AI will fundamentally change business through the intelligent automation of many routine – and non-routine – tasks companies and executives perform all the time. Those who make decisions about the relationship between AI and their companies have to get this right: AI is easily the most important technology in a generation. Question and answer sessions can help.
Questions And Answers About AI
This is part 2 of a 2-part analysis of how executives should think about AI. Part 1 listed the questions. Part 2 provides some answers. The questions and answers should be filtered through the vertical industry the executive is in and the position in that industry the company occupies. In this case, it’s the marketing/public relations industry.
Here are the questions – and answers – about how every executive should think about AI:
1. Perspective
What is machine learning and generative AI?
Here’s how Gemini answered the question:
With machine learning, “instead of programming it with specific instructions, you provide it with a lot of data. The computer then figures out patterns and relationships within that data, and uses those patterns to make predictions or decisions … like for image recognition (where) you show a computer thousands of pictures of cats, it can learn to recognize cats in new pictures … or language translation (where) a computer can learn to translate text from one language to another by studying millions of translated documents.
MORE FOR YOU
“Generative AI can generate new content like text, images, music, or even code. It's like having a creative assistant that can help you brainstorm ideas or produce something entirely original … it works by analyzing vast amounts of data and learning patterns. Then, it uses that knowledge to create something new that is similar to what it has seen before, but also different and original.”
What’s so special about AI?
AI has the capacity to automate more tasks than you can count which means it can help you save money and make money. It’s a strategic technology that can make you more competitive. Applications can be “taught” and can “learn” how to generate new content, video, code and images. AI is as special as the internet was twenty years ago.
How should our marketing processes and business model be “matched” with AI’s capabilities?
Marketing processes are straightforward and include at least: market research, product development, advertising, social media marketing, sales, customer service, brand management and public relations. What you need to do is decompose these functions into tasks and sub-tasks. You then need to identify the capabilities of machine learning and generative AI and match the two columns. For example, “advertising” consists of commercials, print ads and online everything. Can AI perform these tasks?
What should I be doing about AI right now?
You should immediately develop a Task Force which should launch a competitor analysis, assess in-house talent and begin the marketing functions, activities, processes, tasks and sub-tasks AI matching process to identify appropriate pilots. You should also find resources to assess the contributions AI can make to marketing. You need a team and a budget.
Can I just ignore AI for a few years?
Only at your own competitive risk. AI is moving at a pace we’ve not seen before, so tracking progress is required. There’s no evidence that delay is a smart strategy. You do not want to be the executive that sat on the fence as your competitors exploited all that AI can do.
2. Augment or Replace
What kinds of marketing jobs will AI augment or replace?
Lots of marketing functions, activities, processes, tasks and sub-tasks will eventually be automated. The ones that will be replaced the earliest include marketing content creation, customer service, social media management and personalization. The activities likely to take a little longer to automate include strategic planning, creative direction, relationship building and crisis management. But as always, it’s a function of time.
What kinds of jobs will be “safe” from AI?
For now, lots of marketing jobs. But within a few years, very few will be safe. This is based on the trajectory of the technology and the power the technology continues to demonstrate. But “safe” is the wrong word. You should be assessing all marketing functions, activities, processes, tasks and sub-tasks along a continuum that moves from “safe” to “automated.”
What’s the “schedule” for all this? When will it start?
Within three years many marketing functions, activities, processes, tasks and sub-tasks will be upended by AI, machine learning and especially generative AI. Within five years the marketing profession itself will be reimagined. Within seven years the majority of traditional marketing functions, activities, processes, tasks and sub-tasks will be automated.
How should I manage the elimination of jobs?
