The Future of Marketing in an AI Moderated Digital World

16

October

2023

No ratings yet.

Currently companies can apply online marketing in various ways. Ads are placed all around. Websites are constructed in ways that humans respond to best, as data informs of how consumers behave and how to improve statistics such as conversion and click-through rates (Fogden, 2023). Search engines form something of a marketplace for webpages where online auctions are used to determine what will be shown. Marketeers use search-engine optimization (SEO) and paid search-engine advertising (SEA) to win auctions and feature amongst the first results (Eology, n.d.). This is how it has been. Now, one development shows potential to change it all. Since February 2023 Bing is powered by Open AI’s GPT-4, an advanced AI that Bing uses to improve its search engine and to act as a copilot and chatbot (Mehdi, 2023). This approach can forever change the way we use search engines and how results are generated. It could even change the way we interact with the internet altogether.

Right now, marketeers target us directly. Their methods are based on getting information directly to their target audience in the most appealing way. The intervention of AI may change this. Granted, an AI like GPT-4 is trained on pre-existing datasets and does not have direct internet access, so its responses cannot be influenced so easily. I asked Bing AI, which connects to Bing’s search engine, whether its response is influenced by paid search advertising, and it still categorically rejects the possibility. According to it, Bing AI and even the search results it might draw on only try to use the most relevant and reliable sources to provide information. Additionally, it says that any ads and sponsored links are filtered out by its internal tools. Perhaps advertising does not influence it, but still search engine optimization can help websites appear more relevant and end higher up the results. In this way internet marketing already focuses on convincing the search algorithm of a website’s value, not the user directly.

Regardless, we might get to a point where we rely on AI assistants to get all our information, or where search engines are run entirely by an AI which browses the internet, filters through information, and presents us with the best results. AI certainly has a promising future in real-time content moderation (Darbinyan, 2022). And, according to Santiago (2023), marketeers can even use it to protect their brand. But when marketeers themselves are the ones who need to get information through an AI gatekeeper, how will they respond? Many current strategies can make content appeal to humans, but what will the AI respond to? If AI is the middle-man, marketing efforts might have to be constructed in such a way that AI filters and retells information in the way marketeers ultimately want to reach consumers. It becomes important to consider what the AI will respond to, what it will need to see the information’s value.

Perhaps AI ushers in the end of traditional online marketing to consumers. Perhaps AI will simply assess an offer at its true value, and recommend it only if there is a good fit with the consumers needs, be it a new chair, information, or entertainment. Think about recommendations on social media platforms such as TikTok, which use algorithms that carefully select content that a user will probably like. This might be a preview of how we will receive all information: moderated by AI, and it could be a win-win situation. Marketeers could rely on AI to do the work of targeting, personalizing, and distributing content to the right audience, while users can rest easy knowing the information is of utmost relevance. AI could moderate content better than either marketeers or consumers themselves ever could.

References

Darbinyan, R. (2022, June 14). The growing role of AI in content moderation. Forbes. https://www.forbes.com/sites/forbestechcouncil/2022/06/14/the-growing-role-of-ai-in-content-moderation/.

Eology. (n.d.). SEA know-how: How to use search engine advertising unerringly! Eology Magazine. Retrieved October 16, 2023, from https://www.eology.net/magazine/sea-know-how#jump_function.

Fogden, T. (2023, April 14). What makes a good website? 12 must-haves. Tech.co. https://tech.co/website-builders/what-makes-good-website.

Mehdi, Y. (2023, March 14). Confirmed: the new Bing runs on OpenAI’s GPT-4. Microsoft Bing Blogs. https://blogs.bing.com/search/march_2023/Confirmed-the-new-Bing-runs-on-OpenAI%E2%80%99s-GPT-4.

Santiago, E. (2023, April 7). AI content moderation: How AI can moderate content + protect your brand. HubSpot. https://blog.hubspot.com/marketing/ai-content-moderation.

Please rate this

The Best Position: Winning a Google Ads Auction

7

October

2021

No ratings yet.

