Welcome to the fifth article in my ‘Leadership in Marketing Cloud‘ series. I created this series to address a major gap in reference to and documentation around leadership inside the specific context of Marketing Cloud. When I searched through the community and official places, there was nothing really available!

Please reference here for all the articles currently available for this series.

In the last article, we went over Conflict Resolution and Tough Conversations. This was quite a heavy topic and gave a peek into the ‘darker’ side of leadership. I wanted to shift the narrative and address a topic that is highly relevant and highly impactful right now: AI (Artificial Intelligence) and all the opportunities and innovations associated with it. I know previously this was supposed to be an article around Delegation and Enablement, but I felt I had to interject my regularly scheduled post topics with this topic first.

As this topic is so large and has so much to consider, I will be splitting it into 2 articles. This one will focus on what AI is and should we use it, while the second will go into impacts and repercussions as well as a general overview on what we, as leaders, need to consider around AI.

AI is the new hot thing across many fields, most especially marketing and technology. Although AI is not new, the huge advancements that have been coming up recently and all the new tools like ChatGPT and Salesforce’s new MarketingGPT and more have given so many new opportunities for leaders to consider. Although most of these can be a net positive option, each option does also contain inherit risks, significant costs and potential pitfalls. As a leader, it is our role to protect our people and our business while also looking for the most innovative and efficient solutions overall.

First, let’s take a look at AI and what it is.

(also as a fun little note, each image in this article was created via Generative AI)

Artificial Intelligence

What is Artificial Intelligence (AI)? Well, that is something that is extremely hard to define, especially with all the new advancements. It changes all the time. In general, AI is broken down into a simple definition of technology that allows a computer to think or act in a more ‘human’ way. It can include the creation of computer generated content, actions and assets utilizing prompts along with learned logic and processes, and not explicit instructions.

Now there is a lot to go over in AI to truly get a strong grasp on what it is and how it can be used, etc. but for this article I am going to keep it to a high level and concentrate on Generative AI which is most relevant to us. This includes ChatGPT and Salesforce’s MarketingGPT as well as a ton of others.

Generative AI is focused on the creation of new content that comes from algorithms such as ChatGPT. Essentially it uses two neural networks competing against each other to generate new content based on the input it receives. These algorithms perform statistical analysis at scale, processing millions of bits of data at incomprehensive speeds (this sentence stolen from the great Genna Matson). This content provided can be defined as not just what we usually mean in marketing as copy or images, etc., but also as analytic and predictive insights and decisioning.

There is so much more to AI than just churning out visuals, code and copy. It can provide personalization at scale as well as campaign optimization and strategic insights that go well beyond the scale, we can achieve utilizing other current tools.

Now that we have a general idea of what I mean by AI, let’s explore what it can do.

What can AI do?

Generative AI can do an amazing amount of things. It can:

• Create high-quality and relevant content at scale
• Assist in personalization
• Streamline Copy and Design
• Enable more robust chatbots and personal assistants
• Assist in market research and analysis
• Social listening and sentiment analysis
• Predictive analytics
• and more…

The core places that we will look at in this article are involving the creation of content such as scripts, copywriting, images and documentation as well as the predictive and analytic aspects. These are likely to be the most impactful aspects for those in the SFMC community.

AI has reached a level that with simple prompts, you can have it code an entire email from scratch, including ISP hacks and platform specific scripting. This is huge for many places that may not have easy access to email developers or only have a limited output available. It can create all kinds of new opportunities and capabilities with little to no extra cost/efforts. It provides a scalability to your organization that you likely could not match in any other way.

That being said, as I will go into in the next section and the second article, it comes with a lot of caveats. For example, the coded email provided may not be as accurate, optimized nor fully capable as what would be produced by a human developer, which can cause major issues if not monitored.

You can also use Generative AI to generate visual assets and copy for these emails, to create hyper personalization and content agility, as well if you want your own developments and/or designs instead of what is created completely by the AI. This allows you to get past the bottlenecks of utilizing overworked graphic designers or copywriters on something that can be a simple task and instead allow them to concentrate on the larger and more impactful projects. This also provides a scalability to personalization and customization due to the quantity of modular content you can create.

There are so many accelerators and efficiencies in content creation that I could go on for days around the possibilities here. Rather than do that, and make a long article even longer, let’s move forward and take a look at the predictive and analytic aspects.

