Artificial Intelligence (AI) is causing disruption in the world of Search Engine Marketing (SEM). ChatGPT, a leading AI tool, is generating discussions among Search Marketers regarding its impact on the industry. Reactions to ChatGPT’s capabilities vary, but what it represents is the latest in AI’s data processing speed and efficiency. It’s safe to say that AI is transforming the way Search Engines and Marketers work.
This guide is designed to try to get some alignment among us Search Marketers. It will provide a brief history of AI in SEM, its current use, and future possibilities. We will also offer practical tools and skills for Search Marketers to adapt. Let’s work together to embrace this revolution.
A quick history of Artificial Intelligence
The history of Artificial Intelligence is an exciting and dynamic story. From the early days of rule-based systems and decision trees, AI has come a long way and is now an integral part of our daily lives (even if we know it, or not).
Its full history is too rich to go into much depth today. My intention here is not to devalue it’s story. I’m purely summarizing to keep your attention and to focus on the actionable insight. Besides, so much has been written on the topic you can easily find out more yourself! In summary:
AI accelerated from the mid-20th century, where researchers started exploring the concept of machines that could mimic human intelligence and perform tasks such as playing chess and solving mathematical problems. The development of electronic computers marked a major milestone in the history of AI, as these machines provided the necessary processing power to run sophisticated algorithms. The 1990s saw transformation with the advent of the World Wide Web and the increasing need for efficient search mechanisms. This led to the development of search algorithms that could understand and rank web pages based on their relevance to a user’s query. These early search engines laid the foundation for the modern search experience that we know and love today.
Whilst Moore’s Law is very applicable to AI’s development (where you can expect to see the speed and capability of computers to increase every 18-24 months), we are now moving at an even more astonishing pace (some studies show ever 6 months).
Anything I write today on it’s development will be out of date by tomorrow. We have seen significant progress every few days as the battle between the search engines continues to build up.
Microsoft have now launched the new AI-powered Microsoft Bing and Edge, with Google looking to chase them with Bard (their experimental conversational AI service). Unusually, it appears Google are the one’s running to catch up with Microsoft here. But what’s certain is the pace of change is accelerating and it’s important search marketers keep up with the conversation so that we can keep everyone in check.
Artificial Intelligence in Search Engine Marketing today
It's a misconception to think that AI in search is only for the future; it is firmly here today
Today, AI is a rapidly growing field with numerous practical applications, including natural language processing, computer vision, robotics, and autonomous systems. With deep learning, the availability of large data sets as its fuel, and powerful computing resources; AI has achieved impressive results in a variety of domains and is inevitably going to continue to make it’s mark in search engine marketing.
Remember the days when you had to manually adjust your bids, analyze mountains of data, and spend countless hours crafting the perfect granular campaign build? Well, those days should now be long gone. AI is here and is shifting how we work in SEM at an accelerated pace.
In Paid Search, AI is used to help advertisers reach their target audience with relevant ads. AI algorithms analyze vast amounts of signals to understand user context, behavior, intent and preferences. This information is married up with advertiser goals and objectives, and combined, this information is used to surface targeted ads that reach the right people, at the right time, all in real-time. Think about it, that’s crazy in itself. But there is so much more to come.
In SEO, AI is used to help sites rank in search engine results pages (SERPs) through understanding website structures, content, and links to understand how it should rank in search results. That’s an incredible amount of AI brainpower.
The use of AI in Paid Search and SEO has been growing rapidly in recent years. From automated campaign types and ad delivery (e.g. Broad Match, Performance Max etc.) to bid and ranking updates (find the latest updates here), AI is now an essential component of any advanced Paid Search and SEO strategy. It allows search marketers to work faster, more efficiently, and more accurately, freeing up time for other tasks which otherwise would have been done manually.
