Project Magi: Google's Response To ChatGPT-enabled Bing - What Marketers Should Know

For the first time in the history of this newsletter, I did a poll to gauge the interest of my audience on the topic I was planning to cover this week. Out of the 16 people who voted, 15 were interested in reading about Project Magi. That was enough encouragement for me.

So here I am, as promised with a detailed article on the topic. And I have tried to cover as many different dimensions as possible about the topic (including some predictions for the future).

Setting the context

We have spoken about the application of AI in marketing many times in this newsletter. But none of those topics was as critical as this, given that these recent developments have put Google in a mode of panic.

The new AI-powered Bing has seen a 25% increase in monthly page visits ever since the integration of ChatGPT and GPT-4. There was also a potential threat of Google losing a multi-billion dollar contract from Samsung.

Project Magi is Google's response to this sudden onslaught from its rival Microsoft. With Microsoft extending its partnership with OpenAI (the creators of ChatGPT and GPT-4), Google couldn't afford to wait anymore to accelerate its efforts in the AI space.

In this edition of the Elevate Your Marketing newsletter, we will learn:

• How the new AI-enabled Bing and Edge are different from their old self.
• What does Project Magi entail, and what are the AI tools Google is currently working on as a part of the initiative?
• What could these new developments in AI mean for the future of search?
• And above all, what does this shift mean for marketers, and how you should be prepared?

Before you read on, I would recommend you to check out my article titled The New Era Of Generative AI In Marketing - GPT-4 And Beyond. It would be a great precursor to this content. It covers in detail how generative AI and GPT-4 are changing the phase of modern-day marketing.

So, let's roll.

What are the new AI-powered Bing and Edge?

Microsoft has been very vocal about its investments in AI recently. The tech giant sees the new AI tools it is developing to enhance search and browsing as an 'AI copilot for the web'. As the company explains, the new Bing and Edge come with the following exciting features:

• Better search with more relevant results and a sidebar for more comprehensive information.
• Complete answers by summarizing the content on a topic.
• A new experience with interactive chat to provide complete answers with better clarity and details.
• A creative spark with all the inspiration, information, links, and sources you need for information gathering.
• A Microsoft Edge with a new look and two additional features called chat and compose.

According to Microsoft, the new Bing also makes search more fun. For example, I tried to search using the below query:

I want to give a welcome party to new joinees at my company. Show me some ideas for this.

I got the following results for it:

What's really cool is that I can choose a conversation style from creative, balanced, and precise. I also added a follow-up query which gave me the following result:

The beauty of this kind of experience is that I have to spend several minutes going through a full article only if I need more details. Since a large percentage of search queries might be answered on the chat interface itself, I get exactly what I need than having to read through an article that might have a lot of information I really don't need. In short, this is personalized search at its best.

Why is Google panicking about this new development?

We need to look at this situation from two different angles - one is technology and the other is information gathering.

From a technology standpoint, incorporating AI into the search experience could be a game-changer. Though the current model of search involves artificial intelligence and machine learning algorithms, they are used for providing the best search listings on SERP. The user has to then spend more time going through the content on multiple websites to get the info he/she needs.

Leveraging AI to 'generate' answers in the form of content on the search interface itself eliminates the need for all these.

From a search experience standpoint, the ability to perform search interactively could be the future, especially given that Gen Z was born into a world of chat-based communication. And even for Gen Y and Gen X, chat offers a more easy-to-use, convenient, and enjoyable experience.

Bing's recent rise in popularity and usage validates these trends. So, Google is very likely to lose market share to Bing if it doesn't build AI into its overall search experience.

To add fuel to the fire, reports that came out on April 17 suggested Samsung considering making Bing the default browser on its smartphones. Though this is not confirmed yet, this led to a 2.95% decline in stock prices for Google in one day according to BestStocks.

In addition to this, as a part of this AI meta trend, a large ecosystem of AI tools is being built (especially using GPT-4, ChatGPT, and DALL-E 2 - all from the same creator OpenAI). HubSpot's introduction of ChatSpot is a classic example of this.

Given all these, there is no surprise why Google is under fire. It definitely wouldn't want to become another Nokia or Kodak.

Project 'Magi' - Google's pivot towards AI-enabled search

Apparently, Google has declared 'code red' within the company and the founders and the CEO have asked various teams to up the ante when it comes to developing AI-based products. This whole initiative has been termed 'Project Magi'.

The NewYork Times says Google has 160 people working on the project. These engineers are reportedly working in what is called sprint rooms to get things moving fast through continuous testing.