As we have already seen to some degree with the Einstein capabilities in SFMC, there are some fairly robust predictive intelligence options that can come from AI. This can include such things as send time, real-time and predictive personalization, content and copy recommendations and more. All of it is based on your data and the data collected via tracking and interaction. This can open up many doors to enable your campaigns to be the most successful and impactful as they can be without the need of manual data analytics and predictive models being built, run and reviewed by your analytics team.  And of course, with the more data you provide it, the more insights and recommendations it can provide. You want to find out what women over 45 interact with most from all your products? Simple prompts into the AI will get you a robust analysis returned in minutes.

With that being said though, we need to consider the old saying of “Garbage in, Garbage out” as without high quality data and learnings being fed through, what is produced is not likely to be of the quality you expect it to be.

Should we use AI?

Short answer is yes, but as leaders we need to do a lot of research, exploration, analysis and more to validate when/where/what/why/how we will use it.  AI is the future and will have a huge impact. Think of the amount of impact and change that came from the introduction of digital marketing so many years ago.

This change will have impacts not only to the way we run marketing but also the jobs and skillsets needed to successfully run a marketing plan. These changes are still a bit ahead of us though as AI has not reached the maturity level that we can place full trust in what it returns back – nor have we fully fleshed out the requirements needed on our end (as the meatbags) to best enable and teach/guide the AI on what we want or need. This combined with inherit bias and learning guard-rails can have unintended future consequences that we have not considered yet. (Check out the story of Tay and Twitter here for reference – please note that the messages that the bot shared contains some vulgarity that is not likely work appropriate, etc.)

We can easily see a large boom on needs for skills relating to database management/architecture as well as QA and content review. And, of course, we will need many engineers and developers to upkeep, innovate and maintain the AI engines and neural networks to help it continually learn and get better. This will have huge impacts to resourcing and operational procedures and processes. This all being said, although there will be a shift in where the resources are needed, it will not in any way replace jobs like copywriters, developers, designers or any of the other careers. These specialized skills will be paramount in ensuring the success of AI and the success of the company in general. Although the number of resources will be reduced for those roles, the required skill level and importance of those positions will likely be higher.

As AI is still on the ‘cutting edge’ and not matured, there is a good bit of risk involved with utilizing AI. So, this is where we, as leaders, need to review the feasibility and capabilities of AI based on:

• the context of your company
• how much your company is willing to invest
• length, cost and impact of implementation on current state
• what you want it to do
• what ability you have to enable it
• what other applications or software is needed to implement
• quality of your data and current tracking information
• security and risk mitigation you will have in place
• what review and QA process you can have to ensure quality

If even a single one of those is not able to be met, it can very quickly become a ‘no’.

Of course, though, the high level of cost not just for implementation but also continual maintenance and updates can skew your return on investment or cost to value analysis to levels that it won’t make sense to invest right now. This along with any other investments in required applications or software can easily price out many organizations in the current market. There are piecemeal parts of AI that you can purchase or utilize third-party applications or services for those not willing to invest, but those also come with additional risks, considerations, limitations and more.

I do also want to bring up the ethics of AI and the transparency with customers and subscribers on the usage of AI. Both are paramount to continuing the trust you have already built with your customers as well as building more trust and comfort by open disclosure and ethical behavior. AI, although a powerful tool for good, can be used in malicious or otherwise unethical ways very easily, and honestly, there is not much in place to prevent or protect against this. By keeping high ethical standards and utilizing the tools appropriately, we can better form AI to be the great technological advancement for the betterment of the world and not a tool that can throw our world into chaos.

I think the 4 major factors to focus on as a leader in SFMC are:

• Cost and ROI
• Security, Legality and Risk Mitigation
• Data and AI enablement
• QA and Review

Without strong support on the above factors, you can see some major repercussions for using AI which can cause massive blowback for you and your company. There is so much to consider, but if one of these factors are not met, then it is likely that utilizing AI for your company or team should wait until you can figure out a new strategy that allows that factor to be met.

This is the end of the first article! The second article will dig into the impact and repercussions of using AI as well as an overview on what, as a leader, we should be doing in regards to AI. If you have any thoughts or feedback on this, please let me know and I would be happy to discuss it with you.