Optimizing Google AdWords for Non-Profit Organizations: A Case Study, Study notes of Marketing

Download Optimizing Google AdWords for Non-Profit Organizations: A Case Study and more Study notes Marketing in PDF only on Docsity! Google AdWords: A Guide to Acquiring, Optimizing, and Maintaining an Effective Web Advertising Campaign By Peter Anderson and Ben Brookins Presented to the Department of Economics, University of Oregon, in partial fulfillment of requirements for honors in Economics. Under the supervision of Prof. William T. Harbaugh June 2008 Google AdWords: A Guide to Acquiring, Optimizing, and Maintaining an Effective Web Advertising Campaign Abstract: Our goal for this project is to provide a model by which non-profit organizations can use Google AdWords online advertising to increase visibility and, ultimately, donations. By examining current research on the subject and looking at select case studies, we have arrived at a practical and theoretical solution that requires little maintenance and can be applied to nearly any organization with ease. By following these guidelines, we believe a non-profit organization can increase its visibility – at nearly no additional cost – while achieving its ultimate goal: increasing donations. Approved: __________________________________________________ Prof. William T. Harbaugh Date 3 volunteer hours. It competes with both local and national food banks; its main competition comes from the United Way. Though they have had success increasing donation figures for the past few years independent of Web advertising, they have expressed interest in advertising more effectively. It is our shared belief that an online ad campaign through Google AdWords will not only increase local visibility and donation totals, it will allow for relevant demographic data collection, further enhancing FFLC’s knowledge of its local market and eventually permitting more targeted advertising methods. Food for Lane County is still in the early stages of this process, awaiting organizational approval to apply for an advertising grant. Before we go any further, it is important to understand what Google AdWords and a Google grant are. Google AdWords are sponsored advertisements that are displayed during a Google search. See below: 4 The advertisements in the boxes labeled Sponsored Links, both above and to the right of the standard search results, are the “AdWords” themselves. Specifically, they are combinations of search phrases that an organization pays Google for in order to better advertise its products or services. The idea behind this advertising is to attract the attention of Web searchers looking for specific search terms and visit a site that they otherwise would not have known about. For example, the search above is for the iPod music player. Notice that Apple, the company that makes the iPod, has paid for an advertisement, but so have many distributors of the product: Circuit City, MacMall, eBay, AOL Shopping, Overstock, and Amazon. This is an example of how AdWords can effectively increase awareness (and ultimately sales conversions) for its users. Standard AdWords accounts must be paid for, including a small startup cost and a cost each time an ad is clicked on, but each year, Google gives out numerous AdWords advertising grants to non-profit organizations that apply for one. These grants allow the non-profit to use the AdWords program for free, though the amount they can spend is limited. The grant application process is relatively simple; an explanation of the Google Grants program can be found in Appendix B. Other Google Programs Now that we know, in general terms, what the AdWords program is and how it can be used, it should be noted that we used other Google programs and applications for the duration of our project. For the sake of familiarity, we will summarize what those programs are, and how we used them, before we move on. 5 Google Analytics – A Javascript-based tracking script that allows Web administrators to track visitors to their site by a variety of criteria, including information on page views, time and duration of visit, number of pages seen, number of new visitors, how visitors got to the site, and so forth. It also collects demographic information on visitors by geographic area, language, operating system, Web browser, and other relevant factors. Within a given geographical area, visitors can be tracked by country, state, county, city, or other specified geographic region. We recommended adding an Analytics tracking script for the purpose of data collection, with the eventual hope of compiling demographic information on donors. Google Checkout – Google’s financial transactions software, Checkout processes credit card transactions in a secure manner, much like PayPal or other secure transaction programs. A Checkout button can be easily created as HTML code and added to a Web page; when clicked, the visitor is taken to the checkout phase of the transaction. We encouraged the use of Checkout in order to offer donors a safe, secure, and simple way to donate to a charity, as well as for the purpose of linking demographic information with donor information for a more complex data analysis. For a complete list of all Google Programs relevant to this paper please see Appendix A. Literature Review While most of the research here is highly applicable to any project in which marketing objectives are to be met, we will summarize the conclusions of previous research in the body of this paper. For a more in-depth examination of the following research, please refer to Appendices E and I. 8 In Advertiser A’s case, their bid of 40 cents and their Quality Score of 1.8 give them an Ad Rank Score of .72. Advertiser B, despite having a much larger CPC bid of 65 cents, falls into the second position – this is a direct result of the Quality Score being significantly lower than Advertiser A. Conversely, Advertiser C has a higher Quality Score than Advertiser B, but has set its minimum bid too low to compete for a higher ranking. The actual CPC paid is the amount an Advertiser has to pay in order to maintain their current ad position. Let’s look at Advertiser A, whose actual CPC is 37 cents. If Advertiser A were to bid anything below 37 cents (say, for simplicity, 36 cents), the Ad Rank Score would fall to .36 * 1.8 = .648, dropping them below the Ad Rank Score of Advertiser B (currently .65) and lowering their ad position. Similarly, if Advertiser B were to lower their bid to anything below 38 cents, Advertiser C would take over the second position. Please note that Advertiser C, with no other competitors for ad space, pays the minimum amount possible (1 cent) for their position. Should another competitor come along, that cost would increase, but in this simple model with only three bidders, they pay only 1 cent. Users have a budgetary cap on both CPC expenses and daily expenditures, at which point their ads will stop being shown until midnight of the next day. For non- profits, the maximum CPC bid for any search phrase is $1, and their daily expenses cannot exceed $330; for standard accounts, the limits are $100 and $250,000 respectively. It should also be mentioned that, though a user may not “win” the auction, their ads will still fluctuate up and down in terms of ad position over the course of a day. This happens for a variety of reasons: constantly-updating Quality Score calculations, 9 budgetary limitations, specific requests for ads to be shown during certain times of the day, and so forth. There also exists a minimum CPC bid for search terms, relating directly to a search phrase’s Quality Score. Phrases with a Great Quality Score can be bid on cheaply, usually between 5-10 cents, because they perform well and will make money for Google thanks to their high CTR and search volume. Keywords with an OK Quality Score must have a minimum CPC bid of around 30-50 cents; there will be decent search volume and a reasonable CTR, so a moderate bid will be sufficient. Phrase with a Poor Quality Score, by extension, are the most expensive, usually requiring a minimum CPC bid anywhere from 80 cents to $5.00, depending on how low its search volume is. These ads are almost never clicked on despite numerous Impressions, and thus Google