Let me tell you a quick story. I request your patience for this one. Trust me, you will see a point in all of this.
When I started my career as a marketing professional, I realised very quickly that writing was a necessary skill for me to develop. I had good writing skills, but they needed work to write copies that could sell. At that time, I was working in a B2C business. Then I got hired by a global B2B services company to take care of marketing for one of their BUs. Soon I learnt that writing sales copies was not something they needed. They needed someone who could ghost write articles, get industry reports published, and then use them for brand awareness and lead generation through e-mail marketing. Then came social media, which was another ball game. In all of this, I spent countless hours sitting next to highly creative designers to understand the power of design and visual communication.
Sub-consciously, I learnt how the integration of all these things added to become something bigger than the individual marketing assets that they were. They became a brand. They became a perception that our clients and prospects formed in their minds.
Automation of some of these things growingly became important. And contrary to each of us thought, it increased the speed of what we did, creating an empty space for more work to flow in. That taught me the nature of work itself. But that conversation is for another day.
Any marketer with any amount of experience has to learn that up-skilling and re-skilling, learning and unlearning are something they just need to do. It has never been a choice, it has been the only way our work could mean something and hopefully, leave an impact.
And today, prompt engineering is another skill that stands in front of us marketers, waiting to be embraced.
Let's meet prompt engineering
Unless you are retired or have been living in a cave, you must know what prompt engineering is. For the ones who don't know, prompt engineering is a part of generative AI that creates prompts or instructions that guide generative AI models to work to get the desired results. To put it simply, it's the art and science of creating commands and prompts that guide the AI models to generate what you want.
For example, a marketer, working with a B2B technology services company, can create a prompt to generate engaging blogs that will be published on the business website and promoted through social media and e-mail marketing channels. A prompt can be created to guide generative AI to write a 1000 word blog on 'Top 10 challenges in onboarding a knowledge management system'. The prompt could include information on the target audience, targeted keywords for effective SEO, and tone of the language. The prompt could choose to include OR exclude factors that may be relevant for developing the blog.
Let's stay with this example a bit longer. Even before creating the prompt to generate the blog, another prompt could be created to ask generative AI to provide information on the key challenges the target audience might be facing, suggest topics for the blog and which SEO keywords would be most valuable to target through the blog content.
All of these could be 3 separate prompts, or carefully crafted into a single prompt. Based on the response generated, the marketer could continue to ask follow up questions or modify the prompt till the output is acceptable.
The entire process of prompt engineering is fundamentally about translating the user intent into commands that generative AI can execute and generate engaging content at scale.
Applications of prompt engineering helping marketing professionals
The generative AI technology has made every business function highly permeable and is reaching every nook and corner of the function. New use cases around productivity, automation, and insights are cropping up every day. And mind you, they are not some research papers published on the Internet about possible applications of generative AI. Rather, they are quickly transforming into case studies with measurable and demonstrable outcomes with clear contribution to the top of the funnel, middle of the funnel, bottom of the funnel, or all of them.
From a marketing perspective, I think the applications of generative AI and prompt engineering can broadly be classified into 4 buckets - customer experience, growth, content generation, and productivity. Content generation and productivity could be part of a single bucket, too. But let's stick with these 4 for now.
Starting with customer experience or CX, there is so much that AI can help marketers achieve. In the 2024 world, customer journeys have become more fragmented, complex, and difficult to track. Today, text search on Google or clicking a link in an e-mail is not necessarily the starting point in the digital marketing landscape.
In later part of 2022, it was reported that Gen Z are preferring to use TikTok or Instagram as a search engine instead of Google. This was substantiated by a Google VP in a public remark at an event that about 40% of young people use TikTok or Instagram to search for a place to have lunch instead of Google Search or Google Maps.
Before you pass off this as a behavioural pattern exhibited by the younger generation, think about your own online journey today. Does your online journey of information discovery start ONLY with Google Search? If you need latest news, would you go to Google or your favourite news app or X? If you are looking to purchase a pair of socks, are you most likely to search on Amazon or a local eCommerce app or Google?