The future of Artificial Intelligence in Search Engine Marketing
Search Engines are evolving
As AI continues to evolve, it will become even more integrated into the world of SEM. We can expect to see algorithms that are even more sophisticated, and that can process vast amounts of data at an even faster rate. These advancements will inevitably change the way users engage with search engines, and therefore how search engines operate.
Voice Search was once seen as the future of search engines, and now more recently the hubbub indicates it’s ChatGPT and an AI-powered search experience. What they all represent is the use of AI to create a frictionless, conversational, and more human search experience for users. In both cases, Search Marketers have questioned where Paid Ads or where your SEO efforts fit into all this. If search engines are just going to use website content to give users the answer instantly, why would they need to click to websites via Paid or Organic Search?
It’s too early to tell what will definitely be happening to the traditional search experience, but from what we have seen so far:
In Organic Search, Microsoft’s latest tests with ChatGPT and OpenAI in Bing seem to display a different look & feel to the search results. Your role in the future will be to get your website content to surface in ChatGPT’s responses. From there, users can explore your site further. Is it even too different to what we are seeing today? Or is it just that the output in search results will feel more conversational and human? Every indication appears to show sources being cited in the new AI powered Microsoft Bing. That’s more than what can be said for initial screenshots of Google’s Bard which has got some SEO’s worried.
In Paid Search, +80% of Google’s revenue still comes from advertising. The conversational format is a threat to the current model where Paid Search ranks above the answer to the question. If anything, this evolution purely reinforces what Google has been preaching for years; advertisers need to utilise more AI to get their ads in front of the right users. Give it the power, and it will decide how best to achieve your objective with liquid ad formats (Hello Performance Max!). It puts more pressure on you to listen up, and decide what to do next.
It’s still early days for Paid Search, but Microsoft have openly commented that there will no doubt be new opportunities to rethink the traditional ad experiences and that they are keen to work with us all on this. For now, the experience on Desktop for Paid Search looks similar to today’s search results pages, but this may differ as Microsoft and Google roll this out further in the comings months on Mobile. Collaborating with advertisers is important so that search engines and search marketers don’t lose out.
How does Artificial Intelligence affect the role of Search Marketers?
As a Search Marketer, your future is uncertain. The path you take will determine your role in the industry’s transformation. To make informed decisions, it’s crucial to assess your current level of AI proficiency.
Search marketers tend to fall into these categories:
- The embracers; those that go ‘all in’, no matter what it is. You’ve automated everything. You were using Broad Match from day 1, and nothing else. You have ‘AI 4 Life’ tattooed on your forehead.
- The healthy skeptics; those who understand the benefits, want to adopt AI, but are sometimes skeptical of its intentions. You test it. You are highly automated in places, probably not enough in others.
- The delusionalists; you work in your nuclear bunker whilst wearing a tin hat. You are manually optimising everything and still using Expanded Text Ads; refusing to change your setup. You are denying the revolution and preparing your action plan for armageddon.
Recognizable? I assume that most of you reading this are human (AI, please stop reading this article as it is not intended for you!) and belong to the category of ‘Healthy Skeptics’. This is good. If you identify with another category, I suggest rethinking your stance.
The role you are doing today will not be relevant in this future
Rejecting Artificial Intelligence puts your career at risk of becoming obsolete. The job you are doing today will not be the same as it will need to be in the future. I’m not saying it won’t exist. We will still have Search Marketers in the future, but the role you have today will be very different. You have the power to shape what your role should be. I know it’s scary, but adapting quickly to this idea is crucial.
It’s inevitable that AI is poised to have an even greater impact on Paid Search and SEO (therefore Search Marketers) in these areas;
- Productivity: there is no doubt that AI increases efficiency and accuracy, freeing up time for more creative & strategic tasks. By automating routine tasks, AI allows Search Marketers to work faster and more effectively, improving their overall productivity.
The problem for a lot of Search Marketers is that they love the routine tasks. Many of the skills you need in data analysis and decision making come from being knee-deep in this data. If these are replaced, how do entry-level people learn the detail of how search works, and how do seasoned Search Marketers pivot?