cannot justify letting them be purchased cheaply. We will also mention ad text in brief, even though it is more of a marketing component and slightly beyond the realm of this paper. Each Google AdWords ad itself is made up of five components, as seen in the image below: 10 The five components – Headline, Description Lines 1 and 2, Display URL and Destination URL – all have short character limits, placing a premium on brevity and amplifying the importance of writing effective ad text (though this is a marketing-based plan, suggestions on the most effective ways to write ad text can be found in Appendix I). From a technical standpoint, ads that Google determines to be more relevant to both the keyword selection and destination site have a higher Quality Score than less-relevant ads. With this explanation in mind, we can add on additional research to enhance our understanding even further. Specifically, with the knowledge of how the program works, we can begin to understand ways in which AdWords can thus be broken down and optimized. Several papers suggest ways this can be done. A study by Feng, Pennock and Bhargava (2007) concluded that advertisers should use “a rank-revision strategy that weights clicks on lower ranked items more than clicks on higher ranked items,” which “is shown to converge to the optimal (maximum revenue) ordering faster and more consistently than other methods.” In other words, lower-ranked phrases are usually more specific to an organization and face less pressure from other bidders, so they will be far less expensive. Daniela Danciulescu (2007) examines a small business that sells “baby toys, clothes articles and accessories.” Like Africare, the goal of this business was to increase Web traffic and, subsequently, sales volume. It also sought to collect e-mail addresses and other information on prospective buyers and compile them into a database, which would be used to send out a company newsletter as well as special offers. Her findings reveal that: A) Negative keywords, or keywords that AdWords is told to ignore when 13 were extremely important in successfully marketing a non-profit: integrity, nurturance, sophistication, and ruggedness. These four characteristics not only affect the likelihood of contributing but also the amount contributed. In order to solicit more donations, it is advisable for non-profits to try to project these traits on both their Web site, and through its advertisements. There is also literature on the differences in giving between men and women, as well as how experience in charities as a youth may increase charitable donations. For more information about the characteristics of charitable givers, the reasons why they give, or more about any of the above studiesm please see the more extensive literature review in Appendix K. Finally, we examine simple ways to entice searchers by examining what they are most likely to respond positively to. A recently published article by Melissa Tooley (2008) found that online donations go up when even the simplest of changes are made (results seen graphically below): Increasing the size of the donation button and changing its color to red increased donations by 25% (and green buttons performed even better than red); using “polite” header text rather than forceful header text increased donations by 22%; and removing only the title and suffix boxes from donation forms increased donations by 31%. Not only are these changes easy to integrate, they are easily testable, as we will explain later. Results of Tooley’s Study (2008) Table 1.3: Email Landing Page Results Pe aaa Copy (Black) (Red) 125% “Polite” Header* + Appeal Call-to-action Header™* -22% Copy (Black) (Black) “Polite” Header* + Appeal Call-to-action Header™* 12% * Polite header: “Please make a tax-deductible gift today to stop the abhorrent practice of extraordinary rendition” followed by appeal copy. “* Call to action header: “Donate Now! Help us end extraordinary rendition!” Title and Suffix in Personal Info Form BE Cree Eee CU meet Mey No Title or Suffix in Personal Info Form International” Header seconds” Header SUBMIT 18% High to Low Gift String Low to High Gift String +6% No Intro Copy- Paragraph Intro Copy- Bullet Point Format Format 12% No “Donate to Amnesty “Donate in just 60 1% No Siler Chart 3.2: Search Landing Page Results eae +22% DONATE NOW! 13% aout ss4 cenrineates +6% No Funding source copy* No funding source copy +5% No * Funding source copy reads: “Amnesty depends on the generosity of individuals like you to fund our life-saving work. To maintain our impartiality, Amnesty does not accept any money from governments.” 14  15 Similarly, research done by Haynes, Thornton & Jones (2004) has found that advertisements that appeal to an individual’s sense of guilt (i.e. negative imagery) is significantly more effective at attracting attention than those that appeal to an individual’s sense of warmth (i.e. positive imagery). This is of particular interest to both Africare and Food for Lane County, as both are charities that try to combat issues in which negative imagery is bountiful – malnutrition, AIDS research, rampant poverty, and so on. The Application Process for Non-Profit Organizations Now that we know what AdWords is, how it works, whom the ads are targeting and why that targeting has been shown to be effective, we turn to our own case studies as empirical examples of everyday organizations that can benefit from the knowledge presented here. This process begins with the application for a Google grant. Though this step is shorter and much less involved than the next step, creating an effective AdWords campaign, it is crucial for a non-profit to get right – Google Grants does not deny very many applicants, but those that it does deny are denied forever. (For a review of the details of the application process, see Appendix B.) As stated, this is the step where Food for Lane County currently sits: previously unaware of the Google AdWords program on the whole, but highly interested in increasing its local advertising. As such, it has been encouraged by our research to apply for an AdWords grant. Once again, this process is fairly straightforward; we will assume throughout the rest of the paper that an organization already has an AdWords campaign, 18 Only when all of these components are considered can the goal of optimizing the campaign begin. As such, we will describe all four of these factors and explain how to create a uniform campaign, using our work with Africare as an example. Keywords Selecting the proper keywords is arguably the most subjective step of the four – and the most difficult to get right. Keyword selection varies depending on the type of organization, its focus, what points it wishes to emphasize or feels will generate the highest search volume, competition for the same phrases from multiple sources, and so forth. Like Feng, Pennock & Bhargava, we found is that organizations that emphasize specific search terms with lower search volume actually see increased traffic and lower costs than those that opt to buy common, high-traffic search terms. Selecting organization-specific terms, which face much lower demand than common terms and are more relevant to the site, will result in a lower CPC bid and improved Quality Score. Additionally, these specific terms almost always result in a higher CTR, furthering increasing Quality Score and lowering total CPC. Africare has, logically, purchased several search terms relating to the continent Africa. A snapshot of its “Africa” campaign can be seen below (sorted by CTR): 19 As seen here, its top nine search terms all involve giving aid to Africa. Its lowest- performing search terms all have little to do with the goal of the site itself and are, in most cases, too broad or generic. In the case of especially popular phrases, Africare’s ads are even Inactive for Search, meaning they will not show until the minimum bid is increased drastically (and in the case of a non-profit, impossibly) higher. Because these terms are only tangentially relevant to Africare’s site, they suffer from low CTR, high CPC and Poor Quality Score ratings. The following chart shows each phrase in this campaign and its respective Quality Score (sorted by CTR): Phrase Quality