Sunday, September 20, 2009

Google as a cheap market research tool

For years I have been involved in getting information about 'awareness' levels for charities. This would be done through polling 1,089 people (give or take) and asking them to name ten charities. This would give us an 'umprompted awareness' measure.

A PR / awareness campaign would usually be preceded by one of these polls and then followed with one. The relative impact on awareness would be measured.

One of the most useful methods of measuring the effectiveness of PR and awareness campaigns.

I am not a fan of such campaigns if their objective is to raise money (fundraising tends to work better, and I prefer targeting the 'right' audience,not the 'general public'). However, such campaigns are useful for social marketing (like anti-smoking messages) and this unprompted awareness measure is probably the best metric.

However, I wondered if there is a much cheaper way of achieving the same aim?

Surely, my theory goes, Web Search terms will 'rank' organisations in the same pattern as polling. Using Google Insight, I looked at five charities from the top ten fundraisers to compare the number of searches over time. Red Cross peaked during the Victorian bush fires (Nearly 100% more searches than back in Tsunami times).

But apart from the peaks, the trends do seem to look like what we would expect from awareness polls.

(Click on the picture for a clearer image, or if that doesn't work, click here). Please note 'TOTALS' are not absolute numbers, they are relatives.

Now all I need to do is find some awareness poll data from the past few years and see if there really is a correlation. If there is, and if it can be substantiated by other results, we could be looking at a massive saving on costs for lots of charities measuring their awareness campaigns.

If you work in a charity and willing to share with me a few years of unprompted awareness polling data, please contact me!



Latinblog said...

Hi Sean, good article. Did anyone come back to you with poll data to prove correlation? I work for the salvos but not sure if we got some, will dig into the other depts. files. cheers, Esteban

'Sean is always learning' said...

Hi Esteban

No, please can you dig for me!? Thanks a lot!


Jeff Brooks said...

Hey, Sean. I think BOTH polling data and Google trends will provide you with fairly bogus data. Polls only tell you what people think (or claim) they know. Google only looks at searches, which are done by a subset of the population.

On the whole, I'd far prefer the Google source. It's free, and it's at least a scientific look at behavior, not just what people say.

Disaster Fundraising Guide download it here