Search is a tremendously expensive game to try and win. It's economics are such that the more search share you have, the more money you earn per search. This is a critical point, so I'll spell it out a little further.
All search companies have more advertiser dollars than they have searches to spend them on, Google, Yahoo and Bing included. It's a supply constrained marketplace. The search ROI is so good for advertisers that they all want to spend more money at their current CPCs, but there aren't enough searches. This supply constraint leads to a problem for the smaller players. Search revenue is driven by having lots of advertisers compete in every auction. The larger the share, the more clicks each advertiser will get, and thus the more advertisers you attract. The smaller scale players don't drive enough clicks for some advertisers for it to be worth their time to set up and manage campaigns on them, while the larger scale players it is worth their while (the return they get exceeds the fixed cost of advertising in the marketplace). So with fewer advertisers, there are fewer bidders in the 2nd price auction, and the revenue per search is lower for the smaller scale players.
So how does this apply to Microsoft's online division? Well, if they want to catch Google, they're going to have to do it at a scale disadvantage, meaning that Google is going to make more off of the same searches than Microsoft will simply because they have a bigger marketplace. To beat that, Microsoft has to commit to spending lots of money to try and close that scale gap by buying share through distribution deals and spending a ton on technology to differentiate the search product while accepting that they don't monetize the searches they do have as well. If they can eventually build a product that will pull enough marketshare from Google to be roughly equal, then they should start to see better monetization.
The valid questions are: 1. Is it possible to catch Google? Or are the market dynamics such that without a transformative difference in how the product works that Google will never be caught.
2. If it is possible to catch Google, how much money will you have to spend, and what will your eventual ROI be when you get there.
Since Microsoft is a company that does 60B in revenue, it has to look at big businesses to drive a 10% growth in that revenue. Your hot little startup that does $100MM doesn't make a dent. Even Facebook only does 1-2B, depending on which report you believe. Search is a 10B going to 20B market, and if Microsoft can spend 5B over 5 years to get half of that market and earn 10B every year it's worth it.
Of course, the division has been horribly mismanaged for years. Qi Lu now runs it, and he's a different breed from most Microsoft execs. So time will tell if it's a good bet or not for Microsoft.