From Prompt to Professional: How GenAI Images Are Saving 951 Years of PowerPoint Prep
RevExpanse Proudly Presents PowerPoint Pic Tuesdays!
In the early phases of a technological revolution, it can be difficult to appreciate the trajectory of change that is coming. Generative AI (GenAI) is no different. Many, rightfully, have a healthy degree of skepticism when it comes to GenAI. In the recent past, people saw the hype around blockchain crashing down. Blockchain evangelists touted the technology as a revolution in financial transparency that would eliminate the need for traditional audits. Yet this promise proved hollow when Sam Bankman-Fried's crypto-exchange, FTX, collapsed, concealing massive fraud despite its use of the "trust technology."
AI itself has been around for decades. More than once, the hype around AI ended up with disillusionment, causing research funds to dry up, resulting in an "AI winter." The first major AI winter struck in the 1970s after early promises fell short. Another followed in the late 1980s when expert systems failed to deliver expected breakthroughs.
Will this time be different?
Although we are likely in the "1995 Phase" of technology (to borrow from the Internet timeline), we can see that there are real benefits and uses of this emerging technology. Specifically, image generation. When it comes to AI images, the focus has often alternated from memes to deep fakes. The latter, to use a risk and compliance lens, is a legitimate issue and requires vigilance. That being said, we also need to see the nascent, but powerful upside to image generation.
As any seasoned accountant, auditor, or consultant will tell you, there is a desire to bring life to PowerPoint presentations, reports, or other communications. The problem? The lack of time and energy. We all know that these deliverables are prepared in the wee hours of the night or at the end of a very long workday.
The result?
Finding the first free stock image that we can find online or in Microsoft’s library.
Rethinking images in the Age of GenAI
A picture is worth a thousand words, but in the age of AI, a few words are worth a thousand pictures. And in this amplification effect, those few prompt words unlock millions of words of meaning - as each generated image still carries its own thousand words of depth, emotion, and context.
With the ability to generate a prompt, we can almost effortlessly find that perfect pic to bring out the text that’s on the slide, on the page, or in that deliverable.
Is it important to find that perfect pic? Does anyone care?
According to research, the answer is a resounding “yes”.
A 2015 study published provides compelling evidence for what many of us intuitively know - visuals significantly outperform text-only presentations. Researchers Kernbach, Eppler, and Bresciani conducted an experiment with 76 experienced managers to measure the difference between bullet-point lists and visual presentations (using metaphors and diagrams) when communicating business strategies.
The study's overall conclusion is definitive: "Visualization was significantly better than text in terms of the achieved attention, agreement, and retention." While we see potential in AI’s ability to generate images, this research reminds us that there's real science behind why visuals matter in business communication.
Other key findings from the study include:
1. Attention scores increased by approximately 25-40% when visual formats replaced bullet points - measured on a 5-point scale where participants rated how often they focused on the presentation. This confirms what we've suspected: visuals genuinely capture and maintain audience focus in ways text simply cannot.
2. The perception of the presenter was dramatically improved when using visuals - with 68.7% of how an audience perceives a presenter being directly attributable to the quality of their visuals.
3. When comparing high-performing companies against underperformers, researchers found that 73% of outstanding performers clearly communicated their strategy visually, while only 28% of underperformers did the same.
In addition, the researchers noted that "bullet point lists and written statements are not the best way to communicate complex strategic ideas," suggesting instead that diagrams or images are more effective. They also observed that visual elements help to "focus, accelerate, and improve recall" of information - exactly what we need in our content-saturated business environments.
Perhaps most tellingly, the study concluded that "by using diagrams and visual metaphors instead of just text, the perception of the presenter by his or her audience can be improved." In other words, your audience doesn't just understand your content better with visuals - they think more highly of you as the communicator.
(for the paper see: Kernbach S, Eppler MJ, Bresciani S. The Use of Visualization in the Communication of Business Strategies: An Experimental Evaluation. International journal of business communication (Thousand Oaks, Calif). 2015;52(2):164-187. https://journals.sagepub.com/doi/abs/10.1177/2329488414525444 )
AI and Images: From Midjourney to OpenAI’s 4o Image Generator
In addition to generating accounting memos about the cash balance, image generation was one of the early use cases. More importantly, it created that awe and amazement about how the future is finally here. Midjourney was clearly the pioneer in the field. To generate images users needed to join the Discord server, which was great for early community building.