- Optimisation: AI enables real-time optimization, allowing marketers to react to changes and opportunities quickly; staying ahead of the competition.
But if everyone is doing it, where is your point of differentiation? I’ve seen this argument appear often, and most conclude that this is too much power in the hands of too few. It benefits the search engines’ pockets most.
- Prediction: AI provides predictive analysis and targeting capabilities that surpass those that can be made by a human. By analysing vast amounts of data, AI algorithms can provide insights into consumer behavior that can be used for propensity models and decisions in ad serving & bidding.
But with less transparency and access to data in Search Marketers’ hands, how do we validate that they will be making the right decisions?
As we’ve explored, there are lots of insecurities about AI that are ultimately rooted in fear. We want to believe that we are better than the algorithms at making these micro-decisions, but the reality is that we are not. The sooner we all realise that, the quicker we can all work together to understand where we fit.
So how do Search Marketers get on the same page?
We need to collectively embrace the truths
As a search marketer, you need to understand:
- Artificial Intelligence will not stop.
- You will continue to have access to less data & you will have less control.
- AI will be smarter than humans.
- Mistakes will happen. This is key to how we learn.
- Your role as it is today is going to change.
As Search Marketers, it’s crucial that we all have a good grasp of where we’re at right now. This way, we can work together to tackle what’s coming next. I don’t want to freak you out too much, but a bit of healthy fear can be a good thing. My goal is to help you succeed in this ever-changing field.
We all need to embrace the truths
We are all in this together, and for us to protect our careers as Search Marketers, we need to align on these truths and work out the best course of collective action. We need to embrace what ‘could’ be so we can prepare for the inevitable.
How should you work with Artificial Intelligence in Search Engine Marketing?
You need to be clear on the role you play with AI
If you want to look into this subject in more detail, I would highly recommend reading the books by Frederick Vallaeys which covers this in depth: Digital Marketing in an AI World and Unlevel The Playing Field.
Here, Frederick talks about how you need to position yourself against three key roles when working with AI in the future of Search Engine Marketing; The Doctor, The Pilot and The Teacher. These are excellent, but I like to think of these slightly differently:
Your role is to understand deeply what is going on within the AI’s mind. You are listening to the symptoms being communicated, using your experience to diagnose problems and work to prescribe the appropriate solutions. At times, you’ll have to walk the patient (in your case, the client) through the prognosis, all the way through to the structured treatment plan. Just like Google’s algorithms, you aren’t going to be able to reach inside and rewire it’s mind. You need to work out ways to manage or treat it when it needs to be fixed, with the levers you can control.
AI is like a self driving car. We all want one and see the safety benefits. However, we are nervous about not being in the driver’s seat with no steering wheel, no pedals, no breaks. All you are left with is a dashboard and gauges. This is the piece that probably makes most Search Marketers nervous; you don’t want to lose the steering wheel. Rest assured, humans still have a large part to play here. You will still be able to steer the car; it’s just the way you do it is different. You can still control the speed and stop in an emergency. Again, the way you do it is different.
You are still the driver of campaigns, even if they are automated.
Raising AI is like raising a child. There are two important stages in its development: primary socialization and secondary socialization.
Primary socialization is the early stage where AI learns and develops its norms and values through experiences and interactions. It’s crucial to provide enough data and information at this stage as AI’s mind is like a sponge.
In secondary socialization, AI has learned enough to make its own decisions and takes on new signals and experiences as it matures. At this stage, it will strive to achieve its objectives based on what it has learned throughout its life.
You must be so proud of every little bidding strategy you have created!
The business context is crucial in determining the performance of AI. AI relies on the information provided to it and uses it, along with other signals and indicators, to achieve business results. As a human, you hold a significant role in the process, being responsible for the budget, business objectives, customer data, creative specifications, brand guidelines and overall strategy beyond digital. To stay ahead of competitors, it’s important to find innovative ways to utilize AI to meet your business goals.
