SaaStr, a prominent voice in the SaaS industry, recently integrated Salesforce Agentforce into its AI stack, specifically targeting a common yet costly problem: neglected sales leads. In just over a month, this AI agent has delivered "transformational" results, achieving a remarkable 72% open rate on emails sent to previously "ghosted" contacts and successfully closing deals that were once considered lost opportunities. This early success demonstrates the powerful potential of AI in revitalizing sales pipelines and optimizing lead follow-up processes.
Addressing the "Ghosted Lead" Gap
SaaStr had already deployed over ten AI agents across various functions, including outbound sales and instant on-site conversations. However, a significant gap remained. Following the recent 2025 SaaStr Annual + AI Summit in May, the team discovered approximately 1,000 leads—individuals who had explicitly requested information about SaaStr Annual—had received no follow-up from the sales team. These weren't cold leads, nor were they actively engaged in real-time conversations; they were interested prospects already in their Salesforce CRM who had simply been forgotten. SaaStr needed an AI solution capable of leveraging their extensive Salesforce data to address this specific "middle ground" use case.
Why Agentforce Was the Right Fit
Agentforce stood out due to its native integration with Salesforce. Unlike other AI agents that might connect to Salesforce, Agentforce's inherent ability to access and utilize all existing CRM data was a game-changer. This meant the AI could draw upon years of past interactions, company information, previous event attendance, and any notes within a contact's record. This deep contextual understanding allowed Agentforce to send highly personalized follow-ups at scale, eliminating the need for manual management of each contact and effectively reviving the 1,000 ghosted leads.
Setting Up Agentforce: Easier Than Expected
Despite initial concerns about requiring a dedicated Salesforce administrator, SaaStr found the setup process surprisingly straightforward. Their Chief AI Officer, not Salesforce certified, received crucial assistance from Salesforce's Forward Deployed Engineers (FDEs). This hands-on support proved vital, demystifying the deployment of complex enterprise AI agents. SaaStr also discovered that once proficient with one AI tool, the underlying concepts often transfer, allowing them to adapt successful prompts from other AI agents for Agentforce's specific use case.
Early Results: A Month of Transformational Impact
Launched at Dreamforce just over a month ago, Agentforce has already yielded impressive metrics:
- 72% Open Rate: A stark contrast to the 0% open rate for emails that were never sent by human sales reps.
- Highest Response Rate: Among all of SaaStr's AI platforms, Agentforce achieved the highest response rate.
- ~3,000 Emails Sent: To the initial list of 1,000 ghosted leads.
- Deals Already Closing: From contacts who had received no follow-up six months prior, generating significant revenue from ticket sales to sponsorship packages.
The 72% open rate is particularly notable, far exceeding typical generic email marketing benchmarks and underscoring the power of highly contextualized, personalized outreach.
The magic behind these results lies in Agentforce's ability to pull rich context directly from Salesforce data, including:
- Past interactions with SaaStr, often spanning several years.
- Comprehensive company information.
- Records of previous event attendance.
- Any notes or data stored in the contact's record.
This contextual awareness enables the AI to craft messages that resonate deeply with recipients. For instance, an agent referenced a lead's past interaction and company details, prompting an immediate response even if it was a "no thanks, but I'll attend again." Such engagement was impossible when these leads were simply forgotten.
The Strategic Value of "Low-Hanging Fruit"
While some might argue that starting with "raised-hand" leads is an "easy" use case, SaaStr views it as strategically sound. These leads, already in their Salesforce CRM with existing context, presented an ideal opportunity for AI to acknowledge past engagement and offer relevant assistance. This "middle ground" audience—neither ice-cold nor actively conversing—was perfectly suited for Agentforce to re-engage with personalized notes, effectively turning forgotten prospects into active opportunities. Salesforce itself recognized a similar opportunity, identifying 100 million leads that had never been followed up on.
Campaign Triggering and Automation
SaaStr can trigger Agentforce campaigns at the lead, contact, or campaign level. Initially, campaigns were managed manually to ensure no mass spamming. However, Salesforce has since streamlined the process, allowing SaaStr to:
- Add contacts to a Salesforce campaign.
- Activate the entire campaign on the AI agent.
- Let Agentforce intelligently manage send times, considering time zones and optimal engagement periods.
This smart queuing ensures personalized delivery without overwhelming recipients.
Current Limitations and Future Plans
SaaStr has deliberately limited Agentforce's autonomy in its initial phase, for example, by not empowering it to book meetings directly. Instead, it asks for meetings, allowing human oversight for complex, multi-threaded inquiries (e.g., questions about speaking and sponsorships in a single email). This cautious approach builds trust and understanding before granting the AI agent more advanced functionalities, a strategy SaaStr applies to all its AI deployments.
The Reality: AI Requires Time and Investment
SaaStr emphasizes that Agentforce, like any complex AI agent, is not a "set-it-and-forget-it" solution. It demands ongoing investment in time, training, and iteration. The quality of the agent's output directly correlates with the effort put into reviewing draft responses, refining prompt builder instructions, and strategically defining target or exclusion campaigns within Salesforce. While the agent can operate on autopilot, consistent engagement from human experts leads to superior performance and higher response rates. AI agents clone best processes but require continuous refinement of those processes.
Costs and Budgeting Insights
Enterprise AI tools like Agentforce typically fall within the $50,000-$100,000 annual range, factoring in licenses, training, and onboarding. SaaStr funded Agentforce by reallocating headcount budget during natural team attrition, highlighting a strategic shift from human roles to AI agents when opportunities arise. Beyond direct costs, the most critical investment is the time dedicated by top personnel to ensure the AI's effective operation.
SaaStr's Advice for Agentforce Adoption
SaaStr offers clear recommendations for organizations considering Agentforce:
- Do:
- Start with a clear, specific use case, such as re-engaging ghosted leads.
- Leverage support from Salesforce FDEs during setup.
- Adapt successful prompts from other AI tools.
- Be patient through the learning curve.
- Insist on speaking with technical experts, not just sales representatives.
- Don't:
- Assume you need to be a Salesforce admin.
- Expect the AI to magically fix broken processes.
- Attempt to deploy all functionalities at once.
- Give up after a single negative experience with a sales rep; seek technical assistance.
Fun to see the SaaStr himself Jason using Agentforce for sales — and finally calling customers back! 🤖 I’m doing that too… it’s incredible what happens when your AI actually helps you sell! 🤯💰 https://t.co/3YTvuDeq47
— Marc Benioff (@Benioff) November 12, 2025
For further insights into AI SDRs, SaaStr recommends reviewing Marc Benioff's perspective: Do AI SDRs Work? You Better Believe It, Per Marc Benioff.
The Bottom Line
After just over 30 days, SaaStr is still in the early stages of its Agentforce deployment, with plans to significantly expand its use. However, the initial results—particularly a 72% open rate on a segment that previously received zero follow-up—are not merely incremental; they are transformational. Those who dismissed Agentforce as overly technical or exclusive to large enterprises with dedicated admins are missing the point. Like any valuable AI tool, it requires investment in time, training, and iteration. For organizations willing to commit, Agentforce offers a powerful solution to bridge the vast "middle ground" between cold outbound and real-time inbound leads, effectively re-engaging forgotten prospects and unlocking new revenue streams. SaaStr anticipates sharing more data in Q1, but for any business grappling with neglected leads or contacts that have fallen through the cracks, Agentforce presents a compelling opportunity.
Want to see all our agents in action? Check out saastr.ai/agents for the full stack. And if you’re at SaaStr AI London next month, come find me – I’ll be diving deeper into all of this, including marketing and CS use cases we haven’t covered yet.
