promotional tools – just not crude ones.
5. FORM AND CONTENT: How does Google’s offering fit within the “New MR” movement? Its form is radical – single-page surveys and (deliciously) a 200 character limit on question text. As Lenny points out, this is a mobile-ready version of research. But in content terms GCS is trad, not rad, and makes some very old school assumptions about respondent capabilities. Take its suggested phrasing of a question about orange juice, swapping “Did you drink OJ?” for “How many times in the last year have you drunk OJ?”. This is switching a dull but answerable question for something very few people could answer accurately! That’s hardly a problem with the tool itself, of course.
6. THE RESEARCH EXPLOSION: Despite all this I think GCS should be a big hit – it fills a need which has been obvious for a while: reliable, quick, affordable research which will be especially useful to small businesses. The effect will be a massive widening of the research customer base – most of whom won’t ever need, and could never afford, a traditional provider. Which, if you believe in the idea of market research, has to be good news.
7. LIVING IN A POST-GOOGLE WORLD: How will this shake out for bigger agencies? I don’t pretend to be a business analyst but I suspect a lot of them will be loyal Google customers! It looks like it might be a cost effective way for us to ask simple or boring questions, which should make our projects leaner and more interesting. More generally, I think the best comparison is Google Analytics. It’s the foundation of understanding your digital audience, and if you’re a small business or site it’s almost certainly more than good enough. If you’re big, though, you’ll push up against its limits and you’ll find it doesn’t give you a lot of strategic insight. I suspect GCS will fill a similar role – useful for everyone, sufficient for many, and a cornerstone of any added value analysis the rest of us do. Of course, it’s also worth remembering that Google Analytics is unrecognisable now compared to its launch – if Google Consumer Surveys is a hit, it’ll add features well beyond the obvious ones.