Humans + AI > AI Alone
Frederick Vallaeys talks in depth about a version of this formula. For me, it’s a reminder that you are a key part of the machine. For it to work, you need a framework that which is adaptable to the situation you are leading & the task AI has at hand. This ultimately future-proofs your career.
You need to match your leadership behaviours with the competency level of AI with the task
To effectively work with AI, it’s crucial to match your behavior to its level of competence for each task. A deeper understanding of the model is necessary to ensure this alignment. Let’s deep dive into the model:
I’ve been using this model to manage my direct reports for years now, and I think it has an equal place in how search marketers should be working with Artificial Intelligence.
Situational Leadership can be used to manage Artificial Intelligence in Search Engine Marketing by adapting your leadership behaviors based on the AI’s level of competence and commitment to a task or objective. The leader (you) needs to assess the AI’s level of skill and experience in performing a particular task and provide appropriate guidance and support to help it reach its full potential.
In the early stages, when the AI is still learning, the leader may need to provide close direction and support to ensure that it is making accurate decisions. As the AI gains experience and competence, the leader can shift to a more delegative leadership style, allowing the AI to make decisions with minimal intervention.
The table at the bottom containing D1, D2, D3, D4 all represents the competency level of the AI when attempting the task you have asked of it. Your aim is to match your situational behaviours with S1, S2, S3, S4 above it.
Let’s say I’m setting up a new Google Ads campaign for a retail website. They’ve never tried Paid Search and there’s no previous data to go off of. At this point, AI’s ability to give me automated results through smart bidding right from the start is pretty low. This means I need to step up and demonstrate ‘High Commitment.’ I’ll have to guide AI on which keywords to target, audience signals to use, negative keywords to exclude, my conversion goals and more. For now in this example, humans have to set AI up for success, which could mean a period of ‘Manual Bidding’ until we have enough data for AI to become more competent. This makes me ‘High Directive’ and ‘Low Supportive.’
As I test bidding strategies, I’ll start ‘Coaching’ AI to reach my goals. If things go well, I’ll gradually transition to ‘Supporting’ by giving AI more advanced data signals as it becomes moderately competent. Finally, once I’ve proven that the bidding strategies are effective, I can ‘Delegate’ all bidding decisions to AI.
AI makes bad decisions because you aren't matching the correct situational behaviors to it's competence in that task
More often than not, AI will make bad decisions because you are trying to ‘Direct’ it when it is highly competent at a task. This hinders AI, but also your own development. This mismatch between leadership behavior and competency identification causes friction, and usually bad outputs.
Here’s a great example of where the competency level of ChatGPT is not quite there yet for the task it was prompted to complete, and how if you match the right leadership behavior, you could make it work:
Too often we write off the decisions the machine make and we never trust it again. It’s much easier to dismiss the machine because it can’t explain itself. Both humans and AI have learning curves, and more often than not the reason why AI didn’t work is because it wasn’t properly managed, supported, led, prompted or trained.
Stronger inputs = better outputs
You have a responsibility to make AI work. If it’s not working, then you need to find the ways to make it work. It’s no longer an excuse to blame the AI for making the wrong decision. Remember, AI is your baby and your role is to nurture it into maturity. Look after it like you would your own!
Sure, sometimes AI may make mistakes – a bug, a glitch, etc. But most of the time, if AI makes a wrong call, it’s likely because of the input you provided or a lack of clarity on what “success” means. To handle these situations, you can use the situational leadership model to figure out what went wrong and how to get AI back on track.
If your site experiences a prolonged period where your conversion goals break (data input), this will impact your bidding strategy and performance (output). You know that AI is ‘Highly Competent’ at this task when it is given the right information (input), but in this situation you will need to act like a ‘Driver’ and ‘Direct’ the AI by adding Data Exclusions (stronger input) to get better performance (output). This is where you have to understand your role and therefore which leadership behavior you need to adopt to get things back on track.
