News

Intimidated by artificial intelligence? That’s understandable. But it’s not an excuse. With AI introducing new capabilities literally every week, businesses must develop an understanding of how it works and develop policies to protect themselves.

We get it. We’re in the intellectual property business, so we must be extremely careful about how and when to use AI. Our agency policy prohibits using it for final-facing content or uploading client data since AI networks are not proprietary and locked down. You may want to establish your own corporate AI policy – if so, we can help.

But first, this issue provides step-by-step instructions to experience this game-changing tool for yourself. From enhancing team efficiency to uncovering insights within datasets and inspiring your own creativity, we’ll make your AI experience both easy and fun — we promise.

Enhancing Team Efficiency With AI

Zoom, Teams and the like are facts of work life. Truth be told, we don’t always pay strict attention to everything everyone says from their little screen squares. That’s where AI can really help.

But first, a warning: Any content an AI platform transcribes becomes part of its dataset. This is why Bader Rutter disables this feature for all of our agency Zoom and Teams meetings. Still, it’s important to know what’s possible.

Take these five steps to use AI to transcribe a virtual meeting, then organize the contents into a bullet-pointed summary. Sound good?

Let’s get started.

1. Pick an AI-transcribing Platform.

While Teams and Zoom both feature built-in AI functionality, you can also use a number of third-party platforms as well. For this example, we will use Otter.ai. Otter.ai has both a paid and free version; download it at www.otter.ai.

2. Set Up an Account.

Before you can use any platform, you must set up an account and fine-tune your settings. You will also want to connect Otter.ai to your Teams, Zoom or Google Meet accounts for easy use. Otter has a “pilot” function that will monitor your meetings to listen and transcribe notes.

3. Review Your File.

When your meeting is over, the transcription will be saved in your dashboard on the Otter.ai website. Here you can view and correct the transcription, highlight key tasks, view insights and takeaways, add comments or set action items. You can also use the Otter chat feature to ask questions about the meeting. For example, you can chat @otter “What are the action items from this meeting?” The bot will scour the transcription and create a bulleted list of action items from the call.

4. Download and Share the Document.

Once you’ve reviewed the document and feel it is accurate, share the notes out as a post-call report to other meeting attendees.

5. Marvel at the Beauty and Efficiency.

Wasn’t that easy? And so fast! Platforms like Otter.ai, and its built-in equivalents in Teams or Zoom, allow you to focus on the meeting first and let the technology take the notes. It’s fast, efficient and helpful, though with this, and any AI-generated work, it’s your responsibility to review and edit the content for accuracy and to ensure the content of your meeting is safe to make public.

Uncovering Insights With AI

AI-driven insights are revolutionizing marketing campaigns by unearthing compelling narrative possibilities from vast quantities of human and behavioral data. For most businesses, that data resides in Microsoft Excel documents.

Here is a quick guide to using AI to draw insights from Excel data, but remember: Any data you upload to an AI platform may be used to train that platform. For this reason, any proprietary information you share should be anonymized to protect it from potential public exposure. Bader Rutter prohibits uploading Excel data to AI platforms without client or agency permission, even if the data have been anonymized.

Got that? Good. Now gather an Excel spreadsheet and give it a try.

1. Define the Objective.

For example, you may want to uncover patterns in customer buying habits over the past year. You will get better results by defining your objective more precisely. 

2. Prepare and Upload Data.

Ensure your Excel file is formatted clearly with relevant columns. Each row might represent a transaction or purchase date. Before uploading, ensure all data respects privacy regulations and guidelines; scrub or anonymize personally identifiable information (PII). Always handle sensitive data with care, using secure methods for uploading and analyzing.

Example: Your initial Excel sheet contains customer emails and the times of their purchases. Given your focus is on buying habits and trends, “Customer Name” and “Email Address” are not necessary. Remove or anonymize these columns to protect customer privacy before proceeding.