Score Help Africa Great Help Needed in Africa Great Giving to Africa Great Helping Africa Great Gifts for Africa Great Gifts to Africa Great Aid Africa OK Reaching out to Africa Great 20 Africa Aid OK “Africa” Poor Africa Culture Poor Africa People Poor Life Quality Africa OK Africa Agriculture Poor Africa Poor African Agriculture Poor Poverty in Africa Poor Faces of Africa Poor Africa Face Poor [Africa] Poor African People Poor African Culture Poor The conclusion is supported by the facts: every search phrase that directly related to the mission of the site was evaluated as either Great (seven phrases) or OK (two phrases), while the broad, less relevant terms were evaluated as Poor (12 phrases) and OK (one phrase). The rest of the data for each phrase follows a predictable trend: CTR is highest for the best performing phrases, and average CPC among specific phrases is lower than common ones (85.1 cents against 90.5 cents). The rest of Africare’s campaigns corroborate this story: every phrase in the “Africare” and “Donation” campaigns is evaluated as Great or OK, and each one is Active and currently being shown. Consider also the “Water in Africa” campaign. The search phrase “’African Well Fund’” has an OK Quality Score while the more general “’Well Fund’” has a Poor Quality Score. Though in this instance it does not negatively impact Africare – the latter search phrase has a lower CPC (50 cents against 70 cents) and a higher average ad position – it serves to show how a simple change in a search phrase to include the word “African” immediately increases the Quality Score by increasing ad relevance. As stated before, vague search phrases do not perform as well: phrases like 23 searchers who otherwise wouldn’t have known about them, a lower bid can lead to a chain reaction of positive events. To begin, there is a lower total cost. Ads with lower bids will fall in ad rank, but the benefit is that a click-through will cost less than before in terms of both actual cost and average CPC. What’s more, an unlimited budget for any organization is exceedingly unlikely, so the ones that set higher bids will earn top ad rank priority early in the day, but that will diminish over the course of the day as their budget depletes, pushing lower-ranked ads up in the rotation. In other words, patience will lead to about the same results over the course of a day with a lower cost. To summarize, if the daily budget is unmet, setting each phrase at the maximum CPC bid is the optimal strategy. Once the budget is met, though, CPC bids should be lowered with more common words being lowered before more specific words. Click-Through Rate Though this component has already been displayed in previous sections, it is important to discuss here due to its importance in ad effectiveness, as well as its impact on Quality Score. Overall, Africare’s CTR is not very high, sitting at 1.65% for all of its search phrases in all of its campaigns. While its best-performing keywords are not an issue here, the ratio of those keywords against poorly-performing keywords is low, which is primarily to blame for the low overall CTR and the relatively high average CPC – especially considering that the maximum CPC for Africare is $1. While CTR is an important all-around measurement, it is not something that can be directly controlled by an organization. Instead, it is affected indirectly by the other key factors, and it is our 24 belief that following the changes we recommend later in the paper will increase the overall CTR by a drastic amount. Quality Score Though it has been discussed at great length already, Quality Score remains the most secretive factor in AdWords, and the hardest to account for. The Google AdWords support database offers an outline of Quality Score computation but not much more; recall this information can be found in Appendix F. None of these descriptions come with any information on the weight assigned to each factor, nor does it offer insight into what the catchall term “other relevance factors” means. This list does, however, narrow down the prerequisites for a high Quality Score along the lines of ad text, CPC bid, landing site quality, historical CTR and account history, and keyword selection. In effect, future success compounds upon early success; getting a good start and maintaining a quality ad campaign will lead to high marks in all areas, which will only further improve the Quality Score and improve the value of the campaign. Ad Text Creating relevant ad text is a simple but effective way to help drive traffic to a given site; the more compelling the ad text, the more likely a click-through will be. While this paper will not attempt to look at this from a marketing perspective, we will say that mission-specific ads written with “short, punchy copy” have been shown to be the most effective in attracting searchers’ attention. 25 From the technical standpoint, relevant ad text helps improve Quality Score by establishing a more relevant link to the destination site. This can be a difficult process given the strict ad parameters (25 characters for the Headline, 35 each for lines one and two of the text and the displayed URL), but creating ads custom-tailored to specific keywords will ultimately be the most successful method of writing ad text. Africare’s current advertising methods follow the same formula for each campaign: two ad variations that relate to the overall mission of the site, and one that is campaign-specific. See the example below: The first two ads are the general ones that appear in every campaign, while the third is specific to the “Food” campaign. In most every campaign, the first two lead the campaign-specific ad in terms of clicks, impressions, and CTR. While this seems counter-intuitive on the surface, consider that the goal of this paper is to create relevant ads, not necessarily popular or attractive ones. Also consider that AdWords automatically displays better performing ads more often; with two ads that are seen in every campaign and higher CTR scores in most every situation, success compounds upon success and these ads are shown more often than the campaign-specific ones (this can be seen in the % Served column). 28 An Ideal Campaign We have examined Africare and its current AdWords campaign at great length, and we have explored methods of optimizing a campaign using a handful of simple and effective changes, so the final step (essentially) in this process is to help them optimize their campaign for maximum value and performance. Our recommendation to Africare on optimizing its campaign can be found in Appendix G; this section, though, will summarize the key points of establishing an optimal campaign from start to finish for other organizations in general. Apply for a Google Grant – With an Effective Application This is a step specifically for non-profit organizations wishing to get started with AdWords but without the budget to pay for a standard account. Besides meeting the basic requirements to be able to apply for a Google Grant, such as 501(c)(3) status and little to no political affiliations, it is important to write a good application for the Google Grants program. Although it seems that few charities will be rejected from this program, it is still important to write as quality of an application as possible to ensure the best terms of a Google Grant. Most of the application consists of fields that are meant to be filled in by the applicant; some of the fields are simple, like the main contact within the organization or its address. Other pieces of information are more technical and may require certain documents to make sure the company has everything it needs to fill out the application. Examples include: Employer Identification Number, annual operating budget, Web site traffic figures, and Google AdWords customer ID if the non-profit is already an 29 AdWords customer. The majority of the application, however, is a few essay fields that need to be filled out. The rest of this section will focus on the aforementioned fields. The first essay field asks for the organization’s Mission Statement. Most organizations already have a Mission Statement, and it is advisable to simply use that. There is a character limit on this field however, so if the