What Does ChatGPT Know About Lead Scoring?
Companies need to make sales to survive. In order to sell in the most efficient way possible, they need to attract and engage with those potential customers that are the most likely to buy from them. But how can a company know who these people are? One way is lead scoring.
If you wanted information on lead scoring, you could type various questions or keywords into Google and hope to find a relevant article or two. But nowadays, more and more people are turning to ChatGPT for instant information. But does ChatGPT know lead scoring? Let’s find out….
Step 1: Asking ChatGPT Basic Questions
Prompt: “What is a two sentence definition of lead scoring?”
Ok ChatGPT, keep it short and sweet.
Assuming I knew nothing about lead scoring going in, I would come away with the knowledge that it’s a way for companies to prioritize their efforts by targeting those who are most likely to become customers (qualified leads). This is done by assigning points to leads.
Prompt: “What are some limitations or drawbacks of lead scoring?”
Lead scoring sounds like a good idea, and can obviously save on time and other resources. Also, it would obviously lead to higher conversion rates. But what are some possible downsides?
I was only expecting two or three, but ChatGPT was able to identify a total of 8 different drawbacks in just two regenerations (a few were repeats). This is a good illustration of how much more efficient searching for information with ChatGPT is over traditional search engines.
Some of the drawbacks include being limited to available data, bias in lead scoring models, an over-reliance on lead scoring (and by extension, a neglect of other measures), cost, and the tendency to oversimplify a lead’s potential based on just a few characteristics.
Prompt: “What are the different types of lead scoring models?”
We have a basic definition of what lead scoring is, but let's get more information on the different variations.
A few repeats, but it’s clear lead scoring is not just one thing. There are different methods and data points for determining who might be a good customer fit, and it’s up to each company to determine the proper mix of lead scoring models for their specific situation (referred to as “combination lead scoring”).
There are scoring systems based on general demographic information, how someone engages with a company’s content (behavior), historical patterns, explicit and implicit actions, and more.
As a side note, it’s always good to compare multiple responses for consistency. For example, if ChatGPT had defined “behavioral lead scoring” differently from one response to the next, then that would be a good sign that additional research would be necessary.
Step 2: Asking ChatGPT More Specific Questions
Prompt: “How is lead scoring different from lead grading?”
These terms are often used interchangeably, but they aren't the same thing. Let’s see what differentiations ChatGPT makes.
The good news: ChatGPT did a good job of staying consistent with its definitions from one response to the next and identified a difference between the two terms.
The bad news: ChatGPT talks about how lead grading is “based on specific attributes such as company size, industry, or job title”, but this is exactly how it defined demographic lead scoring in a previous response. A little confusing.
You get a B- for this answer, ChatGPT.
Prompt: “What're some indications that lead scoring is being done well?”
“What does success look like?” is one of the best questions a company can ask itself.
These answers are a little vague (I guess I was hoping for specific numbers, which may not be a realistic want considering every company is different). ChatGPT does identify, in both responses, that conversion rates are king.
- “Ultimately the success of a lead scoring model is measured by its ability to convert leads into customers”
- “The ultimate goal of lead scoring is to increase conversion rates”
Prompt: “What are the signs that lead scoring is being done poorly?”
I’m really hoping for more than the opposite of the above answers. We’ll see.
It’s interesting to note that in basically all of these responses, ChatGPT defines lead scoring, without being asked, before actually answering the question itself. I Like that.
I find it rather curious that after identifying high conversion rates as the ultimate indicator of success in the previous response, ChatGPT does not explicitly mention low conversion rates here. At all.
TL;DR on these answers: inconsistency, insufficient data, and misalignment are indicators of bad lead scoring.
Prompt: “Write out a 5 step plan for how to effectively do lead scoring”
So far, we’ve heard a lot about what lead scoring is from ChatGPT, but how do you actually do it?
This is definitely a prompt worthy of regeneration to see how the recommended steps may differ each time. ChatGPT’s steps were similar (started by defining ideal customer profile and both included assigning scores/values to leads and following-up and adjusting the system) for each response, but only step 1 talked about identifying a lead scoring threshold.
Prompt: “How might a CRM aid in lead scoring?”
A CRM was mentioned in a previous answer by ChatGPT when referencing available data, so let’s get more specific.
Ok, long story short, if you’re not using a CRM for lead scoring then you’re doing it wrong.
ChatGPT identifies centralizing customer data, automating the scoring process, segmenting leads, providing detailed reporting and analytics, prioritizing and focusing your sales efforts on those with the highest potential for conversion, and more accurate scoring as benefits.
Put those spreadsheets away, son!
Prompt: “How will continued automation affect lead scoring?”
One of ChatGPT’s last answers mentioned automation, so let’s dig a little further into it.
It’s interesting that without specifically prompting for pros and cons, ChatGPT listed a bunch of good things and only one bad related to automated lead scoring (including mentioning zero bad effects in the second response).
The good: automation in lead scoring improves efficiency, accuracy, consistency, personalization, and integration with other tools. Also, it will result in less biased scoring than if it were human-led.
The bad: it may overemphasize quantitative metrics over qualitative ones.
Step 3: Asking ChatGPT Some Fun Questions
Prompt: “Write the opening scene of a noir film called ‘Lead Scoring Legends’, which is about how lead scoring can improve conversion rates for companies”
I don't really know what I’m expecting here, but maybe we’ll get something new about lead scoring from ChatGPT.
Eh, that was kind of weak. Although, I low-key want to keep prompting for additional scenes to see how it all turns out for Detective Collins.
Prompt: “Pretend you’re the salesman on the cheesiest 1990’s infomercial ever and you’re trying to sell lead scoring”
I’m a huge infomercial fan, so I'm looking forward to this.
If you’ve ever stayed up until 3 am and watched a 30 minute advertisement for some ridiculous gadget you had no idea you really needed, this answer has all the feels (“But wait, there’s more!”).
If you're an expert in lead scoring, this prompt won't really provide any new or ground-breaking information, but if you’re clueless about the subject it can be a fun intro to it instead of the boring “what is lead scoring?” prompt.
The Final Word
Overall, ChatGPT provided good, solid answers to questions about lead scoring. If you went in with zero knowledge of the topic, you’d come out fairly well-versed in it.
So, compared to just searching Google...
Is ChatGPT quicker? Yes.
Is ChatGPT more fun? Yes.
Is ChatGPT as accurate? Depends. There’s misinformation in articles you may read from a Google search, so it comes down to always double and triple checking. Just as you might read several articles to check info on Google, you can regenerate the response several times or slightly change the prompt on ChatGPT to try to verify accuracy.