How is AI being used in the ETA process?

searcher profile

March 04, 2024

by a searcher from Duke University - The Fuqua School of Business in Raleigh, NC, USA

Hello everyone,

I'm always looking to borrow best practices and insights. If you don't mind, please let me know if and how you are using AI to support your ETA journey. Thanks!

JP

2
10
155
Replies
10
commentor profile
Reply by a searcher
from Loyola Marymount University in New York, NY, USA
There are platforms out there that provide services like deal sourcing (Smobi) that have AI working on the backend. But outside of that and giving ChatGPT some context and asking it questions or having it summarize content, the tasks and data sources for ETA are more nuanced and difficult for AI to handle. Some of obvious use cases:
1. Automated lead generation: LLMs (large language models) can be trained to scan websites, business directories, news articles, industry reports, etc. to identify companies that meet specific criteria such as revenue size, geographic location, and industry sector.

2. Enhanced market mapping: RAGs (retrieval-augmented generation) can assist in creating detailed market maps by retrieving and synthesizing information from a wide array of sources. This can help searchers understand the competitive landscape, identify niche markets with fragmented competition, and pinpoint companies that are potential hidden gems for acquisition.

3. Automating deal flow: AI models can be integrated with Customer Relationship Management (CRM) systems to automate the logging and tracking of interactions with potential targets. LLMs can summarize the key points from interactions, suggest follow-up actions, and even predict the likelihood of a deal progressing, helping searchers manage their deal pipeline more effectively.

4. Legal document/contract review: AI-powered tools can streamline the review of legal documents, contracts, and agreements, identifying key clauses, obligations, and potential liabilities. This includes flagging non-compete clauses, indemnity clauses, and other critical legal factors that could impact the acquisition.


I've found that off-the-shelf AI products are really, really bad at extracting and structuring data from CIMs, memos, etc. And due to the constraints of existing models (read: the context window), I've had to build my own automated solution to extract and structure data from websites, CIMs, etc. to be input into our systems for pipeline management. Honestly, it works phenomenally. With the click of a button I can provide the AI a brokerage website (with listings) and have dozens of new qualified leads in my database in a matter of minutes. This was a custom build though, and I was only able to achieve this by having some technical knowledge.
commentor profile
Reply by a searcher
from Emory University in Atlanta, GA, USA
Along the lines of what Matthew Sarni mentioned in #1, I have found GenAI to be invaluable in day to day tasks such as scraping sites, organizing data (spreadsheets, etc.) and cleaning up datasets. Although I am a programmer by (long ago) background, ChatGPT and other platforms have made it incredibly easy to write scripts and programs that can make a lot of the smaller, repetitive daily tasks much easier and simpler to manage.
commentor profile
+8 more replies.
Join the discussion