Mission Statement exceeds this character limit it is advisable to shorten it while still addressing the key points. Ideally, it will coincide with Google’s primary objectives of helping the world in areas such as science and technology, education, global public health, the environment, youth advocacy, and the arts. The next section also has a character limit of 500 characters and asks how Google AdWords will help the organization, which is not a lot of room, all things considered, so brevity is key. Before starting the application, the organization should have outlined the goals it wants to achieve, including getting more volunteers or soliciting more donations. Briefly describing each goal the organization would like to accomplish and then explaining how Google AdWords will help achieve that goal is the easiest method. The next essay question asks for a target audience. This has a character limit of 150 characters, which is only one or two sentences, so hitting the key points – who are the donors, where are they from, and why should they give – is the optimal route here. The last parts involve advertisement writing and keyword selection. Much of this paper focuses on these two topics. The first step is to consider the goals the company decided it wants to accomplish through this campaign. If it wants to solicit more volunteers, make sure to suggest keywords that deal with volunteers; if it wants donations, it should use words that encourage donations. Perhaps the most important 30 thing, however, is to suggest words that relate well to the organization. As far as advertisement writing goes, repeating the search phrase in the Headline as well as the text of the advertisement has been shown to be effective. See our examples for Food for Lane County below; once again, to see FFLC’s sample application, please see Appendix C. Integrate Google Analytics and Checkout ASAP These two Google services are immensely important in evaluating the effectiveness of the AdWords campaign. Using Google Analytics and Google Checkout together would allow a non-profit like Africare to track not only which search terms are most effective and which lead to the highest amount of conversions, it will allow the collection of demographic information, including geographic location, contact information, and donation amount and frequency. Though Africare currently collects information on its AdWords campaigns, having this demographic information would be instrumental in moving forward with a more sophisticated data analysis of donor behavior and characteristics. It is advisable, then, that any group, whether non-profit or not, add these two scripts to its site in order to start tracking site usage data immediately. 33 underperforming terms that are not specific to the site and do not garner a high enough search volume will cause an increase in every major AdWords statistic in short order. Extensions Though our guide has been relatively comprehensive in terms of obtaining and maximizing the efficiency of an AdWords campaign, we have taken note of several areas that would act as a good jump-off point for further research. There were several obstacles in our research that prevented us from pursuing these opportunities. There is significant work to be done with, specifically, keeping track of donors and their characteristics, and custom-tailoring ad campaigns to them. The most important area in which we feel our project could be greatly improved is identification and tracking of donors. The primary obstacle in this case was the lack of data. With the Google Analytics software we are able to get geographic data on the visitors to the site, as well as their demographic information. By extension, we could use this as a proxy for the characteristics of those visiting a site, allowing us to determine what characteristics donors to Africare typically possess. This information will allow us to better target this audience with both ads and perhaps geographic-targeting. An unanticipated obstacle that we ran into in achieving this goal was the charity’s need to get the addresses and keep track of the donors using its own tracking system. This interfered with the use of Google Checkout, which would have been preferable as far as extracting the most amount of data from the process. That way, we could track people from search to amount of donation. However, both organizations we worked with did not want to move away from their current system. In Africare’s case, Sage Systems has 34 generated significant interest among Africare staff as a potential database management system. Sage Systems promotes a donor management system that, though relatively costly, offers uniform collection of donor information and additional e-mail and physical mail targeting. While it would only be ultimately useful for experimental purposes if its data were linked with Google Analytics’ demographic data, Sage Systems is still a possible point of exploration, as it is a professionally-run database management system. Conversely, if we could figure out a way to integrate Google Checkout into the existing AdWords campaign so that the addresses of those who used it could be shown to the charity, it would be very helpful for both the non-profits and our own analyses of the effectiveness of the advertising campaigns, especially in convincing them of the effectiveness of Google Checkout. The only information we currently have is whether or not someone reached the donation page; we do not know whether or not they actually completed a donation. It has been shown that, typically, less than half of the people who navigate to the donation page actually complete a donation, oftentimes only a small percentage of total visitors to the page. Without further knowledge of who actually donated, all we can really do is try to increase traffic to the donation page, not necessarily increase donations. If online donor management can be arranged, it would be ideal to link repeat donations as well. It is very conceivable that someone might find the charity through an online Web search, but maybe they get involved in this charity and give more money later on. In this case, only the first donation is attributed to the Web search, when in reality there may be several batches of repeat donations that are actually linked to the advertisement. Likewise, someone may already be donating to the charity and simply 35 uses the sponsored ad as the portal to the Web site, in which case the AdWords campaign did nothing to solicit more donations in that case. Either way, there needs to be a way to track recurring donations that originate through the web search. Another important aspect that could be greatly improved is writing effective ads. There is significant research in advertising as to what people most respond to in ads. Additionally, there is research on what types of characteristics non-profits should exhibit to attract more donations. With this in mind, one could theoretically develop what types of ads that should extract the most donations from potential searchers. Furthermore, testing the effectiveness of these ads is an important area of discussion. AdWords has software that allows different ads to be displayed randomly. Using this tool, one could randomly display two different ads on any given search term and see which one solicited more donations (for an explanation of this process, refer to Appendix H). Finally, an ideal next project would take the groundwork from this paper and apply it to for-profit organizations. There are a variety of issues that for-profits have that non-profits do not, not the least of which involves paying for its own ads. Subsequently, revenue earned by non-profits is essentially equal to profit. For-profit businesses not only have to pay for the ads out of company revenue, they also have to pay for the cost of the good and many other expenses associated with increases in advertising, resulting in marginal cost and marginal profit concerns. There are also concerns about the return on investment. Could this money be better spent on a different advertising campaign? The recurring revenue aspect is even more important when you are dealing with profitable businesses. They may “lose” money on advertising campaigns, at least