My question is, how should we decide how much money to spend on advertising if the return on this advertising isn't easily measured?
For example, spending it on an adwords campaign to acquire new leads in an online lead-gen business when the leads are included as part of a subscribers monthly fees. Getting our subscribers more leads obviously should increase retention rates but its hard if not impossible to measure.
If we did decide to stop advertising because we couldn't prove the value was there, then we'd be losing out hugely to our competitors in the same space who are blowing out our search terms on adsense and consistently showing up in the #1 paid spot. We know we are losing market to them but its very hard to say how much exactly. We are concerned that if we allow this to go on for much longer that it will really erode our business. Hence we are considering additional spend to curb this erosion even though it is difficult to measure. But how much spend is the question.
One idea would be to shift our pricing model so that we charge per lead, which would make it easier to track whether or not the spend was paying for itself, but we aren't sure we want to do this just yet.
A second example woud pertain to SEO, buying inbound links to increase SEO relevancy. You dont really know how many links will bump you up to that next spot in the SERPS, nor can you even accurately estimate how much money an additional spot up will mean for your business w/o taking a wild guess. So, how much to spend?