👋 Hey, it’s Sundar! Welcome to a free article of experiMENTAL: a weekly newsletter on B2C Marketing & data science how-to guides, frameworks, and stories

How Marketers can use AI

Last week, I wrote about how Data Scientist should use AI and had a few marketers reach out about my thoughts on how Marketers could use it, so here we are.

To explain this, I’m going to revisit a concept from finance called Alpha and Beta that I used last week.

Example: You buy Apple shares and it’s up 25% this year.

Beta → The beta of a portfolio is how much of your returns are related to the broader market. Example : The S&P 500 is up 10% this year. That would generally be considered “the market” and so that’s the Beta

Alpha → The alpha of a portfolio is how much of your returns are related to your decisions. You chose to only buy Apple instead of the broader market and so you’ve created a 15% Alpha.

10% (Beta) + 15% (Alpha) = 25%.

So, how does this work with Marketing? The Beta will essentially be the expected and the average performance. The Alpha is what you can layer on top to create better performance than your average peer. Using this concept, we’re going to explore how Marketers should use AI and how you can find Alpha and Beta.

But first, let’s examine the job of a marketer.

The job of a Marketer

Regardless of the type of marketer you are, there are some overlaps on what your job is. Lucky for us it’s quite simple:

1. Understand your customers

2. Build creatives that cut through

3. Optimize budgets to maximize ROI

No big deal. Easy enough.

Understand your customers

Understanding your customers is one of the toughest jobs for marketers because:

1. It’s complex

2. It requires an analytical skillset

3. There’s either too much data or not enough

So, let’s now revisit the Alpha + Beta concept.

If you’re below the Beta (aka below the average) about understanding your customers then you can use AI to help analyze the data set that you have.

For example, run all of your reviews (internal + external) through GenAI to create a wordmap or a synthesis of the biggest pain points

If you don’t have a data science team then another great use of AI is to run your data through AI. For example, a few questions you could ask it to analyze:

1. What’s the average time from signup to 1st conversion

2. Are there any attributes of people likely to convert (mobile vs web, iOS vs Android etc.)

Now, here’s where it gets good. If you don’t know what to analyze then ask AI. Again, AI is built to produce the best average using all the data it has so if you’re below the average then use it to catch up.

“Hey, what type of questions should I be asking about my customers?”

“What attributes would you look at to understand your customers?”

Then you can use these answers to analyze data. A pretty great feedback loop.

Now, the limitation is that AI doesn’t have all the context about your customer. It doesn’t have all the context about your business so there’s still a bit of a limitation here.

If you want to achieve Alpha (aka be better than the average) then you still need to talk to customers. Go do customer interviews. Learn how to probe and probe to better understand their pain points.

Within the Beta + Alpha framework:

Beta (average) → Use Gen AI to help you analyze data to understand your customers.

Alpha (beat the average) → Use Gen AI to do the above but you still need to talk to customers to get the alpha advantage. You can also use AI here to spot check your understanding of customers and see if the data aligns with what your interviews say or vice versa. Use it to supercharge the customer obsession you have.

Summary: Good at getting to Beta but not for finding Alpha

Build creatives that cut through

Creatives represent almost 55 - 70% of a campaigns predictive performance so it’s a make or break for campaigns. Now if you have a good sense of what the customer wants and their pain points then creating break through creatives becomes easier (but not guaranteed).

What AI helps with is 3 major things:

1. Increase speed and lower cost of creation

2. Depth / Breadth of creation

3. Scoring / predicting

No other technology in humankind has given Marketers the ability to speed up and lower cost of creative creation. Whether it’s images or real looking humans or whatever your creative needs, you can now build this by yourself in a matter of seconds. That is insane. Even videos (through Sora, Synthesia, Veo etc.) are now significantly more accessible.

AI can also help with the depth / breadth of creation. Say you’re a creative director and you want 20 ideas for the creative direction you want to take. Great, you can make that in 2 seconds. While all of the work for a creative director will come after in terms of pairing it down and choosing the best direction, it still helps with the brainstorming phase.

Now, if you’re a large agency or company and you have access to a team to brainstorm then maybe you don’t need this. But imagine you’re a smaller team or even a 1 person team … you can now brainstorm!

Finally, a last part of the new AI wave is the ability to see how your creatives would perform and potentially predict how it’ll be received. This is a super dangerous slope as a common case is when creative has done well in pretesting but then not in the real world. Things change on a dime so it’s not guaranteed. Still, it’s a cool new tool.

Within the Beta + Alpha framework:

Beta (average) → Use Gen AI to help you build more creatives faster while brainstorming to have more ideas. So much of it is "do more” with less resources.

Alpha (beat the average) → You still have to deeply understand your customers and then understand cultural trends. AI can help you come to market faster but I don’t see it creating new alpha for marketers other than you have less risk to your budget and can therefore experiment faster.

Summary: Good at getting to Beta and can help you find Alpha faster.

Optimize budgets to maximize ROI

So this is the most interesting section for me. I’m really torn on AI’s ability to help marketers here. The reality is that to be able to optimize budgets you’ve generally needed data science + a culture of experimentation.

Why a culture of experimentation?

Let’s say you want to run an incrementality test but your founder says “No way, we can’t afford to risk that”. It’s hard to optimize budgets without knowing how well it performs.

However, on the data science side you can use AI to help you analyze patterns. It’s not a replacement for a data scientist but it’s a step above not having one. A good example is a finance team who passed data to AI and then asked it to estimate what % of direct came from paid (classic attribution issue). It’s not the best solution but it’s a good solution.

Within the Beta + Alpha framework:

Beta (average) → Use Gen AI to help you analyze data and make decisions but be very cautious with blindly trusting it.

Alpha (beat the average) → Again, it can help you beat the average if you’re highly technical and need it to replace eng / data science resources but I’m not convinced it can totally replace the entire practice.

Summary: Okay at getting to Beta and okay at getting you to Alpha. Sorry, this is still going to be the hardest part.

Wrapping up

Gen AI is an incredible tool but I don’t think it’ll ever replace a true world class Marketer. It definitely can help smaller / leaner teams act world class but you have to have the foundations. However, Marketers can absolutely take advantage of Gen AI and find “Alpha” in their work helping their companies / campaigns stand out from others, but as always it’s a tool and the tool depends on the wielder.

Missed my last article?

Here it is: How Data Scientist should use AI

Enjoyed reading this?

Share your thoughts about the article on your LinkedIn . It helps the newsletter tremendously and is much appreciated!