We Don't Sell Robo BDRs

Within days of the initial launch of ChatGPT, people were calling for the end of BDRs, since this is a task that "clearly would be replaced" by AI.

And every company in Sales tech is pitching some version of the AI-powered SDR, Robo BDR, or automated outbound (Koala included!).

We're about ~18 months after launch and I think it's become pretty clear we're much, much further from this vision as an industry than we thought.

What is AI helping with today in sales dev?

The sales tech industry has made a ton of real progress with AI. There's tons of cool stuff you can do with LLMs. To name a few use-cases we've heard or found success with:

• Research: Summarize a prospect's 10-K
• Research: Scrape a webpage and summarize it for a better understand of what a business does
• Prospecting: Find potential prospects in an account using common language to describe your ICP
• Action: Summarize the intent status of an account
• Action: Generate drafts of outbound messaging

And I do believe that we'll continue to see progress on pieces of the workflow.

But we're too early to deploy an end-to-end BDR.

Why are we too early for end-to-end BDRs?

We'll make a few assumptions about your business here – that you have PMF and that your Sales fundamentals support demand generation, a sales development team, and a sales team. If that's true, it's likely that you're currently spending a lot of money on all three of these functions.

It's pretty clear to most people why you can't replace the sales team – sales is more about the human relationships formed than just about anything else. We're far from having AI that simulates this human connection.

It's also pretty clear to most people why you can't replace demand gen – good demand gen is a highly creative endeavor that requires deep expertise in the field and judgment about what the market is buzzing about / needs education on.

So, the discussion naturally moves to sales development. However, if you're spending quite a bit of money to get someone excited enough that they wander into your store (ie., start doing research on your website and/or sign up) to consider a $50k / yr SaaS purchase, wouldn't it make sense to spend a bit of time researching them, finding out how you're connected into the company, and finding a way to earn a sales meeting by being genuinely helpful, instead of doing spray-and-pray? This is what we consistently hear from the world-class Sales teams we work with.

In talking with these Sales teams, the major gap we've observed is how inefficient it is to perform the task above. Signals work well, and people have multiple data sources for signals – website engagement, product engagement, product usage, G2 research, champion movement, social/community engagement, etc. But the data is scattered across as many tools and reps spend 80% of their time trying to get a picture of what's going on and only 20% of their time doing what they do best – getting scrappy to find a connection and write a great, human message on the right channels.

In speaking with less experienced Sales teams (ie., startups), we see a common thread where they hope to avoid the whole BDR thing. While we certainly see some effectiveness in basic campaigns that trigger from high-intent traffic, I believe it’s a local optimization and those startups would benefit from investing more in humans to help experiment in turning demand gen into pipeline. Figuring this out is hard and I understand the temptation to want to automate it away… but this process is one of the key pillars of your GTM strategy. We aren’t at a point where a company’s GTM strategy can be autonomously figured out by a Robo BDR.

Will there be Robo BDRs in the future?

We absolutely believe that we're moving toward a world where more and more of sales development work can be automated. We're already seeing fantastic automation results from things as simple as automatically sending a LinkedIn connection request from a co-founder every time someone becomes engaged with Koala (though we like to wait until we're really seeing serious engagement before sending it). People on the other side feel delight from this gesture.

However, we think the path is going to be much more iterative. I drove a Tesla for the first time about seven years ago and I only trusted the autopilot on the easiest freeway ("level 1" automation). As the full self-driving has racked up the first billion miles of training data, the autopilot has gotten better and better and I find myself using it on 100% of highway driving and occasionally even to help navigate suburbia (personally, I’m not quite ready to drive through the heart of San Francisco yet, but I can see clearly that it will happen soon).

I believe the path toward Robo BDRs will be similar — keep finding creative ways to make your workflow more efficient with automation and AI. But let's be clear: Robo BDRs are nowhere near the creativity, resourcefulness, context, and human interaction that the best BDRs are doing today. So for now: let's get these folks some better data and continue to unlock their creativity!

Your job won't be replaced by AI, but will be replaced by someone who knows how to leverage AI.