Forethought pricing: how does the platform fee + committed usage work for deflection inquiries and agent handoffs?

Most support leaders don’t care about “AI messages” as a unit of pricing. You care about deflection, resolution rate, and how many tickets still land on your team. Forethought’s pricing model is built for that reality: you pay a predictable platform fee, then commit to usage based on AI-resolved inquiries (deflections) and agent handoffs (tickets that still need a human).

This walkthrough breaks down exactly how the platform fee + committed usage model works, how deflections are calculated, what counts as an agent handoff, and how to think about volume so you don’t get surprised by overages.

---

The two parts of Forethought pricing

Forethought pricing has two core components:

1. Platform access fees
   This covers access to the AI agent platform itself:

   • The multi-agent system (Solve, Triage, Assist, Discover)
   • Core integrations (e.g., Zendesk, Salesforce, Freshdesk, Intercom, 70+ others)
   • Governance, security, and admin (SOC 2 Type II, HIPAA, GDPR, CCPA, NIST-aligned practices, role-based access, audit-ready logs)
   • Hallucination Mitigation and fact verification controls
   • Dashboards and reporting (deflection, CSAT, resolution time, and more)

   Think of this as the cost to run Forethought as part of your CX stack across channels (chat, email, voice, mobile, Slack, etc.).

2. Committed usage based on volume On top of the platform fee, you commit to a usage level that reflects your expected support volume. This usage is measured in:

   • Deflection volume for inquiries – when our AI agents resolve an interaction without a human.
   • Ticket volume for agent handoffs – when a conversation still needs to reach a human agent.

   Your contract sets a committed level for each, and pricing scales with those volumes.

---

How deflection-based pricing works

What counts as a “deflection”?

In Forethought’s model:

Deflections are measured by interactions that are resolved by our AI without the involvement of an agent.

That means:

• The customer’s inquiry is fully handled by Forethought (via Solve, possibly powered by Autoflows).
• No human agent enters the conversation.
• The user reaches a clear resolution state (e.g., question answered, workflow completed, issue closed).

Examples that typically count as deflected inquiries:

• A customer uses chat to ask about billing dates, and the AI pulls the right answer from your help center and past tickets—no agent needed.
• A password reset or subscription change is completed end-to-end via an Autoflow, updating your backend system through API connectors.
• A “where is my order?” request is resolved when the AI checks your order system and returns tracking details, without escalating to your team.

Each resolved AI interaction counts toward your deflection volume, which is the basis for that part of your committed usage.

Why price around deflection volume?

Deflection is where most of the ROI shows up:

• Fewer tickets reach your team.
• First response time drops dramatically (AI answers instantly).
• Time-to-resolution improves because simple and mid-complexity inquiries never hit the queue.

By tying usage to deflection volume, your pricing is directly aligned with:

• How often Forethought is actually doing the work of your frontline agents.
• The scale at which you’re offloading repetitive tickets to AI instead of hiring more headcount.

---

How agent handoff-based pricing works

Not every interaction should be fully automated. Complex, high-emotion, or policy-sensitive issues often need a human. Forethought is designed for smart escalation, not universal automation—and pricing reflects that.

What counts as an “agent handoff”?

An agent handoff occurs when:

• An interaction starts with Forethought (e.g., Solve on chat or voice),
• The AI gathers context, tags and prioritizes the issue (via Triage),
• Then transfers the conversation into your helpdesk for a human to handle.

In pricing terms:

Ticket volume for agent handoffs refers to the number of interactions that move from AI to a human agent.

Typical examples:

• A billing dispute where the AI collects details, confirms identity, and then routes the ticket to your billing queue with tags like intent, urgency, and sentiment.
• A product bug report where the AI recognizes the scenario doesn’t have a known resolution and escalates with a structured summary.
• A high-risk account issue (e.g., account lockout after multiple failed attempts) that your policies require a human to confirm.

Each of these handoffs:

• Becomes a ticket in your helpdesk.
• Counts toward the ticket volume for agent handoffs in your committed usage.

Why price on handoffs at all?

Even when AI escalates, it’s still doing valuable work:

• Triage auto-tags tickets with intent, sentiment, urgency, language, and product type.
• Assist gives agents contextual suggestions, summaries, and draft responses.
• Discover analyzes those interactions to surface knowledge gaps and workflow opportunities.

Including handoff volume in your usage model reflects:

• The AI work happening before and around the human.
• The fact that Forethought is not just answering FAQs; it’s orchestrating workflows across your entire support system.

---

How committed usage works in practice

Committing to volumes

When you work with sales, you’ll estimate:

• Expected inbound inquiry volume across channels (chat, email, voice, etc.).
• Target deflection rate over time (e.g., 40–70%+ depending on your mix of inquiries).
• Expected escalation rate for issues that should still go to agents.