Can you imagine processing 6.9 billion queries a day? I certainly could not, nor would any human-being be able to handle that. But then, Google does it every day. It started back in 1996, with a ranking system that rewarded links that came from sources that were relevant and penalized those that were not (Battelle, 2005). Nowadays, it is one of the leading internet search engines, and only the term ‘Google’ already generates 7.2% of the worldwide search traffic (Mohsin, 2020). But how does this search engine work?

Google Ads

Two years after the launch of Google.com, Google Adwords was created, which has been rebranded to Google Ads in 2018. This platform offers pay-per-click (PPC) online advertising that allows advertisers to display their ads on the search engine results page (SERP). Below is an example of a SERP, which displays two sections of Google ads. The red box represents ads of paid search accounts and are usually recognizable due to the little yellow box next to their link. Companies have paid to appear at the top of the search results (Sukhraj, 2019). This process of advertising directly on Google, is better know as Search Engine Advertising (SEA). The green box represents organic search results, which refer to unpaid listing on a SERP (Randolph, 2020). Google ranks these results according to their quality, relevance to the search words and other factors (Randolph, 2020). SEO includes all activities you engage in to ensure your website scores higher in the organic results (Poorthuis, 2019). Both SEO and SEA rank advertisements using auction mechanisms. 

Paid vs. Organic search

How does Google Ads work?

Google Ads work with a PPC model that determines which ads should appear with a lightning-fast ad auction. These auctions take place every time a user enters a search query into Google. To ‘win’ the auction, you will have to optimize your quality score and bid amount. The bids represent the maximum you are willing to pay for your ad to be displayed on the top of the page. In addition, it is possible to set a maximum daily budget for your ad and you will never exceed that amount. Google takes the bid amount and pairs it to your quality score (i.e., assessment of your ad) (Tran, 2020). So, even if your competitor places a higher bid than you did, you can still win a higher position if you have a higher quality score (by using relevant keywords, ads, extension, etc.) (Google Ads Help, n.d.). 

Transition of Auction mechanisms

In 2019, Google Ads transitioned from a second-price auction to a first-price auction model (Setupad, 2021). Initially, with a second-price auction, the winner did not have to pay their own bid, but just 0.01$ more than the rate of the second-highest bidder (Setupad, 2021). In the past years, there were several programmatic developments that contributed to the complexity and operational inefficiency in the second-price auction:

  • Header bidding, which is an advanced method of programmatic ad buying that allows multiple advertisers to participate simultaneously in the auction to win ad space (Tenney, n.d.). 
  • The real-time-bidding (RTB) protocol that enabled buyers to build their own system called a ‘real-time bidder’, which connects to Google servers and generates a constant stream of available ad inventory. 

RTB and header bidding made it possible for publishers to allow multiple ad exchanges (SSPs) to bid on their ad inventory at the same time. Thus, creating complexity and operational inefficiencies. Subsequently, there were often hidden intermediary fees taken by SSPs (Setupad, 2021).

Therefore, Google adopted the first-price auction model, where bidders participate simultaneously in the auction and the highest bidder wins. The main difference between these two auctions, is that publishers earn more revenue in the first-price auction model. Also, the model offers more transparency and works better with the developments that were mentioned above. 

What impact do you think that the transition from second-price auction model to a first-price auction model will have on Google? Was this a smart decision?

References

Battelle, J. (2005, January 8). The Birth of Google. Wired. https://www.wired.com/2005/08/battelle/. 

Google Ads Help. (n.d.). How the Google Ads auction works. https://support.google.com/google-ads/answer/6366577?hl=en. 

Mohsin, M. (2020, April 3). 10 Google Search Statistics you need to know in 2021 [infographic]. Oberlo. https://www.oberlo.com/blog/google-search-statistics.

Poorthuis, C. (2019, December 6). SEO and SEA: What is the difference and what are the advantages of the two? Grid Marketing. https://www.griddigitalmarketing.com/seo-and-sea-what-is-the-difference-and-what-are-the-advantages-of-the-two/.

Randolph, K. (2020, December 28). What is Organic Search in Google Analytics? WebFX. https://www.webfx.com/blog/seo/what-is-organic-search-in-google-analytics/. 

Setupad. (2021, April 22). First-Price vs Second-Price Auction | Differences Explained. https://setupad.com/blog/first-price-vs-second-price-auction/. 