If the fragmented and complexity of customer journeys was not enough, data protection laws like GDPR and CCPA, and higher walls of ecosystem like that of Apple limit the amount of user tracking. This has made capturing data of user journeys even harder. And the available data does not help paint a complete picture of user profile and preferences.
This is where prompt engineering could help leverage the power of generative AI's language processing and machine learning capabilities to identify patterns in customer and market data to segment and target relevant audiences.
This will help analyze and identify high-quality leads, which could then be used to craft highly personalized campaigns. A/B testing could be made easier with follow-up prompts that generate alternative layouts, ad copies, and SEO strategies.
An important point to note here is that the marketers would need a carefully curated and clean dataset for AI algorithms to analyze.
Moving down from top of the funnel, generative AI can be used for making the sales process more effective. A well-crafted prompt can help generate insights into customer behaviour and preferences which can be used for personalized content and messaging. Using the contextual understanding capabilities, generative AI can be used to come up with follow-up e-mails and also be used to deploy conversational chatbots. This will drive up the customer engagement and chances of conversion.
The benefits of generative AI go beyond conversion right up to the customer onboarding stage. It can help generate content right from sending personalized welcome e-mails to training content that can make the customer experience seamless and smooth. Having a conversational chatbot functionality can help shorten the learning curve of your customer and work better with your business.
These are good examples of how prompt engineering can lead to exponential growth, high-quality content generation and improve productivity of the marketing teams.
What does it take to learn prompt engineering?
Now, you might argue that all this is very nice and good but you are a marketer with little or no prior technical background. You could also be someone who is not very tech savvy, and you don't have to be.
Prompt engineering has a technical, programming side to it. But, from a marketer's perspective, the current generative AI models like Google Bard or ChatGPT do not require any programming experience.
You definitely need strong verbal and written communication skills as generative AI models today require you to interact with them in text format. Some are talking about supporting images and videos as an input or prompt as well. I recently read online that ChatGPT 5 is rumored to have advanced multi-modal capabilities.
Of course, you need to have basic understanding of generative AI as well. Knowing technical jargons like language learning models, neural networks, and machine learning will make it easier working with AI. Knowledge of these terms will also help decide which generative AI model will be able generate the best possible output for you.
And finally, having data analysis experience will also be useful. Which, as a marketer in 2024, is something all of us are used to, thanks to all the digital marketing platforms and tools. Having an anlalytical understanding will help identify any possible bias in the output generated and address with modified prompts.
Prompt engineering is ready for a high five, are you?
Most of the marketers today either have these skills or have access to online learning courses, many of which are free, to learn them. Let's look into content generation bit deeper for a moment. Prompt engineering has already proven to be an effective skill in developing engaging content that can help prospects move through all stages of the funnel, nurturing relationships and providing them with a consistent customer experience.
There is a new generative AI-based tool being introduced in the market at such a rapid pace that it is hard to keep up. As far as I know, there are at least 20 tools for video generation like Veed, Runway and CapCut; more than 30 tools for generating text-based content like ChatGPT, Rytr and Grammarly; at least 20 tools for creative design like Canva, Midjourney and Adobe Express; 10+ chatbot tools like Google Bard, Bing AI, and Jasper; and 20+ marketing tools like Brevo, ConvertKit, and Adobe Sensei.
Like I said, these are just few of the names and indicative numbers in a growing list of generative AI tools that are available for marketers today. Frankly speaking, if as a marketer you are not using prompt engineering you need to start learning it now. If you have learnt it but are struggling with use it effectively, keep trying.
I understand if you find it hard. Just as I had to keep learning a new skill throughout my career, I gave it a go multiple times and failed at it. I don't have any shame in admitting that I am particularly not great at it myself yet. May be I'm old-school to an extent. The only thing that keeps me going is the realization that I had when I started playing with ChatGPT, the marketers who will learn prompt engineering and embrace generative AI will thrive because this technology will continue to get better and more relevant for marketing organizations.
Learning something new is always hard because unlearning an old skill is part of the parcel. But the beauty of learning prompt engineering is that it can benefit marketing of your business and reach your customers in ways that you really want it to. So, avoid falling for those doomsday stories Hollywood movies have been telling all this long. Generative AI has already raised its prompt engineering hand, it's up to you to give it a high five.
Comments