I'd like to take this opportunity to make the unusual move of writing in accord of one's thoughts.
Last week Cory Treffiletti talked about the emergence of what appear to be two flighting possibilities for online media: daytime daypart, concomitant with at-work usage; and holiday seasonality. He suggests that these facts might have impact on the effectiveness of the online media and this should inform how it is that budgets are spent online. He's absolutely right.
You see, for a long time now the argument has been made, most vociferously by publishers in the early days, that time of day or day of week didn't really have any bearing on the value of the inventory being sold. An impression is an impression is an impression no matter what time of day or day of week that impression is encountered.
The argument is one born of a direct response perspective and the way online media works, technically. For one, the idea is that if you are there to be in receipt of my message, then that is valuable irrespective of time or day because that reception is my opportunity to sell. Secondly, the way online media works meant that if you were engaging content on a particular property at 4AM, an ad was served to you, that impression just as valuable as any other because YOU chose to be there at 4AM in an active way.
Problem is, this doesn't take into account what Cory pointed out as the "decision cycle" of the target; perhaps I'm not thinking about weekend getaways on Monday, for instance.
When I used to buy radio for a large beverage CPG client, we always had our radio run Wednesday through Saturday, because those days were held to be most in propinquity with the grocery shopping occasion. Ad messages run on other days were deemed wasteful.
Also, the approach of buying blind impressions as if all impressions are created equal assumes that individual members of an audience are created equal. That is to say, what kinds of persons are likely to encounter your advertising on one particular day of the week - or at some particular hour of the day - than on others?
The holiday seasonality Cory points to makes a good example for this point: someone is checking out CBSMarketwatch.com on Christmas Day. Yes, that person is a member of the audience the advertiser is interested in, but the audience the advertiser is interested in is only an aggregate of an idea of a specific potential consumer. That is to say, the general audience of CBSMarketwatch might be exactly who an advertiser is looking for, but the particular person who is on that site on Christmas day may not be who the advertiser is looking for at all. Think about it. The market ain't even open on Christmas.
Of course, what this example also points to is a serious need to segment a site's audience to find out just from whom that audience is really composed. This information in concert with flighting would be enormously powerful.
Before you go down the road to flighting, it would be best to first run a test buy or two for your client before figuring out which days are "most active" and appropriate for your client and product. But once you do, this can be another way to increase efficiency of the direct responses you get as well as the possible "relevance quality" of the impression that merely, well... makes an impression and leads to correlative activity that translates as product sold.