Ben's Bites
← Back
.md

Analyze agency data with AI

How to use AI to create a continuous feedback loop to improve your agency's performance and collateral.

intermediate pro
Tool: ChatGPTTool: Julius AI Topic: AgencyTopic: General

2024-12-09

Welcome to the sixth and final tutorial in our How to run your agency with AI course.

If you’ve made it this far, you’ve automated a large part of your agency GTM operations. In this lesson, we’ll cap it off with processing win/loss data to create a continuous improvement feedback loop for your agency.

We’ll use Julius to create the automated data processing pipeline and ChatGPT to incorporate the insights generated there into our agency collateral.

Steps we’ll follow in this tutorial:

  • Create an automated workflow for analyzing win/loss data
  • Upload CRM data and review analysis
  • Revise your GTM strategy and pitch

Tools needed:

Step 1: Create an automated workflow for analyzing win/loss data

To get started, go to your Julius dashboard and click the “My Workflows” option in the left-side navigation.

__wf_reserved_inherit

Then, click the “New Workflow” button.

__wf_reserved_inherit

Name your workflow (e.g. “Win/Loss Analysis”) and then click the “Generate Steps With AI” button.

__wf_reserved_inherit
💡 Tip: Julius’ AI-generated workflow builder is great, which is why we’re using it. Rather than building a data process by hand, we can quickly describe what we want to do, and Julius will create the step-by-step process.

Describe the type of analysis you want to perform in the resulting text box. We’re going to ask Julius to create a workflow that analyzes win/loss data from our CRM data and subsequently provides insights on how we can win more deals.

Sample prompt:

Analyze the win/loss data from the deals my agency has pitched and provide insights back on how we can win more deals.
__wf_reserved_inherit

Julius will create a step-by-step workflow that includes “User” and “AI” steps. Review the process and make adjustments as needed. When you’re ready, click the “Run Workflow” button to test it.

__wf_reserved_inherit

Step 2: Upload CRM data and review analysis

Your workflow should start with an upload data step. We’re going to upload a CSV of won/lost client data from our CRM.

__wf_reserved_inherit

Julius will then go through many data analysis steps.

__wf_reserved_inherit

Eventually, it will output a summary of insights, including details like win/loss ratio, average deal size, pricing feedback, industry details, and more.

__wf_reserved_inherit

Step 3: Revise your GTM strategy and pitch

Now we can take these insights and feed them back into ChatGPT to revise our strategy, pitch deck, and RFP responses. To do this, click the “Copy” button from the Julius output.

__wf_reserved_inherit

Head over to ChatGPT and ask for feedback on any of your agency’s collateral based on the insights generated from Julius. In this example, we’re asking for feedback on our most recent RFP response document.

Sample prompt:

What revisions would you make to our RFP response based on the below analysis from our previous deals?

RFP Response: [insert sample RFP response]

Analysis: [insert win/loss analysis]
__wf_reserved_inherit

ChatGPT will provide a list of revisions we can make to our RFP response document to improve our win rate.

__wf_reserved_inherit

Now you have a workflow set up to continually process CRM data and improve your agency collateral.

That’s it! You’ve completed the How to run your agency with AI course! Nice work. If you’ve gone through all the tutorials in this course and implemented them into your agency, you should see huge time savings (and win rate increases!) across your business.

This tutorial was created by Garrett.

Upgrade to Pro

This tutorial contains Pro content. Upgrade to access the full tutorial and all Pro features.

Get Pro Access