A question I get a lot is - how do I know what’s good? 

Let’s say you’ve run your representational baseline, you have your analytics, but you don’t quite know how to interpret the results.

Your baseline gives you a lot of insight! Not just the percentage of people detected and the percentage of screen time given across all of the options for all of our identity attributes, but also for intersectional identities. 

To show you an example of the analytics delivered for a single identity dimension, let me present the gender expression insights for the 2023 Super Bowl ads. We saw people with a masculine gender expression detected ~1.5x more than people with a feminine gender expression, and they received ~2x the screen time. We also saw 4 people that were gender non conforming:

And an example of an intersectional identity, here’s the intersection of gender expression and age in that same set of ads. Again, in the 2023 Super Bowl ads, under age 30 we saw a ~ 50 / 50 split of feminine / masculine gender expressions. Over age 30 it shifted to a ~ 25 / 75 feminine / masculine split:

Your representational baseline will look like this, but with data from your ads. So…..  now what???

Of course, you’ll want to uncover those opportunities that are specific to your business so you can act on them and realize the ROI of inclusive marketing. But, if you’re anything like me, you want to know: are my results good? Or are they bad?

I wish we could give you a grade or a score, but the truth is, this is all about driving bottom line impact for your specific business. And your specific business has a specific strategy and target market and set of goals. 

To answer the question: is this good? Let’s look at your baseline statistics through 3 lenses:

1. Who is your product of service for?
2. What is your strategy for this year, what are your objectives for this campaign?
3. What is the right comparison point?

1. Who is your product or service for?

If you sell period underwear or tampons, we’re probably going to see a lot of women in your ads, and that’s okay! We wouldn’t say this is “bad” or say you need to be more gender balanced in your ads because having men in there wouldn’t be additive.

However, even though it may make sense to not include people with a masculine gender expression, that doesn’t preclude you from being inclusive along the other dimensions. Including people from a diverse set of skin tones, body sizes, sexual orientations, ages, and visible disabilities is good business practice.

On the other hand, if you sell toilet paper, well that’s something that all people use. Here you do have the opportunity to be inclusive along all identity dimensions.

Because some products are specific to certain groups, we can’t apply a single bar, a single metric to say what “good” representation is across all companies.

2. What is your strategy for this year, what are your objectives for this campaign

The next thing to consider is what your specific strategy and goals are for your business and for a specific campaign. You may have built your representation baseline and realized, you know what? We actually are doing a great job of inclusion with skin tone, but we haven’t represented many people with a gay or lesbian sexual orientation. 

You may have then set a goal to include more people, more stories, and more ads with people shown in a gay or lesbian relationship. As you go forward and create more content, and as you measure those new ads, your analytics might look like they are OVERrepresenting people in gay or lesbian relationships, but in reality you are just executing on the strategy you set.

Because every company’s strategy is unique, and because it even changes by quarter or year, there isn’t a way to provide a general measure of what’s “good” without taking into consideration your specific strategy and goals.

3. What is the right comparison point

This is where we may be able to provide some guidance.

Some brands are truly meant for the entire population of the United States (or another country! Or even the globe!). With these products - after considering your product’s users and your business strategy - one place to look is the census data for your target country. This can give a few comparison points: gender expression and age have the highest fidelity references. But most of our other identity dimensions don’t have clean (or even messy) mappings to census data.

We think a good comparison point would be the category benchmark for the product you’re evaluating. For example, if your product is breakfast cereal, you may want to know how your ads compare to other breakfast cereal ads (your competitors).

So, two last questions in this blog post full of questions: