What is Google up to with Consumer Surveys?

The irony is that today's lengthy online surveys are not 'real' research at all given the horrendous non-cooperation and non-completion rates. Everyone would like to think it is, but it’s not. In fact, it's better to have one question answered in a representative manner than 100 questions answered by the lone few who have the time and perhaps the questionable motivation to complete a lengthy survey in today's non-stop world. Research professionals need to objectively assess this tradeoff by conducting side-by-side testing before running around and say this is not 'real' research. And once the results come in, all the stakeholders needs to be prepared to determine which measurement is best given the results will surly be different.

Like Reply Report

Nice blog Simon. One correction though if I may be so pedantic. Toluna we've been in the DIY software plus sample for nearly 4 years with our QuickSurveys product (www.tolunaquick.com) which allows users to build and deploy their surveys to our community without so much as picking up the phone to us. I don't see Google and the latest raft of DIY tools cannibalizing the MR industry so much as growing it - in fact Google's move is the clearest validation of the growth and scope of this self-service market I've seen. But you're right - the traditional data collection companies with no response to this news should be tracking developments at Google very closely. Bring it on, I say.

Like Reply Report

Simon, while I agree with pretty much everything you state here, I beileve there is more to be said regarding the pricing and given that Google probably have spent a fair time thinking about how to price this thing, it seems to me that it also tells a lot about what has been going on inside Google. Google´s pricing is, as you say, ridiculously cheap (unless you compare with existing automated online omnibus services, which I believe this should be compared to, rather than buying serviced sample from Toluna et consortes), but it is interesting to note that it is ridiculous only for very few questions and gen pop, or, very few questions and very hard to get audiences. For 6 questions and upwards and with one simple demographic critiera added, say gender or age, something which traditional panel providers typically don´t charge for, Google very quickly becomes "ridiculously" expensive. Why is this? Google seems to have chosen to be disruptive in only parts of the online data collection market. This must in some way be connected to the fact that the surveys will be competing for space within the same ecosystem as Adwords, and thus, pricing must be somewhat in line (regardless of hardware at that, computer, pad, or mobile). A demographically targeted ad, seems then to be something that is quite expensive within that ecosystem, probably as a result of great demand, whereas extremely hard-to-get audiences or gen pop seem to be less expensive (the reason could be that the audience as such is not as important from an advertising point of view - Adwords are not able to get their attention or in the case of gen pop that supply is infinte). Or, the other way around, it could be that Google isn´t good at targeting demographically based audiences, why it is lack of supply that keeps price high (less likely). Either way, Google needs to keep their partner network happy, and if Google replaces Adword revenue with a smaller GCS revenue, that won´t fly, for obvious reasons. If Google is able to show that they are helping walled-in publishers to gain more readers, they might be able to replace revenue up-front with more readers, and then the crowding-out effect would over time be smaller, and the price for "in between" audiences within CGS could go down, but that remains to be seen. Or - it could also be that Google skips the "more-than-six-questions-per-survey" market entirely, based on the fact that the probability of longer surveys being managed by market researh firms and/or panel providers is of course high, but that doesn´t really make sense. Why price yourself out of a market, if it is within your reach, if not immediately? Google might be able to boil some frogs, but to be truly disruptive in this space, they will have to at least part-boil their own Adwords/Adsense frog at the same time. That, I dare say, won´t happen. Thanks a lot for a good read, these are interesting times, indeed. I believe Google entering into this space is exceptionally good - it has shown to everyone, big or small, the importance of listening to your audience and getting the insights you need to keep abreast with customer and market development. This will greatly help in broadening the market for market research. Best / Bo Mattsson, Cint

Like Reply Report

Many thanks for your comments, which bring up extremely good points. Just one clarification: the quotes we got from Toluna and others were on their automated systems (i.e. TolunaQuick). The issue Bo brings up is an interesting one, as it is the same one that AOL faced when it owned DMS. The research side of the equation was constantly struggling to gain real estate because the advertising side was consistently more profitable and a much bigger potential market. If GCS threatens any ad revenue, it will quickly be relegated to the back burner. However, it may well be that Google is big enough to be able to handle both things simultaneously and synergistically, in which case they will pose a big headache for existing suppliers.

Like Reply Report

Recently,I used SurveyMonkey and it cost me over around $600 for my survey with a sample size of 100. Survey Monkey charges you approx $225/ year membership (you have to be a member in order to access their survey group) and $3/ response. I requested only sample size of 100..so all together my size is small and cost me about $600 with taxes. I find Google's method to be more targeted and random. I actually don't know where they get this group of users...It's definitely not as attracted as Google's real users vs paid survey takers.

Like Reply Report

Theres one big point regarding pricing you're missing. If you want Google's broad demographic targeting it's $0.50 per question. 10 questions at basically 100% incidence is $5.00 CPI. This is much higher than any of the sample companies. Make that 20 questions and Google is way out of the ballpark on price. Google allows three tiers of age, gender and wide geographic targeting. With any tier one sample company, your targeting (income, ethnicity, marital status, children in house, zip codes, etc) can be much tighter at half the price of Google.

Like Reply Report

What you all seem to be missing is that this is Google MR Beta. All of this stage is just to test and develop a commercial MR model. Once Google extend the system to opt-in respondents, global markets, online analytics, etc this will be the death knell of Online panel providers. Hand on heart they deserve it. Panel books are full of over-representation of quality respondents, response rates are declining and the industry if full of varpet baggers. The erality is it seems impossible for panel companies to get the kind of granularity that Google will be able to provide. And we all know panels struggle like hell to obtain older respondents, which means major weaknesses in representation of key decision making segments in many markets. One only has to take a look at the massive bias in Panels towards singles over marrieds and under 35's to over 35's. Google not only represents a major challenge to national and international suppliers but will really open the door to targeted local area and segment markets. I love this, it is the biggest sea change in the industry since I entered 40 years ago. Go Google!

Like Reply Report

If Google want to revolutionize the industry they may have some success basedont their access and brand name. But given that most surveys are 5-30 minutes we see that they will have to change the market for mr completely or rethink most of their set up beta or not. 1-2 questions will perhaps be good enough for a limited part of the market within the brand or fmcg part of the industry only.

Like Reply Report

Congrats on the blog Simon! Besides the cheap pricing, what really, really positions Google as a rule breaker here is their audience, potentially they can reach real consumers everywhere on earth in seconds and not professional survey takers coming from the same sources. I know exactly how recruitment works across the industry where most of the panelists come from the same bunch of suppliers who use surveys sites and coupon sites to recruit from. All the major panels out there have been built using the same kind of Affiliate Networks, if you want to run a quick test just search for “ survey money” using Google and you will see all these proffesional survey sites that promote most of the “big” consumers panels out there. Agree with all your points Simon. However, I will add that Google not only got a competitive advantage in terms of price but also in terms of quality and reach.

Like Reply Report

The Quality of respondents is the biggest pain area in online data collection. I think a brand like google entering this space will usher a debate on how to improve the quality of survey takers.

Like Reply Report