AI-image generation, however, quickly attracted several competitive offerings. As the market expanded, each service required its own separate subscription, leaving users forced to choose between paying for multiple platforms or missing out on unique capabilities offered by different providers.
Through Poe, users could access open-source models like Black Forest Lab's Flux, as well as proprietary models like OpenAI's Dall-E-3, Google's Imagen 3, and Stability AI's Stable Diffusion. Poe also included access to Ideogram, which particularly excelled at photorealistic image generation. Here is a sample:
Prompt: “Make an image of the sun setting in a rocky mountainous area with a serene lake and the reddish glow of the sun reflecting on the lake.”
Flux was also found to generate comparable images. The following was generated using the same prompt:
OpenAI’s 4o Image Generation: The Only Model We Need?
OpenAI reinvigorated its image offering in March 2025. As per their announcement:
“GPT‑4o image generation excels at accurately rendering text, precisely following prompts, and leveraging 4o’s inherent knowledge base and chat context—including transforming uploaded images or using them as visual inspiration. These capabilities make it easier to create exactly the image you envision, helping you communicate more effectively through visuals and advancing image generation into a practical tool with precision and power.”
How does it fare? Using the same prompt as we did with Ideogram and Flux, you can judge for yourself:
Perhaps not as photorealistic as Ideogram’s, but it is much better than the previous model that we had access to:
Prompt: “Make a sunset that will be used in a PowerPoint deck to symbolize closing thoughts (Do not include any text in the image produced)"
Though it was good, it definitely has been surpassed by OpenAI’s latest offering.
Where will OpenAI Take Us?
One dramatic improvement is the ability to generate infographics and text. Consider the following:
Prompt: “Create an infographic explaining how financial statements get generated from transaction to journal entries to the different financial statement in great detail. The background should be pure white, and include neatly rendered text labels with step-by-step annotations. Each step should be numbered and connected with subtle gradient arrows, and include both the theoretical explanation and practical observations. Make sure the list of financial statements at the bottom is not cut off.” (Reference)
Also, it allows you to leverage art styles. Here is a rendering of my LinkedIn photo in anime style:
What does this all have to do with PowerPoint Tuesdays?
RevExpanse maintains that generative AI is a tool to amplify our efforts. Let’s look at the math when it comes to generating images. Let’s assume there are 100,000,000 images that are generated within professional contexts. This is extremely conservative as there are about 1 million consulting firms in both Canada and the US. It assumes that each firm generates 100 images a year. This is quite low given that we know decks are a key part of the consulting process, let alone what professionals are producing as part of the operational or financial planning and analysis (FP&A) teams.
How much time does GenAI save in aggregate?
That’s right, 951 years!
Enter PowerPoint Pic Tuesdays!
To tap into this productivity, RevExpanse is pleased to announce, "PowerPoint Tuesdays." Every Tuesday, RevExpanse will release a picture-perfect PowerPoint pic to enhance your deck, report, or other business deliverable. The idea is something that we promote as a part of our task amplification categories that we deliver as a part of our GenAI workshops. Specifically, it fits under “Prep and Polish”, which we define as follows:
“Deliverable Prep and Polish involves using GenAI to enhance the quality and presentation of your work. This includes advanced editing, improving writing style and tone, formatting tables, and generating images or graphics. By leveraging AI for these tasks, you can ensure your deliverables are not only well-written but also visually appealing, potentially saving time in the polishing phase of your work.”
Of course, we can’t wait till Tuesday. So, we’re going to start this one early. Here is a picture that we think illustrates “new beginnings”.
Prompt: “Generate a professional, photorealistic photograph of a golden sunrise breaking over a modern city skyline, warm light illuminating glass buildings, minimal clouds, clean corporate aesthetic, soft gradient in sky from deep blue to orange-gold, symbolic of new beginnings, high resolution, suitable for business presentation”
Follow us on LinkedIn, give us a like and enjoy the pics!