Prepare for calls in 5 minutes

Save 22 hours a week of research

Hi There,

In this issue of the automated salesman newsletter, you'll learn how to cut the time it takes you to find companies to prospect into by 97%. And how to cut the time it takes you to prepare for meetings by 75%.

As always feel free to click any of the links to jump to the part you care about:

Anyone in sales knows research is a huge part of the job.

From the moment you begin your role, you’re expected to start researching.

Who is our ICP? Where do we thrive?

Who are our top competitors? How do we stack up?

Who should I be reaching out to?

And, while there are hundreds of books and resources dedicated to other areas of selling like discovery or pitching.

There are very few resources that teach you how to do proper research.

It’s a part of the job you’re expected to just know.

While there a dozens of research tasks a salesperson does.

There are 2 that are very present in your day-to-day.

Especially if you sell B2B.

  1. Finding companies to reach out to:

Prospecting is the lifeblood of your pipeline.

The more qualified companies you meet with the more money you close.

And if you add the recent economic trends,

Salespeople are doing the most amount of outreach they’ve had to do in years.

Finding good companies that fit your ICP is the key to efficiently filling up your pipeline.

That way you’re not waiting for a miracle come the end of the quarter.

So you research. You prospect. And you land a meeting.

  1. Pre-meeting research

Now it’s time to prepare.

You believe that luck is when opportunity meets preparation.

And you want to get lucky with this sale.

So you research again.

You read 10k’s, and earnings calls.

You browse through websites and articles.

You look for job postings and high-priority initiatives.

You learn about their industry, competitors, and team.

Before you know it it’s been 2 hours.

But you feel ready and confident.

After all, you understand winning the sale is not about how much the prospect knows about you, is how much you know about the prospect.

Research and successful selling go hand-in-hand.

But time spent researching can quickly add up.

It starts limiting how many calls you can take in a week.

And what’s the point of doing all this prospecting if you can’t even take calls?

or even worse it can start spreading like a bad weed.

quickly taking over the time you should be spending with family, friends, or yourself.

Luckily, a new feature from various AI providers can help us cut down our research time from hours,

to a couple of minutes.

What is Deep Research?

Deep Research is the latest feature from OpenAI, Gemini, and Perplexity. It allows simple humans like you and me to easily automate any research process. (did the name give it away?)

It allows you to ask for information in the form of a "prompt", like you normally would ChatGPT. But instead of sharing a simple answer, deep research will visit dozens of sites all over the internet to find the most up-to-date and accurate information.

It's like having your own personal assistant who will search for, enter, and read dozens of websites to give you the best information out there.

So how can we tackle the 2 tasks we talked about earlier?

For these examples, we will use the OpenAI Deep Research feature.

It is by far the best out of the three.

Perplexity, however, came in second and it’s free.

Gemini, in my opinion, showed some promise, but it graded the lowest out of all the available options.

Later on, we will discuss the trade-offs of using each one.

Save 8-12 hours of finding companies to prospect into:

While a lot of material and tools out there focus on finding the right person within a company to reach out to.

You first have to find the right company.

However, finding the right companies can be incredibly time-consuming.

Even more so if you have a very particular ICP.

You can spend hours researching online and reading articles just to end up with 5 companies that match your criteria.

Let's see if we can speed this process up with AI.

Here's a prompt you could try with deep research:

PROMPT
You are my lead search assistant. Your job is to find companies who fit our Ideal Customer Profile and share them with me so I can do outreach to them. My company, {Name of your company}, sells {your service} to {your target customer}. Some other signals we track to see if they would be a good fit are: {Signal #2} {Signal #3} {Signal #4} Based on what I shared, can you find me 25 companies that fit the criteria. Output a table with the name of the company | why I should reach out

Here’s an example of a company selling sales staffing services:

PROMPT
You are my lead search assistant. Your job is to find companies who fit our Ideal Customer Profile and share them with me so I can do outreach to them. My company,sells sales staffing services to high-growth series B and Series C startups in the B2B space that have raised capital in the last 6 months. Some other signals we track to see if they would be a good fit are: Hiring for sales roles Change in sales leadership Announcements of expansion Based on what I shared, can you find me 25 companies that fit the criteria. Output a table with the name of the company | why I should reach out

OpenAI will usually ask some clarifying questions to narrow down what the ideal response would be.

And after you answer them, it will get to work.

ChatGPT took about 8 minutes to respond.

But during that time it scoured 50 websites to come up with a list of 25 companies that matched our criteria.

Something that if we're conservative and say we could’ve done it in 3 minutes per website would've taken us 2.5 hours (3 x 50 = 150 mins or 2.5 hours).

However, this could easily take someone 10-15 minutes per website for a total of

8 to 12 HOURS!!!

Tip: Experiment with different criteria or with one at a time for different results.

The list of companies it came up with is too long to include in this email. But if you want to see it click here.

You’re welcome sales staffing salespeople. ;)

Shorten your call prep time by 75%:

By now you've identified 25 companies you could reach out to and offer your services.

Not only that but the AI has given you the compelling reason you can use to craft and deliver the message.

So you reach out and BOOM.

Meeting Secured.

But oh no...

they selected tomorrow at 8 AM on your calendar.

And you have way too much work to do.

