A pilot of AI-powered email
Robin Izsak-Tseng
Growth Leader: Navigating IPO Milestones, High-Value Acquisitions, and Building Global Teams.
I’ve had a number of requests to walk through how we ran the pilot of 6sense Conversational Email when I was at G2, so here's my attempt to capture it. As mentioned in my original post, we rolled out the pilot across three specific use cases.
NOTE: G2 is pretty unique in many ways so I’ll give our specific use case, then suggest a few best practices that can be applied basically anywhere.
OTHER NOTE: If anyone is using another AI-powered email tool (Clay?) I'd love to hear about it!
Like most things, you're more likely to succeed with focus and clarity about what you're trying to achieve, then roll it out in a limited capacity to start testing.
Use Case #1 - Someone adds a new product on G2
When someone submits a new product, the G2 research team initiates validation, categorization, and then publishes the product on G2’s marketplace.?
Would you believe that before 6sense CE, we would only send a transactional email letting the submitter know their product listing was approved? Eventually we got (marginally) smarter and had an SDR reach out with a 3-email nurture that pushed to a demo. This was manual, not efficient, and created not-great experiences for the prospect.
So when I heard about 6sense Conversational Email, this seemed like a great use case for the pilot.
Instead of the joyless transactional email and 3-touch SDR nurture, what if we triggered an AI-powered flow 24 hours after their product was live on G2?
And what if the email appeared to come from me, simply recognizing an important GTM milestone, congratulating the submitter ??, and asking what they’d like to do next??
I wanted to test the theory that if we engaged these folks at the time they had actual momentum—and we did it in an authentic way that came from a peer (not from sales)—they would respond.?
And they did.
The beauty of 6sense CE is that we found about 60-70% of the time AI could reply to these responses automatically. The tool is built to respond to the following scenarios:
They want pricing info or to talk to sales ??
AI Robin responds and automatically copies the inbound SDR on the reply.
They want to talk to sales, but they add a team member who will own the relationship ??
AI Robin responds appropriately and also copies the inbound SDR to set the meeting.
They’re not ready to talk to sales just yet ??
AI Robin responds with helpful resources to explore at their own pace.
They’re not interested at all ??
AI Robin politely responds with “Cool. We’re here if you need us," and includes links with helpful resources (case studies, videos, etc.).
For the 30-40% of the time that AI could not interpret the person's reply and respond automatically, I’d get a “human review required” notification. This gave me the opportunity to see the variety of questions people had, particularly around how their product was categorized or best practices for getting reviews, how to drive more leads, etc.
I actually enjoyed reading their questions and taking time to respond thoughtfully. I got to know our audience better and learn where people were hitting roadblocks that we needed to address. At the same time, it could be tough to keep up with all the “human review” flags given my calendar, so I partnered with an SDR to help with any responses I couldn’t get to.
If this sounds like too much work, I’d argue that interacting with your prospects and customers is the most meaningful work you could be doing.?
A few ideas and best practices that could apply anywhere
领英推荐
But how is all this different from segmenting an audience and putting them in an automated email flow, you ask?
To me, the difference is the who, the how, and the AI magic. Sending the flow from the “right” person and having the messages feel like helpful conversation starters will get responses. AI will handle the rest, eliminating the need for someone to monitor and respond manually.
?? Integrating with your routing tool on the backend will get the person to the right place. ??
Use Case #2 - Someone downloads content
We all invest in content marketing of one kind or another. When I was at G2, we would release an ebook, a report, or a guide every quarter and, like everyone else, we often used these for lead gen.
So when I learned about 6sense CE, I wondered if there could be a better way to engage people who download content (especially the meatier, more mid-funnel stuff), rather than sending the usual transactional email and (mostly ineffective) follow up nurture.?
Why not wait a day or two and trigger a flow, starting with a message like this:
Hey Tim - I’m glad you grabbed our ABM guide. Occasionally I like to reach out to see what people think and how it's going. I’m always here to swap best practices or share struggles.
Enjoy the read!
Robin
VP of Revenue Marketing, G2
Instead of MQLing these “leads,” dropping them in a nurture, or worse—not doing anything at all, why not engage them in a conversation—marketer-to-marketer? If you write the emails with that outcome in mind, you’ll find people respond.?
And, as outlined in Use Case #1, AI can usually handle the qualification and responses. When it can’t, you’ll enjoy writing back, learning more about your audience, and thinking of new ways to engage prospects and customers at scale.
Use Case #3 - Someone becomes a new admin of their G2 page (or any product for that matter)
New admins of your product should not—under any circumstances— be taken for granted or ignored.
People leave companies or responsibilities shift, and just like that your admins and users are gone (and yet automation is still blasting them with emails that bounce ??).
But new admins and users? These are people you should focus on retaining and growing.
My experience is limited to 6sense CE so I can’t speak to how other AI-powered email solutions work, but with 6sense you’ll create a workflow that will dynamically add new admins (or whomever you want) to a Salesforce campaign, which will queue them up to receive the first 6sense email at the right time.
In 6sense, you’ll assign the email sender and the SDR, BDR, AE or CSM (whomever will be copied on any replies) and connect your routing also.
Once you start thinking about your own use cases, you'll uncover great opportunities to create workflows that deliver the right message at the right time, enabling you to start conversations at scale.
Because we set up the three campaigns to track revenue both influenced or sourced from each, we could clearly see that Use Case #1 was the top performer. That campaign alone generated enough revenue for the tool to pay for itself within the first four months. All of the campaigns had over 70% open rates and solid response rates, with Use Cases #1 and #3 the highest.
There’s a lot of operational stuff I didn’t cover here but hopefully this gives you an idea of how we approached the pilot, how we succeeded, and a framework for considering a tool like this in your own organization.
Have fun! ?? ?? ??
CTO and Co-Founder | Angel Investor | Advisor
3 个月Thanks Robin for sharing with the world different kinds of plays to run and how to be thoughtful about them. We believe that in this new world order of GenAI, humans will focus on what they love and your use case #1 showcases that! I have never heard anyone say "I loved reading the email responses" ??
Chief Product Officer | SaaS | Growing and Scaling Tech Companies
3 个月Thanks for sharing Robin . This is very insightful and shows great now ways of engaging customers in a personalized way - much better thank the traditional email nurture.
Senior Customer Success Manager, Enterprise at 6sense
3 个月Kendra De Berti Lisa Nguyen Huynh Matt Grebow
So is Black Sails!
Head of Product | Building an SDR AI Agent
3 个月Love how you're testing a new way to have more personal conversations with product and marketing leads. We're bringing in more data to make the AI-generated email content even more personalized :)