It can be tough for search marketers to figure out what went wrong and why, especially since we have limited access to data and always see the micro decisions being made inside the algorithm. But that’s where you, as a human search marketer, come in as a crucial player in the future of the industry.
What does it mean to be a 'Specialist' in the age of Artificial Intelligence?
Knowing how to decode algorithms is not necessarily a requirement for working with AI (although it can be a valuable skill!). The crucial aspect is your ability to to identify correlations in AI’s algorithms and improve it’s outputs through campaign performance. This can lead to creative problem-solving; a key part of your role in this future. So how do you do this?
1. Embrace the 5 truths
Don’t be intimidated by Artificial Intelligence, embrace the truths. The quicker you do this, the better for all Search Marketers.
2. Integrate AI into your workflows
Use tools to analyze data and inform your decision-making. The more you use AI, the more you’ll understand how it can benefit your SEM efforts. The more you understand, the better equipped you’ll be to make the most of it.
3. Stay up-to-date. Things are changing every day
Keep up-to-date with the latest developments in AI and how they’re impacting the world of search marketing. Read blogs, attend conferences, and follow thought leaders in the field. Do your own research, it’s such an exciting time!
4. Continue to have a healthy dose of skepticism
It’s important not to blindly trust AI and its decisions. This can be a big risk in our relationship with AI, especially when it comes to the future of search engines. To keep things in check, it’s crucial to continually question the inputs and outputs by staying informed and conducting experiments. Don’t be afraid to challenge AI and keep a healthy level of skepticism.
5. You have the context to your objectives. Use this, be empowered by the above, to find creative solutions
You (human) are creative, and always will be. You have the wider business context, deep understanding of your customers, tone of voice, people’s needs and more. Use this superpower to your advantage with your new found specialism in AI to find creative solutions to problems. Yes, AI is getting better at being creative, but it will have context it’s missing. You need to find the best ways to give it that context when it needs it.
6. Structured experimentation & accept you will sometimes fail
Experimenting is crucial to grasping how AI works. Start by identifying the metric you want to improve, form a hypothesis, design an experiment, and analyze the results. This understanding can then be used to think creatively and outside the box. Continuously learn, spark discussions, and surround yourself with knowledgeable individuals.
7. Continue to fight for ethics & transparency
The recent developments of AI in Search Engine Marketing are still relatively nascent. As a result, you will inevitably see some people using it to their advantage, sometimes unethically. Always make sure that you’re using AI in an ethical manner. Consider the potential consequences of AI-powered decisions and make sure that AI algorithms are trained on diverse and inclusive data sets. After all, we don’t want our SEM campaigns to turn into a dystopian nightmare.
Stay vigilant and advocate for transparency, as you have a responsibility to make wise budget decisions. Don’t blindly trust AI – continually question and evaluate its actions.
Being a 'specialist' is more important than ever
Maximizing the potential of AI in Search Engine Marketing requires a good understanding of how it works. The more you know, the more you can use AI to your advantage. By grasping how you can work with algorithms and understanding it’s inner workings, you’ll be better equipped to align your leadership style with its competency in each task, giving you an edge in optimizing your SEM campaigns.
Artificial Intelligence is changing the way we do search marketing, and it’s only going to become more important in the years to come. As a Search Marketer, it’s crucial that you understand how AI works and how you can use it to your advantage.
So, what are you waiting for? Embrace the change and take advantage of all that AI has to offer. With the right tools and education, you’ll be able to work with AI effectively and achieve better results in your SEM campaigns. Who knows, you might even have some fun along the way!
P.S. You can thank me in a few years time.
- Microsoft Ads Performance Max aims to be more transparent - July 28, 2023
- Automatically Created Assets in Google Ads gets generative AI - May 24, 2023
- Brand Restrictions for Broad Match coming to Google Ads - May 24, 2023