The promises made by AI startups always sound like magic. But the more I’ve used new AI tools, the more I’ve realized that their results rarely live up to the hype. Now, whenever I hear a snappy AI value proposition, my first thought is always: What’s the catch?
However, when I saw that some companies were using AI for sales email outreach automation, I had to give it a shot. No more manually creating every single different email sequence? Fully-personalized conversations based on contact activity? Sign me up. The promise of an AI-powered SDR campaign was too good to pass up.
Setting up one of these tools can take a lot of time—so my goal is to help you avoid wasting your first month’s payment doing that (like I did). Some of these steps below may seem obvious, but once the clock starts ticking, you might find yourself thinking the same thing I did: Why didn’t I plan this out beforehand?
Below, I’ll be outlining the two main components of an AI SDR tool: The persona and campaign.
Building your persona: Plan out these 5 components first for emails that sound like a human.
So, what’s a persona? Think of this as the AI “person” writing the email. It’s the backbone of the entire tool. A well-created AI persona knows your writing style, follow-up approach, value proposition, pain points, FAQs, objections and case studies. I’ll outline all of these below. You can set up as many personas as you want, each with their own style and perspectives. They’re the SDR employees sending your emails… they just aren’t real.
When building your first AI persona, you realize that these tools aren’t magic; they require a lot of input to generate accurate, personalized emails that align with your business and prospects. If you don’t put in the time, there’s great risk of your AI persona generating cringeworthy, definitely-AI-generated-sounding emails that scream “autopilot”.
1- Case studies: Give your persona real examples of your work.
A well-written case study is a great training tool for your AI SDR. The same things that make it work for your prospects are exactly what you want the AI to know— your value props, specifics about how you work, and real results that come with stats. Check out one of ours for reference.
2- Value & offerings: What your persona is going to offer.
You have to be very clear on what your offerings are—and how you can deliver them—when teaching your persona. When feeding this information to the tool, structure each offering as a title, with supporting bullets explaining how you execute.
For example, one of our key offerings is integrated demand generation programs. A good way to frame each offering’s details is to imagine they start with, “We help you…”
- [We help you] Develop and implement scalable, multi-channel demand generation strategies to enhance engagement, improve lead quality, and generate measurable business results.
- [We help you] Eliminate the guesswork from campaign planning with a strategic, data-driven approach that aligns marketing and sales efforts, ensuring you target, reach, and engage the right audience.
Before we get to the next thing you need to think about, it’s worth saying that you probably don’t need to start each of these from scratch! Your AI SDR is serving the role of an SDR—which means your actual SDRs likely have a lot of this information already. Lean on them to help pull this together, and give them a little reassurance that this isn’t going to negatively impact their comp plan. AI SDRs aren’t likely to close on their own. (Yet?)
3- Pain points: How your persona is going to connect with your prospects.
Once you’ve outlined your company’s offerings, define some pain points associated with each offering. When writing a pain point, think of it as an opener to your email, as that’s how the personas typically use them.
For example, here’s a common pain point for integrated demand generation programs:
“Marketing campaigns generate large volumes of leads, but many are unqualified, leading to wasted time and resources for sales teams.”
Your persona would then continue the email with something from your values section, like:
“We help companies implement scalable, multi-channel demand generation strategies to improve lead quality, leading to measurable business results.”
This is a simple example, but it illustrates how these tools work—and why it’s so important to clarify your value propositions and pain points.
4- FAQs and objections.
The good news? The hardest part is behind you. Depending on your business’s maturity, you likely already know the most common FAQs and objections that might arise during a sales conversation. Take some time to map them out clearly.
5- Writing style.
Next, review your past email sequences to identify writing styles that have worked well for your audience. If you’re unsure where to start, or haven’t been successful in the past, plug the work that you’ve done above into ChatGPT and ask for help identifying sample email styles that work well with your Ideal Customer Profile. Be specific. Give ChatGPT your goal, the return format, warnings and all relevant context (such as the fact that you’re using an AI emailing tool). You can do this prompting for FAQs and objections as well.
Once you’ve established your writing style, you can use it across all future personas, as well as any broader FAQs and objections about your business.
Setting up your campaign.
Now that you’ve set up your persona, it’s time to create a campaign.
What’s a campaign? You can think of your campaign as your audience configuration. You have your AI persona “employee” who writes all about your business and sounds like a human, but who are you going to email? Thankfully, setting up a campaign is far easier than building personas. Here’s what you’ll need:
- A finely-tuned suppression list in your CRM—you’ll thank me when you’re not accidentally emailing current customers.
- A segmented ICP list within your CRM.
- A working calendar tool like HubSpot or Calendly.
You can also create cold lead lists within these AI SDR tools, but it’s important to remember that warm leads always outperform cold. Take the time to build a list of leads who have already interacted with your company (aka: contacts in your CRM)—whether by converting on your website or at least visiting recently. Your results will be significantly better.
The final product.
This is the part where you’d expect to see a final product from one of these tools, but each email is personalized to the contact. And that’s the best part: No more generic emails that nobody opens, and more emails that reach your prospect with the right information at the right time.
The Iron Horse insight.
No matter what AI tool you’re using, your output is only as powerful as the information you feed it, and taking the time to plan ahead will save you time and money. Additionally, tools like Common Room can help you track relevant prospect activity signals, which takes these tools a step beyond simply scraping LinkedIn activity, and can help you reach your ideal customers at the perfect time.
