What’s Your Use Case?
Use cases are the new creative brief when it comes to marketing AI implementations.
Marketers charged with implementing AI or who are actively embracing these tools for their organization should learn how to build use cases. Good news: A parallel can be drawn between use cases and creative briefs.
Agencies and many marketing organizations require creative briefs to provide necessary information — including goals, inputs, and deliverables — for advertising and content projects. Like the creative brief, use cases help marketing organizations define AI projects with specific outcomes and resources in mind.
Use cases require marketing organizations to think strategically about AI, as opposed to haphazardly experimenting with widespread tools, hoping that they create success. Smart marketing organizations will take a measured strategic approach to implementing marketing AI with use case-driven pilots and implementations.
By their very nature use cases create a framework for measurable success, such as ROI, cost savings, work efficiencies, and other meaningful outcomes. They outline the steps necessary to be successful and prompt necessary thinking about resources, KPIs, and ultimate success.
Move from Panic to Intentional Action
With so much speculation and an incredible PR battle occurring between big tech giants, it is easy to become distracted, perhaps even panicked about the arrival of AI. This reaction — evidenced in conversations on LinkedIn and speculative media coverage — is inspired by expected dramatic changes caused by AI, including possible job impacts or worse.
First, concerns about AI are reasonable, but I think the overall marketing space — particularly, the content marketing world — is suffering from a healthy case of doom scrolling. Anyone who has used generative AI extensively knows there is a very real need for human guidance.
It’s been said before (see above No Brainer AI Podcast with myself and Greg Verdino) on job loss: Those most likely to have their actual jobs cut are the ones who refuse to adopt the tools. They will be replaced by peers who fully embrace and master marketing AI.
Of course, faced with such a fate it is understandable that people would rush down the AI hill with their fiber optic light sabers out, ready to defend their future jobs. ChatGPT-generated content, here we come! But is that good business?
Moving with intent makes more sense. Each business is unique with its particular approach to its market. Why wouldn’t you try to intentionally create a success that will impact that business with its unique desired outcomes? This is strategic marketing, after all.
Identifying a Use Case
Let’s be clear marketing AI has been around for years. Online ad buying is perhaps the most mature implementation. Still, analytics packages offer insights into basic automation and creative tools have been adding AI tools like sky replacement for several years now.
What’s different is the speed and scale of possibilities. Taking time to examine your actual marketing operations and finding a place where you want more results, may not have the human resources to do it, and the task is repeatable makes a great deal of sense.
Consider a few marketing examples now made possible by generative AI:
- Repurposing content from one medium to the next, for example, creating a narrative script for a PPT
- Versioning messages for individual content forms
- Versioning content for unique audiences
- Summarizing and extending existing content to create new macro pieces, such as white papers
- Generating secondary and tertiary logo schematics
- Extending brand across icon libraries and other presentation arts
These are easy and somewhat obvious use cases. When you build your use case (yes, suggested elements are in the next section) don’t forget to add resources, goals, and KPIs, and make it time-bound (with some flexibility). Your use case should have measurable and intentional SMART goals.
Then consider all the ways marketing organizations can support their sales organization using generative AI, much less ML, when it comes to critical sales functions like competitive and market intelligence, targeting, prospect research, smart content generation based on curated training data, etc., etc.
Typical Features of a Use Case
Use case formats vary from organization to organization, but some clear elements can be used to help forge your marketing case for AI. Let’s take a look at these elements that should be included in a well-thought-out use case.
Solve a Real Marketing Program: Rather than experimenting with AI for the sake of experimenting, solve a real marketing problem or need. These are real challenges many marketing organizations face. Find your weaknesses and address them.
For example, my staff is unable to conduct comprehensive research about our market and potential competitors. We are losing opportunities and market share to better-informed competitors. Can I use AI to research targeted industry and competitor websites and analyst reports to provide better market analysis for sales and marketing team members?
Or our team is unable to do a complete job of versioning marketing campaign elements, and our campaigns fail to reach all of our potential customers. Can I use generative AI to create different content elements to extend the campaign into different media and also version them?
What’s the Goal?: Is there a real outcome expected that aligns with the company’s overall marketing goals? Notice both of the above have an implied or direct goal: Win more deals and market share, and extend campaigns to generate more opportunity.
If you don’t have a goal, drop the cool tech rock fast before you invest significant dollars in an AI implementation that won’t help you achieve your mission.
Don’t Forget the 800 Pound Data Gorilla: Do you have or can you acquire the data (documents, research, images, data, financials, etc.) necessary to train your AI and produce an outcome? Make no bones about it, a great deal of the work will be finding and cleaning or formatting data (yes, there are AI that can help with that).
Include Your Users: Silly thing, but the output needs to help people do their job better or help customers. Is there a core user group(s) within marketing or the larger organization that will use the AI implementation? Are you considering an externally facing application like a chatbot for customers? Understand who your users are and include them in your model so they can provide feedback as you iterate your AI model.
What Barriers Exist Internally?: Is your marketing culture ready to embrace AI? What are their current behaviors and what changes will they need to adapt in their workflow to make your AI project work?
Are there larger organizations and stakeholders that can prevent challenges? For example, do you have to adhere to data privacy regulations or other legal concerns, finance approvals, or IT implementation restrictions to consider?
Calculate the Costs: How much will it cost to build your AI? Consider algorithm licensing, talent (internal or for hire), data costs, and any related cloud, and server costs, for example. Does the end justify the means? Ultimately, most AI implementations need to make money, improve customer service, or reduce costs.
Of course, if your problem is relatively simple and low cost this may be an easy fix with low risk. For example, versioning long videos for various social media can be done dirt cheap and cost minimum amounts of money with a tool like Momento. fm and 10 hours a week of a junior staffer’s time.
Cleaning up your marketing funnel to better understand your customer journey for all five of your project lines, is going to be a major data science lift that’s going to take time and resources.
Scale Your Risk: AI is no surefire bet, so low-risk projects with smaller outlay are good ways to start, then scale accordingly. Of course, by now more marketers are becoming accustomed to AI tools, so it may be easier to take a larger risk to address problems.
Consider how much risk you can stomach, and be willing to fail fast. When you succeed, because you will continue to experiment, take your lessons learned and scale accordingly.
Conclusion
Intentional strategic approaches to implementing generative AI — and AI as a whole — will yield better results for marketing organizations. Building use cases — even for minimal implementations — creates a better chance of success and eliminate wasted efforts on experimentation for experimentation’s sake.
Further, it will provide a competitive advantage as your marketing organization moves towards AI maturity. Meanwhile, your competition will continue to play and experiment with ChatGPT randomly because they read about it in their business publications.
What’s your use case?