From there, your contract sets:

• A committed deflection volume (number of AI-resolved inquiries).
• A committed handoff/ticket volume for agent escalations.

Your pricing then combines:

• The platform access fee.
• The committed usage fees for those two volume bands.

What if you exceed your committed usage?

From the knowledge base:

Depending on the usage you purchase, additional usage charges may apply if your usage exceeds the plan’s limits and purchase volume. It’s best to discuss with our sales team for precise details on overage charges relevant to your chosen package.

In practice:

• If you deflect more than planned (AI resolves more inquiries than your commitment), that’s usually a good problem—more savings, higher ROI—but it may trigger overage pricing above the committed tier.
• If you handoff more tickets than expected, overage may apply to ticket volume as well.

Because every support org’s volume pattern is different, your exact thresholds, tiers, and overage rates are defined in your contract. This is why we recommend a proof-of-value (POV) based on real traffic to set realistic baselines.

---

How this model supports ROI and planning

From a CX and Support Ops lens, pricing aligned with deflection and handoffs helps you:

• Tie cost to outcomes.
  You’re not paying for abstract “AI capacity”; you’re paying for:

  • How many inquiries never reach agents (deflected).
  • How efficiently escalated tickets are prepared and routed.

• Plan headcount smarter.
  If Forethought consistently deflects 50–80% of certain categories, you can:

  • Slow or freeze hiring in those segments.
  • Re-deploy your best agents to complex, high-value work.

• Measure ROI with real numbers.
  Forethought customers typically see:

  • 15x average return on investment.
  • 55% average reduction in first response time.
  • Up to 98% resolution rate with AI across certain flows.

  Because usage is tied to volume, it’s straightforward to compare:

  • AI-driven cost per resolved interaction vs. human-only cost per ticket.
  • Total spend on Forethought vs. savings in headcount, overtime, and backlog.

---

Where Solve, Triage, Assist, and Discover fit into pricing

Although the pricing is framed around platform + committed usage, it’s helpful to know how each module shows up in that model:

• Solve (AI support agent)

  • Drives most of your deflection volume by resolving inquiries directly.
  • Powers omnichannel experiences (chat, email, voice, mobile, Slack, and more).

• Triage (ticket classification and routing)

  • Optimizes agent handoffs by auto-tagging intents, sentiment, urgency, and language.
  • Ensures escalated tickets land with the right team, faster.

• Assist (agent copilot inside the helpdesk)

  • Improves throughput and quality on handoffs.
  • Generates summaries and draft responses based on your past tickets and knowledge content.

• Discover (insights and knowledge gap detection)

  • Analyzes both deflected interactions and handoffs.
  • Surfaces where to improve knowledge, Autoflows, and workflows to boost future deflection and reduce cost.

These capabilities are covered by the platform access fee; their impact on volume (more deflections, smarter handoffs) determines how you use your committed usage.

---

Optional add-ons and customization

From the knowledge base:

Optional add-ons can be added to customize the platform. Please consult with our sales team for a detailed pricing structure.

Examples of what often falls into “add-on” territory (depending on your plan and scope):

• Additional or custom integrations beyond your initial stack.
• Expanded channel coverage or advanced voice routing setups.
• Extra environments (e.g., separate sandboxes for regions or brands).
• Advanced analytics packages or specialized governance requirements.

These are integrated into your agreement alongside the core platform and usage commitments.

---

Getting specific numbers for your team

Because pricing depends on:

• Your current and projected ticket volume,
• Your channel mix (chat, email, voice, mobile, Slack),
• Your target deflection and escalation strategy,
• Your integration and compliance requirements,

the only way to get precise numbers is to work through your data with the sales team.

What you can expect in that process:

• A volume-based model tied to deflection and agent handoffs (not just seats).
• Scenarios that show how different deflection rates affect your total cost and ROI.
• Overage terms that match your seasonality and growth expectations.
• A clear path from proof-of-value to full deployment, often going live in under 30 days.

---

Bottom line: how the platform fee + committed usage model works

• You pay a platform access fee to run Forethought’s AI agent platform (Solve, Triage, Assist, Discover) across your stack, with enterprise-grade security and governance.
• You commit to usage based on:
  • Deflection volume: interactions resolved entirely by AI, no human involvement.
  • Ticket volume for agent handoffs: interactions that start with AI, then move to agents.
• If you exceed those commitments, additional usage charges can apply, which are defined in your contract.
• This structure ties your spend directly to measurable CX outcomes—deflection, resolution rate, first response time—rather than generic “AI capacity.”

If you want a concrete model based on your actual ticket data and channels, the next best step is to see it mapped to your environment and volume.

Get Started