A LinkedIn post with the most comments ever (17,697) was one about AI agents. The hype is real but the question is…
How do you avoid the smoke-and-mirrors and find the right AI Agent for outbound?
💎 𝗦𝘁𝗲𝗽 1: 𝗘𝘃𝗮𝗹𝘂𝗮𝘁𝗲 𝗗𝗮𝘁𝗮 𝗔𝗰𝗰𝘂𝗿𝗮𝗰𝘆
The success of your initiative depends on the data that the AI agents use, such as: 1st, 3rd party data like signals, your CRM data, product usage data, etc.
If the data isn’t accurate, AI agents won’t be effective, or worse, they could damage your brand & reputation. As the saying goes, garbage in, garbage out.
To assess the vendors’ data accuracy:
1. Identify 1-3 specific campaigns you want to start immediately with the AI agent. The results will be more relevant and real for your team. And once you pick a vendor, you can turn them on immediately, reducing the time to impact.
2. Provide each vendor with the same list of accounts and contacts so that you can compare apples to apples.
For example:
- List of Customer accounts
- List of Target accounts
- List of Customer contacts
- ICP criteria
- Persona criteria
💎 𝗦𝘁𝗲𝗽 2: 𝗘𝘃𝗮𝗹𝘂𝗮𝘁𝗲 𝗔𝗜 𝗺𝗲𝘀𝘀𝗮𝗴𝗶𝗻𝗴
Ask each vendor to provide you with 20-30 sample emails for the campaigns to review.
Even though messaging is subjective, you can assess for the following:
- Does the AI articulate the pain points I'm solving, the persona-based messaging, and any industry-specific messaging well?
- Are the emails truly personalized?
“Bad AI” personalization is simple scraping output, like, if [this] then [that] without considering the holistic context around that account or buyer. That’s why they sound robotic or completely irrelevant.
“Good AI” personalization:
- Combines all the data points (signals) of what is happening at that account and buyers and any existing or past relationships you have with them
- Knows which data points (signals) are more important
- Can reference back to its previous messaging to the buyers so that it sounds natural & human-like
💎 𝗦𝘁𝗲𝗽 3: 𝗘𝘃𝗮𝗹𝘂𝗮𝘁𝗲 𝘂𝘀𝗮𝗯𝗶𝗹𝗶𝘁𝘆
One of the key factors for any new initiative is strong user adoption. For AI initiatives, usability isn’t only for the end-users (i.e. SDRs and AEs) but also for the business users & admins (i.e. SDR/marketing/sales managers and Ops)
Depending on the level of AI and technical skills of your existing team, look for solutions that most of your teams can use, maintain, and scale.
Overly technical solutions could cause:
- Challenges in hiring & backfilling the technical resource
- A bottleneck for the existing technical resource
- Your team spending time tooling & bug fixing instead of focusing on the business outcomes
-------
Follow this checklist and you'll probably save yourself & your team a lot of time and headache.