through the 38 Appendix B – Tips for a successful application (Taken from Google) Google AdWords Information Google Grants provides eligible organizations with in-kind keyword advertising using Google AdWords so you can connect directly with your target audience. Through simple, short text ads that run on Google.com, thousands (or even millions) of people can learn about your organization online as they are searching for related information. When someone enters keywords (short phrases specifying a particular search query) into Google.com, ads targeted to those keywords appear alongside the search results. Paying close attention to the information below and the Google Grants Program Guidelines will help you to understand the way our advertising program works, and to successfully prepare your application. How AdWords Work Our automated system monitors the performance of your ads, and displays the most relevant ads in the most visible position on the page. However, some of your keywords may stop showing your ads on search results if they don't have a high enough Quality Score, which typically occurs when keywords aren't as targeted as they could be, and the ads they deliver aren't relevant enough to what a user is searching for. Once your campaign is running, you can see how well your ads perform by logging into your AdWords account to view your online reports, including how often people have clicked on each of your ads and keyword phrases, and the value of those clicks. 39 Effective Ads and Ad campaigns To ensure your ads are effective, we encourage you to review all the topics below: Targeting the right audience (reaching constituents in the region and language you specify) How your target audience finds your organization (developing your keyword lists) Use keyword matching options for greater effectiveness (AdWords keyword tools) Give searchers a reason to visit your site (creating effective ads) Give searchers an easy way to respond (sending searchers to the right landing page) Tracking conversions (measuring the effectiveness of your Google.com campaigns) 1. Targeting the right audience back to top Through AdWords, you can create a variety of campaigns (groups of ads for different audiences or services) as well as target your ads to different geographic locations and languages. The three options for geo-targeting are: Global or nationwide: Your ads will appear to Google searchers located in, or searching for, results in the country (or countries) you select. This option is best suited for global organizations that serve specific countries. Regions and cities: Your ads will only appear to Google searchers in the states and/or regional areas you choose. This option is best suited to organizations whose programs are centered in certain cities and states, or when you need to target searchers located throughout a city or cities. Customized: Your ads will only appear to Google searchers in a specified distance from your organization. This option is best suited to groups serving a limited and specific area. Languages: You can choose "All Languages" or select any of dozens of languages in which your ads can appear. 40 2. How your target audience finds your organization back to top Ask yourself which keywords – word combinations and phrases – you would type into the Google search box to find your organization's programs and services. Then, search using Google.com for those keywords you are considering. Make note of both the search and ad results you see, and eliminate the keywords that return results for information unrelated to topics your organization focuses on. Targeting your ads to the most relevant keywords will help your constituents to find you on Google.com. Remember that the more specific phrases are always better than general keywords. And a short, well-targeted list of words is much better than a long list of general keywords. We've found that our most successful advertisers use a combination of the following four types of keyword matching options: • Broad matching example Keyword: breast cancer If you enter your keyword without any formatting, the AdWords program keyword default is broad match. For example, if your keyword were breast cancer, your ad would show when a Google search includes the keyword breast cancer, regardless of other search terms used or the order in which they were entered. Your ads will also automatically show for expanded matches, including plurals and relevant variations. Broad match keywords can work very well when the keywords are specific to your organization. For example, here are search queries that might display an ad targeted to the broad match keyword breast cancer: 3. Use keyword matching options for greater effectiveness back to top 43 Use a hyphen: -teddy If your broad match keyword is bears and you don't want your ad to show for teddy bears, add the negative keyword -teddy. You can specify keywords as negative matching by preceding an unwanted keyword with a hyphen. This technique works to limit the display of your ad on broad- match searches that don't apply to your organization. For example, if our endangered bear organization wanted to run on the broad-match keyword bear, the following negative keywords should be listed to avoid showing on unrelated queries: -chicago -berenstain -bad news -teddy For detailed help with choosing the right keywords and matching options for your campaign, please try our keyword matching demo. 4. Give searchers a reason to visit your site back to top The keyword matching options above will help you understand how to target your ad, but what about your ad text? Our experience shows that the more closely your ad relates to your keywords, the more likely a user is to click on your ad. When someone enters a search query on Google and scans the search results and AdWords ads, your ad text will help Google users decide whether or not to click on your ad. Our Editorial Guidelines are designed to ensure that your ads attract the right people to your message. We review every ad to make sure it adheres to these Guidelines, 44 so it pays to be familiar with them. In addition to following the Guidelines, these tips will help you create compelling ads: Include keywords in your ad text or title. If your keywords appear in your ad text, Google users immediately recognize that your ad is relevant to their search. Adopt a clear style. Short, non-repetitive sentences work best. Double-check spelling and grammar, and avoid uncommon acronyms and abbreviations. Identify the unique aspects of your organization and services. What makes your website useful, relevant, and different? Call attention to the unique benefits you offer to attract more Google users. Example: "Search our physician database to find a specialist in your area" or "Chat with other breast cancer survivors." Use a strong call to action. A "call to action" prepares your audience for what you want them to do: register, join, subscribe, and so on. Make sure that this phrase is unique and specific to your program so that it is more informative and compelling, and distinguishes you from the competition. 45 Example: "Register for membership," "Volunteer in your neighborhood today" or "Call our confidential help line." One size doesn't fit all. Use our targeting options (country, language, region, state, zip code) to make sure you reach the right audience. If your organization has many different programs and initiatives, focus on one audience per ad (you may create up to 25 ad campaigns and up to 100 ad groups with unlimited ads). View sample keywords and sample ads. Keywords for sample ad #1: cancer surgery options cancer treatments information about cancer Sample ad copy #1: Keywords for sample ad #2: cancer volunteer cancer volunteers cancer survivors Sample ad copy #2: 48 No Employer ID#: Organization type: Other, please describe: Organization Mission Statement / Primary Objective: (Briefly describe the nature of your organization, your activities, your clientele and the location(s) of the services you provide – character limit 1000.) Does your organization operate solely in the San Francisco Bay Area? Yes No Yes, please describe: Annual operating budget: Are you affiliated with political advocacy groups?
 Yes
 No
 If yes, which ones? 