Sukhraj, R. (2019, November 19). Organic search vs paid search: What’s the difference in 2020? iMPACT.https://www.impactplus.com/blog/organic-search-vs-paid-search. 

Tenney, M. (n.d.). What is header bidding? Adthrive. https://www.adthrive.com/what-is-header-bidding/.Tran, T. (2020). A Beginner’s Guide to Using Google Ads (Previously Google Adwords). Hootsuite.https://blog.hootsuite.com/google-ads/.

Please rate this

20 candles for Google

27

September

2018

No ratings yet. logo

In response to the 20th anniversary of (let’s admit) our best friend in high school, I thought it would be a good idea to honour our dear friend Google by writing a post. Considering the evolution that I made myself throughout the years (and I am only one year and two months older than Google), I would like to go through some key elements in Google’s evolution.

Google started off as a start-up between two PhD student at Stanford University, but turned into a giant business worth $766.4 billion (May 2018). Google has therefore the third largest market value, after Apple Inc. and Amazon (Statista, 2018). But how were two students able to go from a simple search engine in the old days to the Google we know today?

Larry Page and Sergey Brin, the two PhD students, started working on a structured search engine in 1996, using a PageRank algorithm which measures the importance of a website by how many times it is linked to. By making use of this algorithm, Google was more convenient than already existing players (e.g. Yahoo) who did not provide any structure. The domain name Google.com was registered on September 15, 1997.

timeline

What I would like to focus on now however, is this amazing search algorithm of Google , however the following examples are only a tip of the iceberg. In 2003 many updates ameliorated the algorithm by changing them and by decreasing spam (e.g. key word stuffing, hidden text and links between co-owned websites). In 2004, Google’s algorithm was able to understand related words and concepts. Over the next 5 years, Google was among many other things able to complete search terms and it increased indexing speed and freshness, while still decreasing spam. And many updates which they even gave names like “Panda” and “Penguin” still followed. In 2014, HTTPS/SSL would give a “lightweight” ranking boost to secure sites. In 2015, “RankBrain” Algorithm used machine learning to impact rankings. And then in 2017 there is also “Fred”, a name to cover all three updates to the search algorithm, which happen every day.

Personally I am really happy with the search results that Google has brought me and I certainly would not dump Google very soon for any other search engine. But does this mean it is too late for alternative search engines to catch up?  I have looked up some alternative search engines and an article by Luke Richards (2018) showed me Bing, Ecosia (mainly based on Bing), Twitter, Slideshare and DuckDuckGo. So far they did not please me as much as Google. For example DuckDuckGo does not store any personal information, which is good for people who value privacy, but definitely has some drawbacks when you value optimised and relevant search results. I also find it hard to believe Twitter and Slideshare would replace Google. There are definitely some drawbacks to consider regarding Google, such as privacy concerns and the fact that Google is able to exploit us with targeted advertising. But for now, I am willing to take these drawbacks in order to enjoy the benefits.

References:

A Timeline Of Google Search Algorithm Updates [Infographic]. (2017, September 26). Retrieved 27 September 2018, from https://www.twinword.com/blog/timeline-google-search-algorithm-updates-infographic/

Google Timeline. (n.d.). Retrieved 27 September 2018, from http://www.datesandevents.org/events-timelines/18-google-timeline.htm

No need for Google: 12 alternative search engines in 2018 | Search Engine Watch. (n.d.). Retrieved 27 September 2018, from https://searchenginewatch.com/2018/05/21/no-need-for-google-12-alternative-search-engines-in-2018/

Please rate this

The Big Brother (Google) is Watching You

4

October

2016

No ratings yet.

Source: Vordaan Advocaten

Imagine it’s 1984. The Big Brother is watching your every step and controlling every aspect of your life. You live in the totalitarian state Oceania where every citizen is under constant surveillance. Sounds familiar?

In 2016, the narrative described by George Orwell in his famous novel “1984“is still relevant but for different reasons. In the past, the role of the Big Brother was played by political parties and figures while today – the Big Brother is represented by large corporations with access to huge amounts of data. Such companies are Facebook, Google, Amazon and others, but this blog post will focus on Google.