Instead of going for a complete foundational rebuild, Google will initially focus on adding new features to its existing search engine while still keeping ads in the mix. The new AI-enabled search engine will involve the integration of Bard (the AI chatbot developed by Google) into search. But the company has work to do given the fact that Bard received mixed feedback from its users.

What constitutes Project Magi?

The primary goal of the project is to make search more user-friendly by incorporating AI. As mentioned earlier, this involves adding an array of features to the existing engine. These include:

• An AI image generator called GIFI.
• A language learning system called Tivoli Tutor.
• Searchalong - a chatbot on Google’s Chrome browser to answer questions about the current web page, similar to Microsoft’s Bing AI sidebar for its Edge browser.

With these new features, the search engine will be able to answer questions about complex topics like software coding - and even write code based on a user's search query.

When will the new search engine be released?

According to The New York Times, the new features of the Google search engine are expected to be released initially to 1 million users in the United States sometime next month. This number would be increased to 30 million by the end of this year.

Embracing the inevitable - the future of search

When the smartphone revolution happened, the move from old keypad-based phones to touch-based interactions was inevitable. When Nokia failed to acknowledge this trend, they were left behind. The same thing happened with Kodak when they failed to adopt digital technology into their cameras.

Many experts look at the advent of AI as a similar phenomenon. And this could pretty well be the time for the long-standing SERP-based search to go extinct. I mean, it is unlikely that Google or any search engine will do away with listing search results completely very soon. But if the AI trendline keeps going higher and higher, there is no doubt that we are going to see a complete disruption in the way search engines deliver information.

That said, I think the following are going to be some of the characteristics of what search will look like in another 5 to 6 years from now:

• Borrowing from the idea of social commerce, search engines are very likely to enable all sorts of financial and purchase transactions on their interface itself.
• Delivering the most relevant results will become even more challenging for search engines. This is because the results obtained in a chat interface usually come with one or a few links to provide more details. How the search engine picks those 1 or 2 websites will be a tough task. This will also make the fight for the top spot more difficult for SEO professionals.
• We jokingly call out Google search as a skill required for jobs. But this could soon be serious and a new role for specialist searchers might arise. As we know, prompt writers for ChatGPT are already gaining some traction (market research companies might find a great use case out of this).
• The accuracy of AI is still questionable. We are very far from perfecting the art. This will likely give rise to solutions that can vet the results delivered by a search engine - particularly in areas like medicine, academics, law, etc. It is also possible to see newer regulations and compliance requirements from government authorities that keep a check on the credibility and reliability of information shared on an interactive chat window for search.
• Just like how many have started using ChatGPT as a search engine, search might take a different meaning altogether as AI gets better at the job. Think about the possibilities of using Q&A platforms (like Quora), content curation platforms (like ScoopIt), information aggregation tools (like SparkToro), or even content writing tools (like Jasper) for search. Where users will go is where they get the best information in the right format.

With advancements happening at lightning speed, it's not so certain as to what will really change (and what won't). It's going to be really interesting to see how search shapes up in the next few years.

What AI-enabled search means for marketers?

As I said, a change in the way search works is inevitable. As it is at our doorsteps, as marketers we need to prepare to future-proof ourselves from going irrelevant. In this section, let us look at what marketers should know - holistically - to keep up in the 'AI race' so that they can continue to produce the results expected of them.

We will dive deep into this along the following topics:

1. Search Engine Optimization (we touched upon this briefly in the previous section, but we will go a little deeper here)
2. Content marketing
3. Demand generation
4. Paid ads

1. Search Engine Optimization

A major shift I expect to happen in the SEO space if AI-enabled search gets widespread adoption is Google adding less weight on metatags and more on the content itself.

Remember those times when metakeywords used to be a ranking factor? But not anymore right? Similarly, I think we are not far from other metatags like title, H1, description, or even alt tag becoming less relevant. The importance of other forms of annotations like schema markup will also fade.

But this is not to say that metatags will not be in the scene at all. They would still have a say in ranking. It's just that they are not going to be as important as they used to be.

The reason for this is that to deliver answers in a conversational manner, search engines will have to analyze content a lot using techniques like NLP (Natural Language Processing).

If this happens, Google's E.E.A.T framework will increase in relevance where Experience, Expertise, Authority, and Trustworthiness will indeed become the core pillars of ranking a piece of content.

The second change I expect is the difficulty in ranking pages, In addition to the reasons we discussed for this in the previous section, more and more websites using generative AI tools for content creation will make it hard for search engines to identify the best content. This opens up new opportunities for marketers to differentiate themselves by creating 'human' content that can put forward fresh and unique perspectives which machines can't do.