3. Choose the Right AI Tool.

To select an AI tool suitable for analyzing buying habits, consider platforms tailored for that specific function or ones that are adept at handling diverse datasets. ChatGPT-4, with its advanced language and analytical capabilities, can be useful for gaining insights, particularly when the data requires nuanced interpretation or contextual understanding.

Example: If you’re unsure about some anomalies in your buying habits data, use ChatGPT-4 to review these patterns and gain a clearer understanding of potential reasons or trends. For visualization or deeper analysis, you might want to integrate a tool like Tableau or Power BI.

4. Run Analysis Using AI.

Depending on your selected tool, set parameters or choose an analysis type tailored to buying patterns.

Example: You might choose a “Customer Segmentation” or “Purchase Trend Analysis” feature. The AI might reveal that a sizable portion of customers tend to buy certain products in the first week of every month or that there is a surge in online purchases during holiday seasons.

5. Interpret and Act on Insights.

Once you’ve extracted meaningful patterns from the results, cross-check them with your business realities to confirm them, then decide on any next steps to take.

Example: After discovering that a large customer segment consistently buys organic products online, you might launch a targeted online marketing campaign for organic products or introduce new organic product lines.

Sparking Creativity with AI

When you are against a deadline, AI can write entire essays for you. Unfortunately, those essays can be riddled with factual errors or plagiarized content. A simple rule of thumb is to never use AI to generate final content; instead, use it to jump-start your own thinking by suggesting topics and angles for stories.

Here are five steps for generating content ideas for a company memo.

1. Choose a Generative Al Platform.

Pick a platform to use and set up an account. Two of the more popular examples are ChatGPT and Bard.

2. Feed Relevant Context and Data.

In the prompt box, provide the model with a topic and some relevant information that you would like to explore. For example, “Discuss the pros and cons of remote work on company culture. The audience for this post is young professionals who are keenly interested but anxious about potential policy changes. While the audience appreciates a professional tone, provide empathetic language in three distinct approaches. Limit each to 150 words.”

3. Generate the Content.

At this point, the model will go to work. Within seconds, you will have suggested subtopics and a range of pros and cons to consider as you develop your own memo.

4. Iterate to Refine.

If you aren’t entirely satisfied with the initial output, add more pointed guidelines to your original prompt. Be specific with your requests to ensure the output addresses any particular topics or issues you want. Then have the platform regenerate a new response that incorporates your feedback.

5. Make It Yours.

Again, AI-generated copy should never be used as final content, for three reasons.

First, it frequently returns errors, charmingly called “hallucinations.” Second, it can plagiarize copyrighted content, which can make the output illegal to publish. And finally, you have standards, doggone it. You don’t cheat on tests, and you don’t pass off the tool’s output as your own thinking. Instead, you take its suggestions and put them through the ultimate platform: your experience and perspective.

And Now, Some Caveats

AI can be applied — and misapplied

AI is only as strong or relevant as the information sets it utilizes. And, importantly, the world of generative AI is still under development. Despite mimicking human intelligence, generative AI is not sentient. It can’t recognize mistakes any child could.

That’s critical for understanding how, when and where to apply AI in our workdays. It’s smart to consider it like an incredibly capable teammate that provides an extension, not a replacement, of your team.

As an aggregator of datasets, AI presents serious issues around privacy and protection. There are both free and paid models available to use and both have their benefits and challenges.

Before you begin any experimentation, you should take steps to protect your company and your intellectual property. Consider forming a usage policy for your organization; if you’re not sure how, Bader Rutter can help.

Regardless of the solution you choose, consider the answers to the following questions.

Here’s a simple chart that toplines the scope of free versus paid generative AI platforms.

Keep in mind: At this point, AI produces many lackluster outputs. That’s the finding of a Harvard Business School study, succinctly encapsulated in this insightful data scientist’s blog here.

We hope this primer inspires you to explore the possibilities of this new technological tool. New possibilities pop up every day, which required us to rewrite this report multiple times over the few weeks we took to prepare it. But it’s worth taking the time to be AI-literate. And AI-savvy.

That will require you to learn, unlearn and relearn. But that sure beats letting somebody using AI to take your job. Or outperform your brand.