Bid management

A bid management is a software for automatic control of the commandments in search engine marketing (SEM). With a bid management tool, any number of search terms (keywords) can be managed via various paid search providers (e.g. Google Adwords or Yahoo Search Marketing ). The systems automatically transfer the bid changes to the respective channels, such as B. Google AdWords, transmitted via API . The bid management should help to determine the optimal bid for each keyword and to continuously adapt it. Most systems use more or less complicated algorithms. For the user, automatic bidding saves time and usually increases performance.

Most bid management tools also offer functions for easier management of SEM campaigns. These include, for example, the areas of keyword expansion, search query analysis and ad text suggestions, but also the monitoring of the landing page or brand protection. Often, bid management solutions are supplemented with campaign management tools. The customer is given the opportunity to carry out intensive analyzes via a login area and to transfer the resulting consequences directly to the campaigns and search networks, for example pausing a so-called AdGroup on Google AdWords.

A trend in current systems is towards expanding tools in the tracking area . More and more tools can not only measure the clicks that took place directly before a conversion , but also the process (funnel) before it, as well as other channels such as display or affiliate. Since the channels sometimes strongly influence each other in terms of performance, this creates better transparency with regard to the evaluation of the performance of the individual advertising channels.

Use

The success of advertising in search engines depends largely on the price that a company wants or has to pay the search word marketer for a click on the ad. This is why the billing model is also called pay-per-click . The cost-per-click prices are determined by a bid system for each search term. The more extensive SEM campaigns are (these can contain hundreds of thousands of search words), the more difficult it becomes to optimize the bids so that the target values ​​are met. Target values ​​can be the net profit, but also the cost of a purchase, i.e. H. Cost-per-order (CPO), or the costs for a new customer contact, i.e. cost-per-lead (CPL). Bid management tools help to find the statistically reliable, optimal bids on the basis of automatic analyzes.

Objectives of bid management - bid strategies

Bid management can pursue various goals that the user can specify in the system:

Click maximization with a given budget

The system tries to buy clicks as cheaply as possible. This is done by redistributing the keywords from which clicks are obtained. For expensive keywords, the bids are lowered and for cheap keywords an attempt is made to generate even more clicks via a higher position. This strategy is suitable for anyone who does not want or cannot take into account a quality difference between different keywords. Usually these are smaller companies and online shops that just want to get traffic to their site. The conversion behavior, for example whether a purchase is made, is ignored with this strategy.

Conversion maximization

When maximizing conversion, not only the cost of the clicks is taken into account, but also the quality. This must be measured using conversion tracking . This tracking is offered both by the paid search providers themselves and by the bid management tools. The implementation of the conversion tracking code can be done by the webmaster . The key figure for evaluating quality in this strategy is the cost-per-order (CPO). It states how many costs are incurred for a conversion (sale, download, registration, etc.). As a result, the keywords with the lowest CPO are the most attractive. The bid management tries to buy as much traffic as possible using these keywords. The optimization process takes longer because now, in addition to the clicks and CPCs, the conversions also have to be taken into account, the number of which are usually much lower than the clicks. The system therefore has to wait longer for data.

Profit maximization

This strategy is also based on conversion measurement. However, the parameter for measuring quality is not the number of conversions, but their value. This can be transmitted with all common conversion tracking systems. Some tools also allow comparison with the customer's inventory control system so that an exact contribution margin can be determined per order or conversion. For the customer, this allows the most precise form of optimization, provided that he knows the customer lifetime value . However, the volatility of the data is highest here, especially if the order values ​​for each keyword vary greatly. Customers with little conversion data are therefore more likely to resort to conversion maximization.

Constraints

In addition to these indirect goals, direct goals can also be specified for almost all systems. These are u. a. Desired position in the respective search network, maximum or minimum bids, compliance with the firstPageCPC and AdScheduling.

functionality

Rule-based bid management systems

Rule-based systems make their decisions regarding a bid change based on rigid rules. These rules can be predefined by the manufacturer or specified and expanded by the user (see bidding strategies ). A combination of different rules with regard to the individual target parameters such as CPO, CPC, position, conversion rate, time, etc. enables a certain adjustment to your own needs. However, the changes in the pay-per-click channels are very high. Therefore, the rules must always be checked and adapted very frequently. One therefore often speaks of "half" bid management tools. One task of modern systems is to find out these rules independently and to continuously improve them. The problem of data scarcity is very high here, because in the long tail , for example, certain rules cannot be applied in the event of missing conversions or a non-existent historical conversion rate. Known systems with a rule-based approach include: a. Aquisio and DC Storm.