The third way - and the most obvious one - in which SEO will change is that websites will have to now focus more on addressing queries than keywords. This is why SEO today needs a holistic approach that includes looking for questions people ask related to a topic (think about what a tool like Answerthepublic provides), creating content that is comprehensive (in terms of covering all possible dimensions of a topic), naturally embedding queries and answers in the content (instead of using schema to mark FAQs), etc.

2. Content marketing

Content marketing is vast. Here, we will try to cover the key changes that are expected due to AI-enabled search than attempting to get into every bit of detail or application of AI in content. Also, content has some overlaps with SEO, and hence in this section, we will be covering only those points that have not been touched upon in the previous section.

The first and foremost trend is something we have already covered in one of the previous editions of this newsletter - zero-click content. With search engines trying to serve information without users having to navigate to a website, the click-based marketing model doesn't hold well anymore. Marketers will now have to focus on building awareness and trust even before bringing users to the website.

The second trend that is likely to be born out of AI-enabled conversational search is making more and more content ungated. The idea of 'hiding the best behind the gate' is less likely to work. You need to learn to give away your best tips and tricks without gating them behind a form - making content more discoverable for search engines. Focus more on creating top-of-the-funnel awareness using your content than focusing on the bottom of the marketing funnel alone.

The third significant shift that could happen is in omnichannel marketing. Today, omnichannel marketing requires you to optimize content separately for each channel (search engines, social media, YouTube). If AI-enabled search becomes a thing widely, it will find its way into social media, video hosting services, Q&A platforms, and the like.

This would mean that optimizing for solving the problems of your ICP (Ideal Customer Profile) will essentially act as the foundational stone of your marketing strategy across platforms. Though there could be slight variations in the way content is hosted and distributed, the overarching strategy remains more or less the same.

3. Demand generation

Wondering what AI-enabled search has to do with demand generation? Here's how.

If more and more information is served within the boundaries of a search engine's interface, your ability to 'impress' a prospect on your website goes down. We are going to see lower traffic to websites but a likely surge in search engine usage and traffic.

What this means is that the probability of a user filling out a form without clear intent goes down. Owing to this, optimizing your marketing activities for demand capture alone is not a good idea. Generating demand for sure needs to be at the top of your list of marketing priorities.

Also read: A Step By Step Guide To Creating A Demand Gen Engine For B2B Businesses

4. Paid ads

When I say paid ads in the context of search, I am referring primarily to Google ads and Bing ads. Reports say that Google's AI-enabled conversation search engine will continue to have ads enabled. This means that your choice of keywords and ad copy (title, description, landing page content, etc.) have to be user-focused.

In addition to this, following are some of the other changes we are likely to witness in the paid ads space in an effort to adapt to the new era of search:

• Using ads to promote educational and thought leadership content will become a key marketing tactic (remember my article on using paid ads for demand gen?)
• Search-based paid ad platforms will eventually pivot to a CPM (Cost Per Mile) or CPA (Cost Per Acquisition) based pricing model instead of relying heavily on CPC (Cost Per Click).
• The line between paid and organic search marketing will get thinner and thinner. While ads will continue to act as a tool to quickly generate and capture demand for your business, the fundamental idea of having to optimize everything for user queries will drive the overall strategy in both forms of search marketing.

Final words

AI-enabled search is for sure picking up. However, as you might have guessed, it will not be relevant for all types of search queries. For instance, somebody who is doing detailed market research or say wanting to learn a new skill would still want to see very detailed information for a search query. This would essentially mean that some part of search will have to still work the traditional way.

Moreover, one of the limitations of making search conversational by leveraging AI is that the user naturally tends to not look at multiple sources to collect and vet a piece of information. This might lead to accuracy issues and disgruntled users.

In short, we are still in a very nascent stage when it comes to AI-enabled conversational search. We need to travel with it for a while to see how things evolve. Nobody could say with absolute certainty what the future holds. So let's wait and see how fast the transformation happens.

Share this knowledge with the world

All my newsletter editions are a result of years of experience, learning, and research. I believe there is a ton of value in the content for founders, CEOs, and marketers in the B2B space. So, if you found it insightful, please do not forget to share this with a friend or colleague who might find it useful.

Also, do subscribe to the newsletter if you haven't already. You will be joining a group of 2400+ learners that includes CEOs, CMOs, and leading marketing professionals.

And if you have any feedback or suggestions to improve my content, please let me know. In addition, if you want me to write about any particular topic in B2B marketing, don't hesitate to share that as well.

As always, until we meet next time, happy learning!