Automating Cold Emails with PandaFlow and AI
Sending Cold Emails can be time intensive task and your response rate will usually be very low, like 1–5%. But if you automate this task and use personalized AI generated content, it can become a valuable tool for getting new leads.
To be able to automate the Cold Emails — each tool, used in the process, requires an API. Here are basic steps for Lead Generation using cold emails.
- Get leads filtered to match the best audience
- Enrich lead information
- Generate personalized subject and email content
- Send the emails
Creating multi-step Automation with pandaflow.io
For automating the whole Lead Generation flow I use pandaflow.io.
Panda Flow is a popular tool for creating integration flows and automating tedious tasks. It has intuitive user interface that allows both developers and business people to create automations.
Get the Leads using Appollo.io
For finding leads I will use LinkedIn.
LinkedIn has an API called Profile API. It is extremely limited and you have to get approved by LinkedIn to use it.
Instead I use https://www.apollo.io/
Apollo has flexible search for leads, that allows dozens of filters, such as:
- Company size
- Company location
- Employee title
- Funding
- Etc…
They have a nice API for getting companies and people, including people’s emails:
https://apolloio.github.io/apollo-api-docs/?shell#introduction
One issue with Apollo API is that they don’t return full Company/Person profile info, so for that we will use another tool. See next section.
Enrich Lead info with PhantomBuster
To get full information about company and lead that I could use to generate personalized email, I utilize PhantomBuster.
Given LinkedIn url, these guys will parse out all the Company and Profile info.
Their API is not very straightforward: https://hub.phantombuster.com/reference
It does allow for creating agents that can be triggered with API, which is what I do.
Generate personalized email content with ChatGPT
Creating personalized email for each lead can take a lot of effort if done manually.
Enter AI. With ChatGPT — creating human-like emails is a lot easier.
It is still possible to detect that email was written by a machine, but it is still much much better than using same email template for all leads.
For generating personalized email I provide prompts to ChatGPT indicating exactly how to write such email. I also include both person and company info that I obtain from LinkedIn.
ChatGPT can be access through OpenAI API:
https://platform.openai.com/docs/guides/text-generation/chat-completions-api
Sending cold emails through Gmail using SmartLead.ai
I would like to use my gmail account for sending emails, at least to start.
When I tried to do that directly through gmail api, it took only a couple of days to get a stern warning from Google to stop sending spam.
Instead there is a nice service called smartlead.ai. It handles all the issues related to sending Marketing emails through gmail, by using deferred sending and other magic.
Smartlead.ai api is located here: https://api.smartlead.ai/reference/welcome
Conclusion
In this article I listed steps and tools that can be used for automating of sending cold emails to generate leads.
Next we will cover each step in detail.