 Explain how you expect Google AdWords advertising will contribute to your organization. (Please limit your response to 500 characters.) Who is your target audience for online advertising? (Please limit your response to 150 characters.) 49 Please provide a sample of an ad you might run through Google Grants. Reviewing the Google Grants information sheet will help you submit a strong application by following our Editorial Guidelines, and showing an understanding of the way the AdWords program works. Line 1: 25 character headline Line 2: 35 character limit Line 3: 35 character limit Line 4: display URL (35 character limit) Example: Please provide a list of keyword / keyword combinations that you feel are relevant to your organization. Viewing our keyword matching demo will help you choose appropriate keywords for your campaign by understanding our broad match keyword default and other matching options. What are your website traffic figures (page views/impressions per month)? Do you have an active paid AdWords account? Yes No Yes, please enter your customer ID number: 555-555-5555 (10 digit) Do you accept online donations? Yes No How did you hear about the Google Grants program? Other, please describe: 50 
 See next page for extended Answers to questions contained within the dialog box Organization Mission Statement / Primary Objective: (Briefly describe the nature of your organization, your activities, your clientele and the location(s) of the services you provide – character limit 1000.) The mission of FOOD for Lane County is to eliminate hunger by creating access to food. We accomplish this by soliciting, collecting, rescuing, growing, preparing and packaging food for distribution through a network of social service agencies and programs; and through public awareness, education and community advocacy. FOOD for Lane County is the regional food bank serving all of Lane County, Oregon. As the second largest food bank in the state, FOOD for Lane County finds creative solutions to hunger and its root causes. We believe a responsive food bank includes programs that help people help themselves. Food banking also requires the participation of the whole community. Explain how you expect Google AdWords advertising will contribute to your organization. (Please limit your response to 500 characters.) The goal of our Google AdWords campaign would be to elicit more donations, both in the form of food and money to better accomplish our core goals of feeding the hungry within Lane county. Another goal is to get the community more involved by getting people to volunteer and participate in more events, whether that be by volunteering in a community garden, or cleaning the warehouse. Who is your target audience for online advertising? (Please limit your response to 150 characters.) The target audience of our campaign would be people within Lane County who can give their time effort and skills to our cause. Keywords: Volunteer Eugene, Volunteer Springfield, Volunteer etc.; Help Hungry, Help homeless, Food donations, sustainable food 53 • You cannot bid more than $1.00 per click for any keyword. • Your monthly ad spend will be capped at $10,000. • If your account consistently reaches the $10,000 per month cap, you may be eligible to apply for a $40,000 per month cap with additional services. Learn more about the eligibility requirements and access the application. Appendix E – Overview of the AdWords program (Some Taken From Google) A user signs up for the program (or applies for a grant, in the case of non-profits) and has several decisions to make right away. The first is the kind of advertising they wish to do: cost per click (CPC) or cost per impression (CPM). With CPC advertising, the company is charged only when a sponsored ad is clicked on. With CPM, the company pays for a given number of impressions ahead of time and isn’t charged until the ad has been viewed X amount of times. For example, if Africare purchased 1,000 impressions for $100, it pays the $100 up front and will not be charged again until the ad has shown up in searches 1000 times. It is important to note that, when using CPM, a search that causes the ad to show counts against the total whether or not the ad is clicked on. It is critical to understand what separates ads that are hardly seen from ads that get immense amounts of traffic, and the answer lies in Google’s automated bidding system. As obtained from Google’s information database: Your ad will compete with ads from other advertisers who also have set their own prices. Your success will be based on how much you bid, the quality of your ad, how many other people want to advertise on the same keywords, and other factors. Very popular topics and keywords (such as real estate or hotels) can cost more because so many other people also want to advertise on them. If you set very low prices on those popular keywords, your ad may appear only rarely or not at all. If you set higher bids on very specific keywords, your ad is likely to appear more often, and in higher positions yielding more clicks, at the price you want. The second sentence touches on the determinants of ad position: namely, minimum CPC bid, Quality Score, and the actions of competitors. Mathematically 54 speaking, ad rank is calculated as CPC Bid*Quality Score. Quality Score is a measure of how relevant the ad is to its respective site, but as stated, the score itself is not divulged by Google. Actual CPC paid (not the amount bid) is determined as the minimum value that keeps a company from falling one more position in the rankings. The most important step in the process is deciding exactly what words to bid on. This can be done in myriad ways, but the overall goal is to minimize costs. Logically, words or phrases that are more common (e.g. Africa, Africa information) will cost more and will mean higher competition for advertising rank than lower-ranked, less common phrases. Fortunately, Google expedites the process by offering convenient estimation tools for prospective word purchases. Here is a sample of Google’s estimation software at work: 55 The phrase entered into the estimation engine was “Africa aid,” which can be seen at the top of the list. The search included potential synonyms as well, which resulted in the ensuing entries. The maximum CPC selected here is $1 – this is the highest Google will allow a non-profit to pay for one click-through – and the actual estimated CPC is given by the Estimated Avg. CPC column. With a maximum bid of $1, Google estimates an actual price of about $0.80. Next to that is the Advertiser Competition column, which roughly indicates how desirable a given word or phrase is. Some phrases, like “aid in Africa,” face almost no competition in Google’s bidding system. Others, like “AIDS Africa,” face heavy competition from other bidders. The selected phrase “Africa aid,” faces moderate competition. The chart also shows search volume for the given phrase, broken down three ways: average search volume over time, average search volume over the past month, and average search volume over the last year. While imprecise, this is an important tool in scouting out potential search terms, especially given Africare’s budgetary restrictions. In the case of any non-profit, the daily limit on AdWords expenditures is $330, so that is one constraint that must be accounted for. As previously mentioned, the maximum CPC bid a non-profit can set is $1, which must also be kept in mind. While this is the general auction process, there are still a few unknowns. First, there is no exact way to account for competitors’ bidding information. Second, Google’s estimation engine uses old data – not that there is a lack of it, but it can’t possibly account for the future. Third, it is unclear how Africare’s status as a non-profit causes fluctuations in search volume; it is entirely possible that current events are doing all the advertising Africare needs for them. 58 efforts,” “African governance initiatives,” etc.) are not costing Africare anything and need only to experience a jump in search phrase traffic to be shown. This