Initial Keyword Bidding for Search Engine Advertising
In a sponsored search auction, the search engine seeks to maximize its own expected revenue by balancing bid price with ad relevance. This relevance is captured in the Quality Score (QS). Ad position (rank) is determined by the weighted bid (Bid × QS), with ads displayed in decreasing order of this value. The "second-price" payment rule is the mathematical core of this system. Under this rule, the price an advertiser pays is the minimum amount required to maintain their current position, calculated as the next highest weighted bid divided by the advertiser's own Quality Score.
In Search Engine Advertising (SEA), the "cold-start problem" represents the primary structural barrier to entry for new campaigns. Unlike mature programs that optimise based on historical performance, new accounts operate in a data vacuum, lacking the empirical evidence required to determine which keyword ranks maximise profit. This often results in inefficient capital allocation, where advertisers oscillate between overpaying for top positions or underbidding and losing high-intent traffic.
Traditional data-driven bidding fail during the launch phase due to two core challenges:
1. Limited Bid Variation: Established accounts typically change bids infrequently, preventing the observation of a complete bid-to-rank curve.
2. Endogeneity of Bids: Live auction bids are frequently correlated with "random shocks" (e.g., demand spikes during specific weather events), leading to biased performance estimates.
Without historical data, it is nearly impossible to accurately estimate the metrics required for profit-maximizing bids. These metrics include:
• Profit contribution per keyword (PC): Revenue per keyword multiplied by the profit margin.
• Conversion rate (CR): The percentage of clicks that result in a sale or lead.
• Percentage increase in prices per click (PPC): How sharply costs rise as an ad moves into better (higher) ranks.
• Percentage increase in clickthrough rates (CTR): How strongly the number of clicks increases in better ranks.
Intuition dictates that higher profit contributions and conversion rates allow for higher bids. However, if the cost to move up one rank is excessively high compared to the resulting increase in clicks, the advertiser should bid lower to maintain profitability.
Advertisers lacking the resources to develop software for the Exact Bidding Approach using data on PPC and CTR increase should utilise the following for effective bidding:
Bid = Profit Contribution per Conversion × Average Conversion Rate × 50%
It is highly effective and requires only the estimated profit contribution and a baseline conversion rate (suggested at 1% for Western markets).
Multipliers for price increases do not vary significantly across industries, but they do vary based on keyword intent and degree of competition. Bidding strategies should be tailored to search intention rather than industry benchmarks. Advertisers must manage SEA and Search Engine Optimization (SEO) in tandem. User intent and organic presence both influence ad performance; thus, budget allocation should reflect the dominant search intention of the target keywords. Using tools like Google Keyword Planner provides a "neutral" view of the market that avoids the biases found in an advertiser’s own historical data, especially regarding price sensitivity and competitive density.
Search intention, presence of organic search results and Search Engine Advertising
The position of an advertisement improves (moving higher up the page) as search intention shifts from informational to transactional, and finally to navigational. The likelihood of a user clicking an advertisement increases significantly as the search intention becomes more specific and deliberate (from informational to navigational). Conversely, the cost per click decreases as search intention becomes more specific, likely because more deliberate queries face less competition. The conversion rate is significantly higher for navigational keywords compared to informational or transactional ones. Similarly, the number of signed contracts is higher for transactional and navigational intentions.
When an advertiser’s organic result appears on the same page as their advertisement, the advertisement receives considerably fewer clicks. Despite receiving fewer clicks, the presence of an organic result increases the conversion rate and significantly decreases both the cost per conversion and the cost per contract. The number of ad impressions and the advertising rank are independent of whether an organic result is present.
Theory and practice must move beyond simple keyword management to include search intention and organic presence as core factors in evaluating and adjusting SEA campaigns. Advertisers can more effectively reach their objectives by segmenting keywords by intention and assigning budgets based on the likelihood of those segments to generate revenue (e.g., focusing on transactional and navigational queries for immediate sales). Search engine advertising should be viewed as a portfolio of different intentions. Advertisers must balance high-volume informational keywords (which build awareness) with lower-volume but high-performing navigational keywords. Advertisers should use Search Engine Optimisation (SEO) to influence organic presence, as this synergy improves conversion efficiency and lowers the cost of acquiring customers, even if it "cannibalises" some paid clicks. Advertisers need to evaluate when they must bid on their own brand keywords (navigational) to defend their advertising space against competitors, while also considering if those ads are simply taking clicks that their organic results would have captured for free.
Overview of online advertising research findings
• Varying Effectiveness and Synergy: Online advertising generally produces positive returns, though its impact varies significantly by product category, customer segment, and ad format. While cross-media synergy is typically positive, the effectiveness depends heavily on the sequence of exposure across different channels.
• Dual Mechanisms (Engagement vs. Mere Exposure): Advertising works through two primary paths: high engagement (e.g., clicks and emotional involvement), which is most effective for formats that command focal attention like mobile apps, and mere exposure, which can improve brand attitudes through implicit memory even when users are not paying active attention.
• The "Double-Edged Sword" of Creative Elements: Attention-getting devices like large sizes or animations enhance brand recognition, but they also increase perceived intrusiveness and annoyance. Effective emotional appeals tend to be positive, high-arousal, or feature complex humour.
• Context and Location Relevance: The benefit of ad-context congruence is nuanced and depends on factors like ad position and consumer goals. In mobile advertising, ads targeting relevant locations are significantly more effective, and consumers in the same physical location tend to respond to ads similarly.
• Personalization vs. Privacy: Personalization improves effectiveness by increasing personal relevance and reducing doubt, particularly during early information-seeking stages. However, it triggers significant privacy concerns because consumers are often unaware they are being tracked.
• Search Advertising Dynamics: CTR is higher for specific, less common keywords and keywords containing the advertiser's name. While top positions gain more clicks, the overall cost increase may not be offset by the gains, making top-rank bidding risky.
References
AbouNabout, N., 2015. A novel approach for bidding on keywords in newly set-upsearch advertising campaigns. European Journal of Marketing, 49(5/6),pp.668-691.
Liu-Thompkins,Y., 2019. A decade of online advertising research: What we learned and what weneed to know. Journal of advertising, 48(1), pp.1-13.
Schultz,C.D., 2020. Informational, transactional, and navigational need of information:relevance of search intention in search engine advertising. InformationRetrieval Journal, 23(2), pp.117-135.