Google

Google can possibly track one’s every movement online as one of its products is the most popular search engine, taking 73.02% of the global market share (NET Applications, 2016). According to Wikipedia, there are about 100 Google apps available on Google Play, although some of the apps might no longer be active or might have different names. As of 2016, Google has 7 products with more than 1 billion users: Gmail, Android, Chrome, Maps, Search, Youtube and the Google Play Store (Popular Science, 2016). I mean, this is the amount of world population using these 7 products!

All the information from these 100 apps and billions of users gets accumulated and stored by one company. All Google apps and products are interconnected so that the company can make use of the large amount of data for statistical purposes as well as to get customer insights by storing the information from each individual separately. Thus, information from all products is stored at a person’s Analytics account. By joining data sources, the tech giant is aiming at providing better and highly customized services.

Find out more about data privacy concerns from the following video featured on CBSN.

Customer- centric

Google is customer- centric indeed, but who is the customer? One would probably think that Google serves individuals and companies for free or for a very small fee. Although the tech giant does provide some of its services and products for free, it actually earns 95% of its revenue from advertising. This is how it works – AdWords enable advertisers to publish ads on Google search results pages as well as on the websites of Google affiliates (AdSense publishers).

Initially, Google management thought that revenue will come primarily from licensing and providing Internet services. However, as technology evolved, Google started using auctions technology to allocate ads, and thus completely changed its business model. So technically speaking, people are tools for acquiring revenue, and companies using the advertising services are the primary customers of Google.

Google Might Create a Problem

One problem that has been caused due to Google’s business model is personalization. Google claims that customization helps it better tailor ads and information to one’s needs. However, in the future, people might get more polarized as Google blocks their access to diverse opinions, i.e. as people fall in the so-called filter bubble. Since Google personally tailors results, two different people who search for the same word or phrase would get different search results. For example, as democrats tend to lick more on links with news about democratic views and democratic politicians, and republicans do the same – each group will get search results based on their interests and thus not have access to other opinions.

Find out more about the filter bubble issue in the following TED talk by Eli Pariser.

Google Might Create a Solution

On the other hand, a problem that Google can solve is the information asymmetry. Imagine that you are using the Google Fit app and Google starts selling health insurance companies people’s data. Health insurance companies would know how healthy one is and would be able to better predict diseases and potential health-related expenditures. This solution would reduce the information asymmetry, but it will also pose some security, privacy and ethics questions. Although currently the company does not provide personal information to third parties, there is no guarantee that it will not start doing it in the future.

Conclusion

The Big Brother will keep watching you and will keep collecting data about you. Sometimes its actions will favor you, sometimes they will favor corporations and/or other people. Although Google provides technological platforms, the data and decision related to the data usage pose ethical questions. As Google algorithms evolve, they might start making ethical decisions. But as people are not impeccable, and machines are made by people, machines can’t be errorless. It is indeed interesting to see how far Google and its algorithms can get.

Here you can find out more about the type of information that Google has potentially stored about you and where it is located.

 

Sources:

“Big Brother Google Is Watching You!”. Voordaan.com. N.p., 2016. Web. 4 Oct. 2016.

“Google Ads Case Study Analysis”. Ecommerce-digest.com. N.p., 2016. Web. 3 Oct. 2016.

“Google Has 7 Products with 1 Billion Users”. Popular Science. N.p., 2016. Web. 2 Oct. 2016.

“Google Privacy | Why Data Protection Matters”. Privacy.google.com. N.p., 2016. Web. 2 Oct. 2016.

“List of Google Apps for Android”. Wikipedia. N.p., 2016. Web. 4 Oct. 2016.

Pariser, Eli. “TED Talks – What FACEBOOK and GOOGLE Are Hiding From The World – The Filter Bubble”. YouTube. N.p., 2016. Web. 3 Oct. 2016.

“Search Engine Market Share”. Net Applications. N.p., 2016. Web. 2 Oct. 2016.

Stiles, Jackson. “Google Is Watching: Find Out What It Knows About You”. The New Daily. N.p., 2014. Web. 4 Oct. 2016.

“The End Of Privacy “The Data Brokers: Selling Your Personal Information””. YouTube. N.p., 2016. Web. 3 Oct. 2016.

Please rate this