Portfolio-based bid management systems

Portfolio-based systems try to increase the degree of optimization through theoretical models, so that not every keyword is subject to rigid limits or rules, but the influence on the overall performance of the portfolio is decisive for the regulation of each individual keyword. Individual keywords can exceed certain limit values, which can lead to an increase in the performance of the entire portfolio if other keywords are in turn adjusted accordingly. This is based on the Markowitz model for portfolio optimization, which is particularly widespread in the financial world and has a significant influence there on the selection of stocks in funds (see portfolio theory ). It makes sense to transfer Markowitz's considerations to the PPC world and thus achieve an increase in performance. However, this model requires a lot of information or data per keyword in order to function correctly. In practice, however, these are usually not available and when they are, they are subject to constant change. Thus, the points of criticism mentioned in the article portfolio theory also apply to bid managements of this type: u. a. inaccurate forecasts as they only relate to historical data and some underlying assumptions are not practical. Well-known systems with a portfolio-based approach include: a. Adobe AdLens and Adspert.

Evolutionary bid management systems

An attempt is made here by constantly adapting the algorithm and the internal rules to adapt the system to the constantly changing framework conditions and thus to continuously increase the performance. The knowledge gained in this way flows into a combination of portfolio-optimized and rule-based approaches. On the basis of an evolutionary algorithm, there is a selection of good rules and settings as well as a suppression of non-functioning approaches. Combined with portfolio optimization, its advantages are used for global optimization of the account, but at the same time it prevents excessive reliance on historical, outdated data and rules. The decisive factor here is constant backtesting to verify the results, as has also been established in the financial world. In this way, the interplay between bid, position, CPC and clicks can be constantly adapted to changing competitive situations and click behavior. On this basis, the systems of this type try to map the entire bid landscape for each keyword and in this way determine the optimal bid. Well-known systems with an evolutionary approach are Adspert, Kenshoo, Refined Ads and IntelliAd.

criticism

The usefulness of a bid management system is often questioned. The task of optimizing the bids seems too difficult and complex. Especially against the background that in practice much more than just the pure data has to flow into the optimization of the bids. Rather, it is also about anticipating which search words will be the top performers in the account in the short and medium term. A SEM manager can use keywords e.g. B. offer the desired position before a certain event. He can act at any time, while bid management systems only react - and only as good as the data on which the regulations are based. If the data is small or fluctuating, bid management can even work counterproductive depending on the quality of the algorithm. Because where the SEM manager can rethink conclusions, the system acts strictly according to the given pattern. Is an article z. B. sold out, then the conversion rate will collapse. Ideally, the SEM manager knows this immediately, the bid management must first "experience" this on the basis of the data.

The use in the long tail is also controversial. By definition, there is not enough data per keyword here to make a statistically significant statement in a sufficiently short time. However, this is also the case with the SEM manager. In practice, the algorithms of the bid management systems should have an advantage here thanks to in-depth calculations. However, the SEM manager can also improve his results by clever combinations. However, this costs him a lot of time in the long tail, which will affect the short tail performance. Against this background, it seems reasonable to assume that a well-optimized account needs both; bid management with good algorithms and a SEM manager who always has control and an overview.

Trends

• Conversion tracking system

More and more bid management providers are building their own conversion tracking systems into their systems. This increases the amount of data and thus the security with which the correct bid decisions can be made.

• onSite tracking

Providers such as B. intelliAd or Refined Ads add their own on-site tracking to their data, so that the bounce rate can be included in the decision-making process for keyword performance.

• Phone tracking

Information from the telephone area, such as B. the number of calls triggered per keyword can be taken into account via a connected telephone tracking system.

• Social media and display advertising

Various providers such as Efficient Frontier and Marin Software are expanding their systems in the direction of advertising and social media.

• Intraday bid management (optimizing several times a day)

This is a way of making even better use of the purchasing behavior of users, as bids can be optimized several times a day. In this way, the advertiser can get even more profit from his campaigns through intraday bid management. One of the driving providers seems to be Adspert.

