- Sales AI
- Posts
- AI BDRs are a sham (mostly)
AI BDRs are still mostly vaporware - despite their proliferation. My goal with this newsletter is to get you to think critically before you trust your account list to a tool that promises you the moon.
First, what comes to mind when someone says AI BDR? An Agent that can, in a human way, find leads at your target companies. Write something relevant to them. Infer their needs. Follow up with them contextually. Handle objections. Basically, take the prospecting load from the AE.
But what you’re going to find in these AI BDRs are something that takes you further away from your goal and necessitates a human BDR. At the time of writing, I found 13 AI BDR tools (there’s surely a lot more):
Here’s what I’ve seen, not from the outside looking in, but the inside looking around. As a BDR, a developer, and one of the earliest developers to use AI to automate BDR tasks, I want to break down the hooplah:
The promise
Sales on autopilot! Automated prospecting! Infinite scaling! Personalized messages! Book meetings like crazy!
Lead lists: Every tool promises to find prospects for you with AI. Perfect account lists. The Reality is every one of these tools can only use data that’s provided to it, and many are pulling from LinkedIn or Apollo (which pulls from Linkedin). They do this worse than a human because they can’t infer certain things about an individual, only read static titles and blog posts.
This encapsulates most of these tools capabilities for features like “Discovering Decision Makers” which is usually code for “Pick the highest ranking person in that department.”
Copywriting: With maybe one or two exceptions, every tool promises to write great personalized emails for you. The Reality is personalized first lines based on a LinkedIn profile (which may or may not be relevant or true) and generic GPT copywriting with no ability to iterate or learn from past campaigns.
So what happens when everyone is using a GPT copywriter? You’re not exactly going to stand out, are you? And without a lot of legwork, GPT isn’t going to effectively iterate your campaigns based on feedback it’s receiving.
Reading and Responding to Emails:
This is one of the areas I think these tools shine - time to respond to leads is super important. If you can get in and schedule a meeting within 5 minutes of receiving their email, you have a much higher chance of actually booking it.
But, many don’t handle the objections well. And the money is in handling those objections. If someone says they’re happy with their tool, you’re going to get insight or a meeting if you can spend the time to dig into what they’ve said and highlight your difference. We’ve not seen AI perform well at that yet.
Signal Processing:
Not every tool has this and I think it’s the best use case for AI SDRs right now. Watching for your prospects to post something and then respond to it quickly is a huge advantage over human BDRs.
The Sum Total
These tools look cool and shiny, but it’s not something you should hand your whole account book to. It will perform worse on lead lists, copy, iterations, and objection handling.
Instead, you should use it with guardrails. Have it purely watch your top 100 accounts and write timely emails when someone posts something. Otherwise, you’re going to do worse on your overall outbound when you have it handle everything.
Finally, as these become more ubiquitous, human BDRs are going to stand out. All these AI BDRs have the same voice (if you’re like me, you get a lot of prospecting from their users) that people are going to become more keen to.
There’s probably an 80/20 here where it can handle 80% of your prospecting “efforts”, but that last 20% of your effort is what’s going to produce results. That’s not truly automateable yet.
Where the space is going
I want this space to succeed. We’ve built a lot of AI to take care of certain sales functions at Sales.co, but a human is always in the loop. Whether that’s high level strategy, judging personalizations, seeing if emails are tagged correctly, AI can’t be trusted to be left to its devices on so wide a use case yet.
So, narrow the use case. I’d love to see one of these tools get really good at objection handling. Or finding you new prospects based on your current customer list. Or coaching your BDR into changes they can make to increase their success rate.
AI is more successful the more narrow you make it. Unless you’re building a custom in-house solution, an all-in-one AI BDR likely won’t cut it for you today.
Love,
Ryan Doyle