Amongst the turmoil surrounding Facebook and their ad services - in particular the Business Beacons - Quirk has been super busy researching the ins and outs of the Facebook services as a marketing platform.

As the final stage of this exploration, I took it upon myself to see how Polls can be used as a market research tool. The results were, unfortunately, rather inconclusive.

After moving through a very simple implementation process, with a relatively high cost ($26 for 100 respondents), I was on my way to finding out “Which eMarketing service is the most productive?”. The first problem I ran into is that you can only select 5 options (which is awful when asking this sort of question, but is in line with Millers’ research on working memory).

The options I selected where:

• Banner Advertising

• Email Marketing

• Viral Marketing

• Search Engine Optimisation

• Pay Per Click Advertising

I then selected a targeted sample of any current or past university students with marketing as a major. After letting the poll run for 24 hours I received the following statistics:
 


No surprises here.

This confirms what appears to be a marketing sentiment for the various core objectives a company could have (branding for banners, interactivity with virals and site traffic with SEO). I was recently reading some of the old posts on GottaQuirk, and I came across one by Rob entitled “What is the best performing eMarketing tactic?”. From this post it seems that market sentiment hasn’t changed all that much.
Facebook polls also provide very interesting demographic stats according to both Age and Gender. This is where I see the real value of these statistics. In this example the various age groups show some interesting outliers and trends.
 


Looking at the outliers I automatically see a potential problem. While it isn’t shown in the graph above, both the age groups of 13-17 and 35-49 had only one respondent. This may be chance, but it worries me that Facebook may be modifying/inflating the stats (just a feeling I get).

Comparing Viral Marketing to Banner Advertising (the two highest scorers) there is also something interesting to be noted. Students (18-24) believe equally in Banner Advertising, Viral Marketing and Search Engine Optimisation, but don’t believe in the advantage of Pay Per Click advertising. The young professionals (25-34), on the other hand, support only Viral Marketing and Search Engine Optimisation. I don’t have a degree in statistics (despite the two years wasted on it) but I believe this to be significant from a trends perspective.

Basically, students seem to have an increased overall awareness of eMarketing services. Once they leave the academic (brand orientated) environment it appears that their perception of the most productive eMarketing strategy changes. While the reasons for this could be almost anything, we can assume it is based on the experience they have gained.

Valuation of Facebook Polls:
Final Grade: C,
I think the service can really be improved upon. Currently you can only ask 100 people, with the possibility of each individual only being profiled on one variable (i.e. age OR gender, etc.). Ideally I would want to really use the available variables and optimize them to the max. Over and above the results, I think the cost is a little high. When converted to Rands the service costs about R4 per person, which makes it very expensive to get a sample of an acceptable size. Bar these “nice to haves”, I think the service is incredibly useful and the 24hour turn around time is a definite bonus. I see this service being used in small businesses and startups where the market research resources are very low. Unfortunately, Facebook Polls doesn’t cut the grade for large business where statistical relevancy is important.

-- End of field report