• Bid management with Google AdWords scripts

In October 2012, Google published the AdWords Scripts for AdWords. This is a programming environment with JavaScript integrated in Google AdWords , as it is already available in Google Apps . This enables you to program your own bid management processes without having to use the AdWords API or complex bid management systems. Google also offers its own scripts and examples that exemplify bid management processes. In the future, this fundamentally calls into question the existence of external bid management systems.

Bid management with Google AdWords scripts also enables smaller advertisers to get started with bid management, as it can now be implemented directly and individually in AdWords itself at significantly lower prices and without having to buy servers or software. Until now, bid management has been the domain of advertisers with large budgets due to the sometimes high fees for the systems. The AdWords scripts are an alternative for these as well.

• TV tracking

Since 2013 there have been approaches to measure the effect of TV advertising on online and especially SEM campaigns. This determines how the traffic behaves after the TV commercial has been broadcast. The additional traffic is assigned to the respective channel or advertising material based on statistical analyzes. In return, this data can be used again for bid control.

literature

• Ashish Agarwal, Kartik Hosanagar, Michael D. Smith: Location, location, and location: An Analysis of Profitability of Position in Online Advertising Markets. , 2008 abstract online
• Animesh Animesh, Vandana Ramachandran, Siva Viswanathan: Research Note - Quality Uncertainty and the Performance of Online Sponsored Search Markets. An Empirical Investigation. In: Robert H. Smith School Research Paper , No. RHS 06-019, NET Institute Working Paper , No. 05-27, 2009 Abstract online
• Google, Anne Beuttenmüller, Thomas Bindl: 2009 Refined Labs case study - long-tail keywords , 2009 PDF, 1032 kB
• Animesh Animesh, Siva Viswanathan, Ritu Agarwal: Competing 'Creatively' in Online Markets: Evidence from Sponsored Search. In: Robert H. Smith School Research Paper , No. RHS 06-064, 2010
• Eva Gerstmeier, Tanja Stephanchuk, Bernd Skiera: An analysis of the profitability of different bidding heuristics in search engine marketing. , 2009
• Anindya Ghose, Sha Yang: An Empirical Analysis of Search Engine Advertising: Sponsored Search in Electronic Markets. , NET Institute Working Paper , 2009 Abstract online
• Anindya Ghose, Sha Yang: Modeling Cross-Category Purchases in Sponsored Search Advertising. , 2010 abstract online
• Avi Goldfarb, Catherine E. Tucker: Search Engine Advertising: Pricing Ads to Context. , NET Institute Working Paper , No. 07-23, 2009
• iProspect: 2009 Search Engine Marketing and Online Display Advertising Integration Study. , 2009 PDF, 238 kB
• Kinshuk Jerath, Liye Ma, Young-Hoon Park, Kannan Srinivasan: A 'Position Paradox' in Sponsored Search Auctions. In: Johnson School Research Paper Series , No. 36-09, 2010 Abstract online
• Zsolt Katona, Miklos Sarvary: The race for sponsored links: A model of competition for search advertising. , 2008 PDF, 322 kB
• Oliver J. Ruth; Randolph E. Bucklin: A Model of Individual Keyword Performance in Paid Search Advertising. , 2007 abstract online
• Oliver J. Rutz, Randolph E. Bucklin: From generic to branded: A model of spillover dynamics in paid search advertising. , 2008 abstract online
• Hal R. Varian: Position Auctions. In: International Journal of Industrial Organization. , Vol. 25, No. 6, 2007, pp. 1163-1178 PDF, 344 kB
• Sha Yang, Anindya Ghose: Analyzing the Relationship between Organic and Sponsored Search Advertising: Positive, Negative or Zero Interdependence? , 2009 abstract online

Web links

• Bid management: guarantee for successful search engine marketing?
• Why good bid management needs a web analysis tool (PDF file; 81 kB)
• Technology of bid management systems
• TV tracking in combination with bid management

Individual evidence

1. ↑ "Bid management software cannot replace humans" ( Memento from August 18, 2010 in the Internet Archive )
2. ↑ http://www.kammerath.net/adwords-scripts-bid-management.html
3. ↑ https://developers.google.com/adwords/scripts/docs/examples/complete-scripts?hl=de
4. ↑ http://www.internetworld.de/onlinemarketing/marketing/mattscheibe-internet-281666.html