practice of selecting words that are inexpensive and in low demand will prove to be a key strategy for an effective AdWords campaign. • Every search term is at the maximum CPC bid right now – this is where the bulk of unnecessary expenditures stems from. Even with a poor Quality Score, the minimum bid is typically 30 to 50 cents; by setting the maximum bid at one dollar, the average CPC ends up costing Africare nearly double that amount (77 cents). • Since the new Analytics tracking code was added to the site, only 2.05% of Google search results have led to a Conversion (defined as the user reaching the Donations page). Search engine searches account for 56.37% of all traffic, which is promising, but the small Conversion rate is less than ideal. However, some of the ad groups, and even specific keywords within said groups, are performing much better than others. For instance, 8.37% of people searching for phrases in the “Donation” group ended up at the donation page. Among this group, 23.81% of users searching for “African donation” ended up clicking on the ad and following it to the Donations page. • Users who click on the paid ads spend an average time of 2:35 browsing the site and see 2.99 pages per visit. Upward of 1,000 of the 1,558 visitors (for this campaign) to the site have access to DSL, Cable or T1 Internet connections (over 400 visitors had unknown connections, but it is safe to assume a majority were high-speed), almost every single visitor is from the Americas (65.3%) or an English-speaking country (79.5%). Among U.S. states, California had the most visits to the site (109 visits), Oregon saw the most pages per visit (6.44) and Oklahoma had the highest conversion rate (20%; 1 visitor out of 5 clicked through). The District of Columbia, where Africare is based, had a mere 10 of these 1,558 visitors, and none of them reached the Donations page. 59 Given this information, here is our list of suggested changes to Africare’s advertising campaign: Do away with the Specific Regions campaign – or at least pare it down. This is a simple way to cut expenditures. The data so far suggests that the ROI is going to be extremely low, if not ultimately zero, for this search group. Only 121 visits out of the 1,679 total have been via this specific campaign’s ads, and none of the visits have resulted in clicks through to the Donations page. Of all the specific ads, only the ones that read “donate to [region]” or “help [region]” have garnered any serious searches, and they all have the highest Quality Scores; the rest of the searches have such poor Quality Scores and search volume that even if the ads were showing, they would be unnecessarily costly. As such, we recommend keeping only those two variants on the region-specific phrases and cutting out the rest, to minimize “Poor” phrases (Google’s Quality Score, which evaluates words based on their CPC bid, relevance to the destination site and other factors; Scores can be Poor, OK, or Great), reduce clutter, and increase CTR. Do away with the “Africare” group under the General Africa campaign. This is the top group in terms of CTR, lowest CPC and highest rate of Conversions, but it is ultimately costing Africare more than it needs to. As touched on earlier, visitors searching for Africare are much more likely to know what they are looking for already; the data helps confirm this thought. Users who searched for 60 “Africare” spent longer on the site than any other ad group (4.01 pages per visit; average length of visit 4:34). Because we can assume users searching for “Africare” already know what they are looking for, having the sponsored ads above or to the side of the regular search results only encourages clicks – even though the first two standard search results both link to Africare.org. Though it has only cost Africare $663.96 to date, it is money that doesn’t need to be spent and could be spent elsewhere. Do not change CPC bid; lower it once the budget constraint is met. This, too, is an easy way to lower total expenditures, and will ultimately save the most money of all these prospective changes. As stated, every phrase currently sits at a bid of $1 per click, the maximum bid allowed for a non-profit organization. While this ensures the highest possible ad ranking (the order in which ads appear is calculated by multiplying the maximum CPC bid by the Quality Score; Google holds the numerical Quality Score figures internally), it is also the most expensive possible method of advertising. It is important to note that Google limits Africare’s account to a daily budget constraint of $330, or $10,000 a month when extrapolated. For the sake of anecdotal argument, consider the following results when using Google’s Traffic Estimator, a tool that predicts search volume and estimated clicks given historical searches for the submitted terms. At the current CPC of $1, it estimates that Africare should get between 327-442 clicks a day from its advertising (parameters included every single search term from the General Africa campaign at the current budget constraint of $330; see chart below). 63 Add Google Checkout to integrate existing Analytics code with financial returns. The existing Analytics code tracks Conversions as whenever a searcher reaches the Donations page from the Google search page. By using Google Checkout along with Analytics, Conversions can be measured from a financial perspective as well. Based on our experience with Checkout, it is simple to set up and requires users to enter a name, e- mail address and mailing address to make a “purchase” (Google Donations works for non-profit organizations, though; we haven’t been able to use this in our test, but it seems simple enough). This level of synergy will make it easier to manage the Africare account (there would only be the one log in name and password) and much easier to track a uniform set of data that is consolidated in one place. Create more Emergency Funds for specific crises. The current “Emergency Fund” campaign has had fairly good success when active, for multiple reasons: search volume during major world events (flooding and famine, for instance) increases as media coverage does, not to mention research suggests that more value should be placed on specific, less common search terms. These phrases will be cheaper than normal words, more topical and thus more likely to be searched for as events arise, and will ultimately provide a solid return on investment. Add more search phrases. Though there are already several hundred search phrases, there is no harm in adding more to the list, especially less common phrases that still relate to Africare’s line of work. By targeting more specific phrases and ensuring the ad text is unique to the 64 given search terms, Quality Score will increase and minimum bid will fall – which means you’ll be paying less for phrases that almost no one else is buying in the first place. If none of the phrases pan out, that’s ok: it won’t cost Africare anything if no one is clicking. It will do no harm and has the potential to be effective. Appendix H – Creating an experiment within Google AdWords (Taken from Google, but simplified) Steps Follow the following route through the account: Campaign Management  Website Optimizer  New Experiment (see image below) For an A/B Experiment (recommended for Africare/FFLC): 1.Choose the page you would like to test 65 Examples of potential test pages could be your homepage or a product detail page. 2.Create alternate versions of your test page Create and publish different versions of your test page at unique URLs so that Website Optimizer can randomly display different versions to your users. These URLs could be bookmarked by your users, so after your experiment finishes, you may want to keep these URLs valid. 