The Economics Of Search - Economists working for the major search engines? Absolutely, and they try to predict where the search economy is going. This session looks at some of their work in predicting and modeling the space. Moderator: Chris Sherman, Executive Editor, Search Engine Land Q&A Moderator: Mike McDonald, WebProNews
Speakers: Peter A. Coles, Assistant Professor of Business Administration, Harvard Business School Michael Schwarz, Marketplace Designer, Yahoo! Research Mark Mahaney, Director, Internet Research - Citigroup Investment Research Hal Varian, Chief Economist, Google
Michael is up first. He's a marketplace designer at Yahoo.
There are 2 questions: How do we balance interests of users, advertisers, and the search engine? The key areas of economics of sponsored search and where the market is going.
There's no tradeoff between revenue and satisfying advertisers and users. Maximizing value delivered to users and advertisers approximately maximizes search engine's profits (it's a theorem). Do interests ever conflict?
What's the future of sponsored search? In the short run - the main competitor of Google and Yahoo in sponsored search - organic listings. If search continues to dominate the sponsored search results, the business model will have to change. Major challenges: making sponsored search more relevant, better measures of user experience, better advertiser tools, and more targeting and smarter advanced match.
Convergence of search and display advertisement: Search is about current intent and display is about demographics. Search for direct response display for branding. Search is spot market display contracts. Historically, search was selling for $0.05 a click and up - display for $5000 and up. The dynamics haven't changed much. Different buyers and different sellers. Convergence is happening as we speak.
Hal Varian, chief economicst at Google, speaks next about how to bid on AdWords.
Choose your keywords - the most popular keywords are not necessarily the most cost-effective. Use the keyword suggestion tool - it is based on real search behavior. Look for long tail keywords: you only play for clicks. More than 50% query strings occur at most once a week. What is the user looking for? Searchers are sophisticated: SKU searches, etc.
Costs per conversion: Note: conversion = acquisition = sale
Value per click and cost per click.
Determine the bid: Profit = value per click * number of clicks - cost of clicks. - vx - c(x) If you increase your bid from b to b' - clicks go from x to x', cost goes from c(x) to c(x')... wow too much math. Then he shows an example. There are way too many numbers for you to understand this. Heck, I'm sitting here and I don't even understand it.
Budget: if your value per click exceeds your incremental cost per click you are making money on each additional click. Wh y do you want to stop? Why do you want a budget constraint at all? If you are actually running into your budget constrain you are doing something wrong. Why? Because by lowering your bid you will get more clicks for the same total cost, so your profit will increase. However, it may make sense to have a budget constrain to avoid surgest in clicks.
Summary: - Use the keyword tool. The more keywords the better. - Use conversion tracking and analytics to estimate value per conversion, conversion rate, and value per click - Determine bids that maximize profit by bidding up until your value per click is greater than your incremental cost per click. If you persistently hit your budget, think of reducing your bid.
The next person up is Peter Coles from the Harvard Business School.
There are 2 sides of the marketplace. Neither wishes to join the marketplace unless there's signifcant representation of the other side. The two sides: advertisers and publishers. It's difficult to enter the platform when other platforms already exist.
The term is multihoming. Can you as a participant join more than one platform at the same time? Publishers: generally not. Which ads do you want to put on the parts of your page that get the best clicks? You can't typically put one network of ads. You can't multihome. Advertisers, however, can sign up for multiple networks. There's a tradeoff though - the reach that you get from the network compared to the costs of signing up for that network. There's a challenge for new networks to enter.
Entry: chicken and egg problem. Mobilzing a network is particularly challenging when there are strong cross side network effects. No publisher wants to be the first one to join the new network that enters. It's hard for an ad platform to make an emergence. Nobody wants to be the first.
Approaches to entry: - Incremental growth? - Big bang (buy another ad platform, maybe) - Using a flagship tenant - focus on one side. - Innovation: double edged sword.
Segmentation: an approach to entry - Focus on a specific market segment (vertical) - Example is stubhub (vs ebay) - Example is technology (Digg.com - now that Digg is going broad, start your own site. Chances are, I'll use it.)
Cross subsidization - Make deals with low profits or margins now to produce benefits elsewhere within the network. Microsoft could outbid to gain third party sites because increased reach might add value to existing sites.
Takeaways: - Strong cross side netowrk effects in ad platforms make entry challenging - Segementation as an entry approach - Differentiation, usually a powerful competitive force - can be a double edged sword.
Last up is Mark Mahaney of Citigroup.
US Internet Advertising Outlook - forcesating 22% year to year growth in net advertising to $126 billion in 2008 E-retail - forecasts 16% to $158 billion in 2008 Online travel - 14% to $106 billion in 2008
Now switch over to search. In 2008, about $11 billion is spent on search advertising. We're in an interesting environment; there's a slowdown in growth. We're seeing an impact in the financial community. What we've seen is that amongst the Us major search engines, Google's share has declined. This may have been one of the reasons why MS bid for Yahoo. Search budget share > search query share. Google's share of budget could exceed the search query share.
How big could mobile search be? 35 monthly searches per installed PC. Mobile search is a fascinating opportunity. In some point, 1 search per handset = $2.5 billion. This is the power of one. 10 monthly searches means mobile search will exceed PC search (by 2010).
Microhoo = it's hard to see this Microsoft/Yahoo deal not happening. Our bookies are all betting on it. Microsoft has to bid higher. This is one of their few opportunities to try to compete effectively with Google. They made this one of their top 3 growth priorities in 4 years and nothing has happened. and this is the only game-changing move that can happen.
3 questions: 1- Will MicroHoo lead to a search query shift among consumers? Will searchers change thier search habits? 2- Will the MicroHoo merger create a better search engine? 3- Could a larger #2 search engine draw search budgets away from Google?
