Brand advertising analytics

I recently worked on a study that tried to understand how much is too much money on brand advertising efforts.

Performing too much analytics on brand advertising and using the numbers as is to determine effectiveness is not a good strategy because of the following reasons:

a) Brand advertising is to create a perception and may not always lead to conversion. 

b) Pure analytics numbers will tell the number of click throughs and frequency of return but it gives no idea about the subconscious brand attachment. Its just not measurable using technology.

c) Campaign analytics is good for validating delivery as per your intent.

So this study involved a brand campaign that was run for a day on Youtube masthead. 

Quick facts:

1) Roughly 6 Million impressions were served to 580,000 browsers in 1 day.

2) So that means each browser saw over 10 impressions! I believe that the reach of the campaign can be increased if there is a way to throttle the ads on Youtube.

3) Only 2.21% people clicked through on the masthead ad and went to the site. I cannot give you the conversion rates for this specific site but standard conversion rates are around 3-4%. A tiny number of browsers may convert, but then chances are they know all about your brand before, your brand campaign just motivated them to finally get to something that they were trying to do later.

4) Final thing that I noticed is none of the 2.21% browsers came back to the site. So which means that brand campaigns may have initial traffic acquisition effect but then have very less frequency.

Now, readers may comment that this is very specific study and I do agree. So please don't start posting comments criticizing my post yet. This is a very specific study and data points are just not enough to conclude anything. I would imagine a proper study would classify their campaigns in following areas: Industry, campaign frequency, time period of study, copy of the campaign,  site flow and merchandizing on the site, reach, frequency and finally spending.

My intent was to make the general reader aware about the possibility of running such a study if they are trying to know how much money is too much spending on brand advertising. The study can be performed if your analytics platform allows mixing two data sources together and you have a way of collecting exposure data with your web analytics data.