This content was originally featured in the March 6th, 2024 newsletter found here: INBOX INSIGHTS, MARCH 6, 2024: IDEAL CUSTOMER PROFILES, GA4 GENERATIVE AI
USING GENERATIVE AI TO BUILD YOUR IDEAL CUSTOMER PROFILE (ICP)
My brain is buzzing and my head is spinning.
Chris and I have been using Generative AI for the past two hours to build our Ideal Customer Profile (ICP). Why? Often we get asked, “Who is your ideal customer?” I realized that we always give a weak, vague answer. So, it was time to do better.
We have a large amount of data. This should come as no shock to anyone, given who we are and what we do for customers. The challenge has always been figuring out how to make sense of this data in an actionable way. Ironic, given what we do for our customers. But, our company is always the last priority. Something like “the cobblers kids have no shoes.” Anyway, we finally set aside the time and went through the exercise.
What did we do? First, we create a user story. I was floundering around with too many places to start and needed to focus my attention. Chris asked for user stories for this exercise.
This was my first user story. Like me, it’s a bit of a mess:
“As a CEO I need a better handle on the characteristics of our ideal customer. so that I know who we’re looking for on LinkedIn and who we’re trying to reach with our content. So we can speak to that person.”
I went ahead and refined it to be more focused. This was the second draft that we worked off of:
“As the CEO, I need to better understand the data points of my ICP, so that I can create tailored content and services for them”
It’s still a complex user story, but it hits all the marks. If we break it out into the 5P Framework, this is the data we know:
Purpose: To understand our ICP. To create tailored content and services for our ICPs. Who is most likely to benefit from our services?
People: Me. This is where we should have created more user stories. I’m not the only one who will be executing the eventual plan. However, since we’re a small company, I prioritize and delegate the action items.
Process: What I stated was “better understand” and “who we’re looking for”. This tells me that we need to create the ICP in such a way that the characteristics create a search query
Platform: I listed LinkedIn. You don’t list your platforms in your user story. What you get is your instructions for what you’ll need. To “better understand” our ICPs, we used a wide variety of platforms where the data lives. To have a well-rounded ICP, we need to bring in a lot more data. So we used LinkedIn, our CRM, our website, our contracts, and other documents we’ve created. We used Google Gemini to help us analyze and summarize the information.
Performance: If our purpose is to understand who our ICP is, that’s what I’ll measure our exercise on. (spoiler: we figured it out).
If I were grading the user story, I’d give it a solid B. It could be further refined, but it gave us enough direction to get started.
Next, we had to collect all the data from the Platforms we listed. I’ll spare you the gory details of how we had to pull everything together. That’s not the point of this post. After the two hours we spent pulling everything together and massaging it through Gemini, we had our ICP. I know have a better handle on company size, revenue, roles, and pain points. Knowing these factors will help me focus our efforts.
What I learned through this process was that Generative AI was not the star of the show. Sure, we got through the data faster than we would have doing it manually. However, there was a lot of iteration with the system. The initial results were always generic. The memory of the system was short term. The capacity to take in data has limitations.
The star of the show was us. Me and Chris. The humans. The brains of the operation with the institutional knowledge. The people who know the company so well that we can reject generic results without hesitation. Not generative AI.
I say this because there is a lot of fear and insecurity that AI will take your job. There are hasty decisions to replace entire teams with generative AI. There is a large misunderstanding that generative AI will solve your problems. It won’t. And it should not replace your entire team.
Had we not worked at it, iterated, pressed on, we would have had very mundane results. We would have had an ICP, but there would be nothing that differentiates this profile from the company down the street. What are the specific pain points of OUR customers? Are the services we offer aligned with those very specific pain points?
You are the star of the show. You are the valuable asset. You know your customers, your company, your outcomes. Generative AI can learn these things, but it will forget as soon as you tell it. Generative AI can come up with a marketing plan, but it will give you the same plan it gave the person before you. You are what makes generative AI successful.
Let me climb down off that soapbox and get back to the point. Well, that was the point. Generative AI made our process move a lot faster, but it didn’t replace the hard work. We still went through the 5Ps. We still went through all our data. We still had to refine the outcomes. And now, we still have to do something with this information.
Now that I have an ICP, I can start taking action on creating custom content and services. I can focus my time. I can streamline the topics I write about in this newsletter. Hopefully, I can close more business. Fingers crossed.
Do you know your ICP? How did you create it? Reply to this email to tell me or come join the conversation in our Free Slack Group, Analytics for Marketers.
– Katie Robbert, CEO
Newsletter
Podcast
