AI head shots that beat out the real ones


Your AI head shot is waiting

The other day, I got a note from Jessica Jobes, a multi-time founder who is currently the CEO of digital marketing firm Vital Velociti. I'm always excited to hear from Jessica, but in this case, I was extra-enthused: Jessica recently completed Viral Data Stories 101, and she was writing to show me her first data story. And oh, is it a good one.

Jessica's story compares the clickthrough rates of AI-generated photos and real photos taken by a professional photographer. I don't think I'm spoiling anything when I tell you that it's a pretty close race. You gotta read the piece. Whether you're a professional marketer or just choosing your head shot for LinkedIn, it's powerful data.

The story is also a great example of what an exceptionally well-told data story can do for your professional brand. Let's take a look!

What the AI Photo story gets right

Right from the jump, Jessica's story aces the fundamentals:

It's clear who the story is for. Jessica's business serves digital marketers, and this story is exactly designed for this audience and loaded with insight they can use. The story also illustrates her own product offering; in other words, the piece is a living case study of the kind of work she does for her clients. She totally nails audience definition.

She has a clear hypothesis, and she builds a perfect data set to test it. A great data hypothesis takes you to a good story even if the hypothesis turns out to be false. In this case, Jessica is comparing the performance of AI head shots with real ones. With the right data, any of these conclusions would have made a compelling story:

  • AI photos outperform real ones
  • Real photos outperform AI ones
  • AI and real photos perform exactly the same

So even before she's collected any data, Jessica's story frame sets her up to share an interesting insight. No matter which way the data goes, Jessica will have a good story.

From there, she builds a scrappy data set that is the right size to test her hypothesis. It's tempting for new data storytellers to believe they need thousands of data points to tell a good story. If you choose your data points well, this is not always true. Jessica collects enough photos to run a good test, no more and no less.

The data visualizations are simple and clear. Jessica shows photos in a grid along with their clickthrough rates, like this:

The data visualizations clearly show the relative performance of AI vs. real photos. You don't need to be a data scientist to understand them.

The bottom line: The story is super well-conceived. She hits a ton of the viral data stories playbook 💯

How I'd turn the AI photos story up to 11

When she sent me her story, Jessica asked for a teardown, and she was cool if I published it here. In that spirit, this is the stuff I'd do if I wanted to take the story from 😄 to 😍😍😍:

💡 I'd include one fewer data slice. This piece is awesome. It's comprehensive. It's clear. It's also very long, and it introduces an extra question that I would drop: whether Jessica should use the winning photo on her company website as well as in her LinkedIn profile.

I'd drop this dimension because it adds words without any new data.. I'd edit to make the main insight stand out more.

💡I'd leave out the "what does this mean" section at the end. Shorter pieces are more likely to be shared. On the other hand, I'd include a clearer link for people to learn more about Jessica's company. After reading the piece, some people will want to hire her.

💡I'd consider launching the story as part of a series. This is a great standalone story as is. But a whole "AI vs. real" series that can be announced in advance and that people can subscribe to ahead of time? That could drive big audience momentum.

The bottom line: Whether you're a professional marketer or simply sprucing up your LinkedIn profile, you gotta check out Jess's story for the data alone. And if you want to tell exceptional data stories of your own, you can set up a free consult with me after you work through the playbook here.

Thanks for reading!

Kieran


Are you serious about building exceptional teams? I wrote down three of my tried-and-true prompts for team meetings and offsites that drive team connection and performance. Ready for your next team meeting, no big budget required.

My latest data stories | Tell your own Viral Data Stories | nerdprocessor.com

kieran@nerdprocessor.com
Unsubscribe · Preferences

nerd processor

Every week, I write a deep dive into some aspect of AI, startups, and teams. Enter your email address to subscribe below!

Read more from nerd processor

Still more data on in-person vs. remote This week, we're taking a final look at the 180 startups I met in the first half of 2024 to see how they're performing 6-12 months later. So far, we've covered: Episode 1: Most startups hire their VP of Sales too early, and their VP of People too late Episode 2: Team sizes around 5 look ideal for company performance; when you're >$5M ARR, it helps to hire an outside COO, but if you're only at $1M, hiring the outside COO hurts more than it helps Episode...

Following the data This week, we're taking another look at the 180 startups I met in the first half of 2024 to see how they're performing today, 6-12 months later. So far, we've seen that: Most startups hire their VP of Sales too early, and their VP of People too late Team sizes around 5 are ideal for company performance When you're >$5M ARR, it helps to hire an outside COO, but if you're only at $1M, it hurts more than it helps This week I took a look at founder demographics to figure out...

Flatter, deeper, faster, stronger Last week, we looked at the 180 startups I met in the first half of 2024 to see whether it helps or hurts your growth to hire a VP of Sales and a VP of People by the time you have $1M in annual revenue. This week, we're going to look at the same companies to figure out how many managers a startup needs if it wants to grow. How big should teams be? Do flatter structures perform better? When should you hire an outside COO? This is Part 2 in a 4-part series. You...