The cost of doing #1 is coming down thanks to inexpensive, professional level DSLR cameras and cheap editing software like Final Cut Pro. Sure, the real expense is hiring the talent to use these tools effectively, but as the tools become less scarce the ability to tell the difference between pro and amateur work is getting harder and harder. The cost of #2 is also coming down if you consider how inexpensive online advertising is. It will still cost a fortune to run a commercial on TV or run a full page ad in Opera Magazine, but those are no longer the only options (and arguably the worst ones). Free can get you pretty far online via YouTube and social media if you have something that is unique and worth sharing. If you want to pay for exposure you can also get pretty far on a limited budget.
So if #1 and #2 are cheaper than ever before is #3 getting cheaper too? It doesn’t look like it, but it should be.
I agree with Seth Godin that when execution gets cheaper, so should planning (and therefore the cost of working with an agency should drop), “The cost of building digital items is plummeting, but our habit is to plan anyway (because failure bothers us, and we focus on the feeling of failure, not the cost).”
If you’re building an online marketing campaign, spend less money on the meetings, sign-offs and planning and invest that money into learning and optimizing. Instead of paying to figure out which idea is best before picking the final winner, invest that money into running them all and optimizing the ones that work best. I think you should publish it all and then let the best filter to the top, more content is worth more than perfect content.