You don't have time to read every report, 10k, earnings call, and article out there.

No worries let's craft a prompt that will give you all the relevant information available,

so you can go into the meeting prepared and confident.

Here's a prompt you could try:

PROMPT
You are my company research assistant. Your job is to search for relevant information about the companies I tell you. My company sells {what you offer} to {ICP}. I have an upcoming meeting with the {job title} of {company}, a {brief description of the company}. The goal is for them to hire us to build their sales team. Your job is to give me a report on the company with relevant information I can use to sell my services. The report should include: What the company does? Who does it sell to? How big is their sales team? What are their goals? What are their financials now? What would they like them to be? What salaries do they pay their salespeople? Status of the industry Comparison to other competitors Any other information you think I should know going into this meeting. This is very important, the future of our company depends on you getting this right.

Here's my example taking one of the 25 companies from the previous exercise:

PROMPT
You are my company research assistant. Your job is to search for relevant information about the companies I tell you. My company sells sales staffing services to high-growth Series B and Series C startups. I have an upcoming meeting with the CEO of Highnote, a fintech platform. The goal is for them to hire us to build their sales team. Your job is to give me a report on the company with relevant information I can use to sell my services. The report should include What the company does? Who does it sell to? How big is their sales team? What are their goals? What are their financials now? What would they like them to be? What salaries do they pay their salespeople? Status of the industry Comparison to other competitors Any other information you think I should know going into this meeting. This is very important, the future of our company depends on you getting this right.

This time ChatGPT took 5 min and browsed through 29 sources of information. Including:

Highnote's web page.

Job Posting sites.

News sites.

and more.

It came back with an 11-page report on Highnote detailing everything I had asked for.

You can read it here.

Let's assume you'd want to be as thorough as ChatGPT and read through all the websites to extract the valuable information. It would probably take you around 5-10 minutes per site or an average of 7.5 minutes.

So reading all the 29 sources would take you around 3 and a half hours.

That's without even counting the time it would take you to search and find these sources of information.

If you take 3 mins to find each website, add another 1 hour.

For a total of around 4 hours and 30 mins.

Per.

Meeting.

Assuming you're averaging just 5 new meetings per week.

this adds up to…

22 hours and 30 minutes a week.

or

49 days per year just on research.

Tip: Experiment with following up on the information. Ask questions about the report, tell it to summarize it for you, or ask it to build a pitch based on the information presented.

Ok, I see the value, how do I try it out for myself?

All you need to do to access deep research is head over to your chatGPT account.

Sign up for ChatGPT PRO.

Find the deep research on the chat and click the button.

then copy one of the prompts I've shared.

replace the brackets with your information.

and hit enter.

That's it!

But wait, how much does it cost?

And... here comes the bad news

To use OpenAI deep research you need access to their pro version.

Which costs $200/month...

OpenAI says that this feature will eventually be available in their $20/month tier.

But I do understand is a costly option at the moment.

So before you click out of the email or start typing an angry response.

Here are 2 alternatives:

  1. Perplexity

Perplexity is another option you could use. They have a free version of their deep research tool.

To use it head to perplexity.com, and from the dropdown on the left corner of the search bar select deep research.

Then input the prompts like you would a Google search.

So why would I ever use OpenAI if the Perplexity feature is free?

Well, the performance of OpenAI is just better.

But if you’re looking for affordability over performance it’s hard to beat perplexity.

However, their solution tends to do something similar to hallucination.

It tries to use past data on companies to predict things that might not be true.

So as long as you don’t mind spending a couple of minutes fact-checking the AI then this might be the best solution for you.

You can see the 25 companies Perplexity came up with here.

And the report on Highnote.

If you decide to use Perplexity, here are my recommendations:

  • Be as specific as possible - I can’t stress this enough.

  • Fact check responses.

  • Consider becoming a paid subscriber ($20 per month) - this will allow you to select more powerful models like GPT4 or Sonet 3.5 to perform the work lowering the mistake rate.

  1. Google Gemini deep research:

Google also has its own version of deep research and it only costs $20 per month.

To use it simply head to Gemini and sign up for Gemini one.

Then once you're on the chat click Gemini Advance on the top left corner.

and from the drop-down menu select "1.5 pro with Deep Research".

While the performance is not as good, it is still better and faster than the average person.

Plus the output is cleaner than OpenAI and it sometimes finds information that deep research can miss.

You can see the 25 companies it fetched here.

And you can see the report on Highnote here.

Gemini also looks up a lot more information and resources than OpenAI so you want to be specific on what you want.

Warnings

While impressive there are a couple of things you should know about deep research that could harm your process.

  1. It could still make up information. After all, LLMs are still prone to making up facts.

  2. It could miss critical information or give you outdated facts.

To mitigate these:

  1. Be as precise as possible with your asks.

  2. Try to ask for the least amount of information that is still viable to you.

And always remember AI is still AI. You should double-check the information you're planning on using by following the links it has cited.

Just like you would with a junior member of your team.

Now you have all the information you need to go out and save a ton of time + become better at the research portion of sales.

If you enjoyed this issue of this newsletter please share it with someone who would get value out of it or share it on social media.

Weekly Challenge: Email us back with how will you be using this new AI feature or what will you do with all your newfound free time.

Logging Off,

The Automated Salesman