3.Identify your conversion page This is an existing page on your website that users reach after they've completed a successful conversion. For example, this might be the page displayed after a user completes a purchase, signs up for a newsletter, or fills out a contact form. For a multivariate experiment: 1. Identify experiment pages 1a. Plan your experiment 2. Add JavaScript tags to experiment pages 3. Create variations 4. Review experiment settings and launch 5. View Report 68 Appendix J – Charitable Giving The early economic literature on charitable giving attempts to explain charitable giving using a model of “pure altruism,” in which the individual derives utility purely from the public good. Olson (1965) shows that using this model that the public good will be underprovided for. An alternative explanation proposed by Adreaoni (1989) proposes the theory of “Warm Glow,” the basic idea that people give to charity not only because they may derive utility from the charity itself (i.e. donating to the art museum in order to get special privileges that allow special entry into the museum) but feel good about the act of giving itself. This model helps to explain giving across income levels, which was a limitation of previous pure altruism models. Hollander (1990) proposed a caveat to the Warm Glow model, suggesting people may not derive utility from the amount of the gift itself, but from the perceived relative amount of the gift. In other words, it does not matter if one gives $5,000 if everyone else is giving $10,000 to the museum. 69 Glazer and Konrad (1996) model public charity as a means of signaling income, and the ability to associate with people of the same or higher social status. Individuals can signal wealth by consuming private goods but they may not choose this method exclusively. Additionally a model was presented in Harbaugh (1998) further explored the idea that benefits from charitable giving may arise in two different ways: interval giving (i.e. altruism), and the benefit derived from other people knowing how much had been given. This helps explain why large anonymous donations are rare, while universities that receive large donations name buildings after donors. It is important to note that the precise amount does not have to be revealed as long as people can guess at the amount. This is supported by the idea that when a university institutes recognition brackets, such as gold donor, silver donor, etc., people tend to give the minimum amount in order to make the next level. Dawson (1998) proposed four categories of motivation for charitable giving: career advancement, income advantages, reciprocity, and self-esteem. The study uses survey evidence to try to explain giving to medical research causes. The results show that for giving to medical research both income advantages and reciprocity are significant, while the other two have the expected signs, the results are insignificant. There is much evidence that taxes play a role in charitable giving. Auten, Sieg, and Clotfelter (2002) show that the trend in inequality of donations by high versus low- income individuals is partially explained by the stronger trend in increased income inequality relative to the trend of increased uniformity in taxes. The subsidization of charities and the crowding out effect have also been studied extensively in economics. Brooks (2000) proposes that crowding out of private 70 donations may exist, but they only occur at high levels of subsidies. This stems from the belief that donors treat grants as imperfect substitutes for charitable giving that follows from models with impure altruism. Adreaoni & Payne (2003) proposes that a second reason, a strategic response of a charity to reduce fundraising efforts after it has received a grant. They suggest that analysts should account for behavioral response of the charity as well as the response of donors when evaluating government grants. Payne (1998) found that when using an OLS regression controlling for heterogeneity and other factors, the crowding out effect is insignificant for nonprofit shelters. However, when they controlled for endogeneity using 2SLS, then donations average 50 cents on the dollar. Appendix K – Characteristics of Charitable Givers It has been found that religion effects charitable giving in both time and money. Brooks (2003) found that a secular person is 23% less likely to give money than a religious person and 26% less likely to volunteer. A possible explanation, he posits, is that secular people may prefer government to take care of social problems, or religion is key in promoting social capital. It has been found that the price of giving is almost zero and completely insignificant for religious donations, though religious individuals do take their marginal tax rate into account when donating to non-religious charities. Additionally, Kitchen (1992) found that both wealth and the age of head of household are important determinants in charitable giving to all charities either religious or non-religious. Adreoni, Brown, & Rischall (2003) found that men and women have different preferences as far as giving to charitable causes. Men prefer to give to a few charities, but give a substantial sum to each, while women prefer to give to more charities 73 Dawson, S. (1988) Four motivations for charitable giving: Implications for marketing strategy to attract money for medical research , Journal of Health Care Marketing , Vol. 8, No. 2, pp. 31-37. Edelman, Benjamin, Michael Ostrovsky, Michael Schwarz. "Internet Advertising and Generalized Second Price Auctions: Selling Billions of Dollars worth of Keywords." American Economic Review 97(2007): 242-260. Feng, Juan, Bhargava, Hemant and Pennock, David, "Implementing Sponsored Search in Web Search Engines: Computational Evaluation of Alternative Mechanisms" . Informs Journal on Computing, Forthcoming Available at SSRN: http://ssrn.com/abstract=721262 Glazer, A. and Konrad, K.A., 1996. A signaling explanation for private charity. American Economic Review 86 4, pp. 1019–1028. "Google AdWords Help Center." Google Inc.. 5 Jun 2008 <http://adwords.google.com/support/>. "Google AdWords." Google Inc.. 5 Jun 2008 <https://adwords.google.com/select/Login>. "Google Analytics." Google Inc.. <http://www.google.com/analytics/>. "Google Grants ." Google Inc.. <http://www.google.com/grants/index.html>. "Google Grants Program Guidelines." Feb. 2008. Google Inc. <http://www.google.com/grants/policies.html>. "Traffic Estimator." Google Grants. Google Inc. <https://adwords.google.com/select/TrafficEstimatorSandbox>. Harbaugh, William T. “What Do Donations Buy? A Model of Philanthropy Based on Prestige and Warm Glow.” J. Public Econ. 67 (February 1998): 269–84. (b) Haynes, Marcus, Jeniffer Thorton, Sandra Jones. “An Exploratory Study on the Effect of Positive (Warmth Appeal) and Negative (Guilt Appeal) Print Imagery on Donation Behavior in Animal Welfare" (2004). Available at Research Online: http://ro.uow.edu.au/hbspapers/80. Hollander, H., 1990. , A social exchange approach to voluntary cooperation. American Economic Review. 80, pp. 1157–1167 Kitchen, Harry. "Determinants of Charitable donations in Canada: A Comparison Over Time." Applied Economics 24 No. 7 (July, 1992): 709-13. 74 Nielsen Norman Group Report , "E-commerce User Experience: Design Guidelines for Trust and Credibility ." <http://www.nngroup.com/reports/ecommerce/trust.html>. Novak, Thomas P., Donna L. Hoffman, and Y. F. Yung. 2000. "Measuring the Customer Experience in Online Environments: A Structural Modeling Approach". Marketing Science19 (1): 22-42. Olson, Mancur, The Logic of Collective Action (Cambridge, UK: Cambridge University Press, 1965). Payne, A., 1998. Does the government crowd-out private donations? New evidence from a sample of non-profit firms. Journal of Public Economics 69, pp. 323–345. Schervish and Havens, 1997. P.G. Schervish and J.J. Havens , Social participation and charitable giving: a multivariate analysis. Voluntas 8 3 (1997), pp. 235–260. Tooley, Melissa. “Perfecting Your Page: Can donation page optimization boost online donations?” (May 9, 2008). Usborne, Nick. Net Words: Creating High-Impact Online Copy . New York City: McGraw-Hill, 2002. Venable, B. T., G. M. Rose, D. Bush, and F. W. Gilbert (2005). "The Role of Brand Personality in Charitable Giving: An Assessment and Validation." Journal of the Academy of Marketing Science, 33(3): 295- 312.