You should begin the process by communicating to the team that several things are happening. The first is the assumption that AI will impact the marketing profession in ways yet to be fully defined – but without question very significantly. You should communicate that your expectation is that the company will be able to save money and make money with AI. The second message is that the company will immediately begin to start matching marketing functions, activities, processes, tasks and sub-tasks to machine learning and generative AI. You should also communicate that there’s every reason to believe that the number of employees at the company will shrink over time, as it will with all service companies that sell concepts, words, images, video and advice. Finally, you should communicate that there’s no precise timeline for any of this – except the decision to start matching and piloting today.
Should I model the good, bad and ugly here? How “good” can it get?
It can get really good, if “good” is defined as cost savings and revenue generation. You should develop good, bad and ugly scenarios with some outcome metrics. You should model the range of outcomes with definitions of “good,’ “bad” and “ugly” clearly defined.
3. Examples
Can you give me some examples of marketing activities that AI will augment?
Within the next few years, AI will augment market research, product development, advertising, social media marketing, sales, customer service, brand management and public relations, among other marketing and public relations tasks and sub-tasks.
Can you give me some examples of marketing activities that AI will replace?
Here are just a few: market data analysis, content creation, logo design, customer service, personalization, market analysis and product testing, predictive analytics and social media management. These activities will evolve from augmentation to replacement – from “safe” to “automated.”
Can you show me some specific tasks and processes that AI can improve, automate or eliminate?
You should track the impact that AI will have on blog posts, social media posts, product descriptions, email automation, customer segmentation, lead generation, marketing campaigns, press releases, behavioral targeting, chatbot development and sentiment analysis, among other activities.
Can you identify some tools that enable the examples?
Here’s a short list: Jasper.ai, Copy.ai, Peppertype.ai, Rytr.ai, Stencil, Phrasee, Lemlist, SocialPilot, Tailwind, Salesforce Einstein, IBM Watson Marketing, Synthesia, Midjourney, Stable Diffusion and Sora, among others, powered by machine learning and especially generative AI. These tools should be used to run some pilots.
Can you show how our competitors are using AI?
You need to understand how your direct and indirect competitors are using AI. You should pay particular attention to so-called “new entrants” into the marketing arena. Historically, most technology-enabled disruption comes from new companies. Make sure you look over your shoulder, not just at the same faces you see every year at industry conferences. Track the investments venture capitalist are making a digital marketing.
4. Assess
Do we have the right people to assess AI’s potential?
Since AI is relatively new and since your longer-term employees were trained in an era when machine learning and generative AI were in their applied infancy, you likely do not have the right team to assess the value of AI. You will need some newer professionals to do your assessments (and pilots), or some consultants you can hire short-term to assess the technology.
Are we capable of making objective assessments about AI?
Maybe you are, and maybe you’re not. The key is to find some people that have little or no vested interests in the outcomes of the assessments. Otherwise you might encounter some due diligence sabotage of one kind or another.
What pilots should we launch?
Start with blog posts, social media posts, product descriptions, email automation, customer segmentation, lead generation, marketing campaigns, press releases, behavioral targeting, chatbot development and sentiment analysis, among other activities.
How should we measure the outcomes of these pilots?
Standard metrics for evaluating pilot applications should be used, such as key performance indicators (KPIs) and objectives and key results (OKRs) that measure cost savings, revenue generation, scalability, customer satisfaction, and the like.
Ultimately, this is about alignment with your company’s operational and strategic objectives, and you willingness to accept AI as transformational and disruptive.
When should we begin?
Immediately.
5. Commitment
How should we fund AI initiatives?
With a budget and team large enough to conduct 5 pilots.
How should we organize to exploit what the technology has to offer?
With a Task Force initially that, depending on the results of the pilots, should expand into an emerging technology unit that coordinates with the product/service management team.
Who should lead these efforts?
A senior executive well-respected within the company.
Readiness
There’s no question that AI is here to stay, though there are still companies skeptical about the impact AI will have. The question-and-answer sessions in part 1 and part 2 can help companies decide about how to leverage AI for competitive advantage, or miss the most important technology since the internet. Think about the questions, and answer as many as you can. None of this is going away.