A mid-market support leader showed me her renewal quote from one of the big AI customer support platforms last quarter. Year one had been $480,000. Year three's projected number was $1.4 million. The volume hadn't tripled. The price had. Her CFO wanted to know why a tool that "automates customer support" was now the third-largest line item in the operations budget.
This is the question every buyer of AI customer support has to answer, and almost no vendor will help you answer it honestly: should I buy a platform like Sierra, Ada, Decagon, Intercom Fin, Salesforce Agentforce, or Zendesk AI - or should I build a custom AI support agent that I actually own? What does either choice cost over three years, and what do I have at the end?
This post is a buyer's guide to that decision for AI customer support specifically. You will get the actual three-year math at three volume tiers, named pricing for the major vendors, a decision matrix you can hand to your CFO, the hidden costs on both sides of the build-vs-buy line, and a clear answer to "when does each model win." If you only have time for the headline: at any meaningful support volume, the math points to build, and the gap widens every year you wait.
The Build vs Buy Question for AI Customer Support, Reframed
Most buyers start with "which AI customer support vendor?" That is the wrong first question. Before you compare Sierra to Ada to Decagon to Intercom Fin to Salesforce Agentforce to Zendesk AI, you have to decide which buying model fits your support function. There are three, and they have wildly different cost trajectories.
Buy: Consumption-Based Customer Support Platforms. Sierra, Ada, Decagon, Intercom Fin, Forethought, and most of the standalone AI customer support vendors. You pay per conversation, per resolution, or per interaction. Costs scale linearly with support volume - or worse, because tiered pricing often punishes growth. Annual contracts typically carry minimums of $30K to $300K or more. You are renting access. Stop paying and the agent disappears, along with every workflow, prompt, and tuning improvement you funded.
Buy-In: Ecosystem-Native Customer Support AI. Salesforce Agentforce, Zendesk AI, Microsoft Dynamics Copilot, HubSpot Breeze, ServiceNow Now Assist. The AI is bundled into a CRM or service desk you already use. Pricing flows through per-user licenses, AI credit packs, or new tiered plans. Integration with the parent platform is deep; integration with anything else is limited. You are extending a system you already pay for, with new cost layers on top.
Build: Custom-Built AI Support Agents. A flat monthly retainer (or one-time engagement) for design, build, and ongoing optimization of AI agents that handle your support tickets, voice calls, and messaging. The agent is built around your specific support workflows, deployed in infrastructure you control, and integrated with whatever helpdesk, CRM, billing, and knowledge-base stack you already run. At the end of the engagement, the software is yours. This is the model OpenNash uses, and it is also how teams like Klarna and LinkedIn approach their internal support agent builds, as documented in LangChain's enterprise case studies.
The mistake most buyers make is comparing a Sierra or Ada monthly bill to a build-side monthly retainer, deciding the platform is "cheaper," and never running the actual three-year math. Run it now.
The 3-Year TCO of AI Customer Support at 30K Tickets Per Month
Scenario: a mid-market company handling 30,000 AI-managed customer support interactions per month - chat, email, voice, and in-app messaging combined. Not a small SMB. Not a Fortune 100. The kind of business that has real ticket volume, real workflow complexity, and a CFO who reads contracts.
Buy: Sierra, Ada, Decagon, or Intercom Fin (Consumption-Based)
| Cost Component | Monthly | Annual | 3-Year |
|---|---|---|---|
| Per-interaction fee ($1.50 avg) | $45,000 | $540,000 | $1,620,000 |
| Platform license fee | $5,000 | $60,000 | $180,000 |
| Implementation (Year 1) | - | $50,000 | $50,000 |
| Total | $1,850,000 |
That $1.50 average per interaction is conservative. Sierra's outcome-based pricing routinely lands at $2-$5 per resolved conversation. Ada and Decagon negotiate per-resolution rates in a similar band for mid-market deals. Intercom Fin is closer to $0.99 per resolution at list. Reinventing.ai's 2026 SMB pricing analysis puts most enterprise AI customer support platforms in a $0.99 to $2.50 range per resolved conversation, with successful resolutions priced higher. If support volume grows to 50K per month by Year 3 - a reasonable assumption for a growing business - the three-year total approaches $2.5M.
Buy-In: Salesforce Agentforce or Zendesk AI (Ecosystem-Native)
| Cost Component | Monthly | Annual | 3-Year |
|---|---|---|---|
| AI conversations ($2/conversation) | $60,000 | $720,000 | $2,160,000 |
| Underlying platform licenses (20 users) | $8,000 | $96,000 | $288,000 |
| Implementation and admin | $6,250 | $75,000 | $150,000 |
| Total | $2,598,000 |
Agentforce is more expensive than Sierra or Ada because you are paying for the AI layer plus the underlying Service Cloud seats it requires. Salesforce's Flex Credit model can soften the per-conversation cost, but only by adding contract complexity and consumption forecasting that most support operations teams aren't equipped to manage. Zendesk AI follows a similar pattern: AI Copilot and AI Agents are priced on top of your existing Suite seats, with autoresolution credits billed per conversation.
Build: Custom-Built AI Support Agents (OpenNash Retainer)
| Cost Component | Monthly | Annual | 3-Year |
|---|---|---|---|
| Build and optimization retainer | $7,995 | $95,940 | $287,820 |
| Infrastructure (your environment) | $500 | $6,000 | $18,000 |
| Total | $305,820 |
At 30K support interactions per month, custom-built costs roughly 12-15% of Sierra, Ada, Decagon, or Intercom Fin over three years - and roughly 12% of Salesforce Agentforce. The delta only widens as ticket volume grows, because the marginal cost of an additional interaction in a custom build is essentially the LLM API call - cents, not dollars. And at the end of the engagement, you keep the software.
The AlphaCorp 2026 pricing guide puts mid-complexity custom builds in the $50K-$180K range for one-time development, which lines up with the retainer math when you amortize across three years of ongoing improvements.
Where the Crossover Actually Sits
Custom is not always cheaper. The honest version of this analysis names the volume range where each model wins.
Below 2,000 support interactions per month, consumption platforms like Intercom Fin or Ada are usually the better financial choice. Total spend is small enough that the upfront engineering investment in a custom build doesn't amortize meaningfully. You are also probably moving fast and iterating on what your support function even looks like. Pay-per-use makes sense.
Between 2,000 and 10,000 support interactions per month is the contested zone. The math depends heavily on your per-resolution rate, your platform's overage policies, and how complex your support workflows are. A simple FAQ deflection bot at 5K conversations per month might still be cheaper on Sierra or Ada. A multi-system support workflow that touches CRM, billing, order management, and inventory at the same volume usually isn't.
Above 10,000 support interactions per month, custom builds win on a three-year basis in nearly every scenario. The variable cost of Sierra, Ada, Decagon, or Agentforce pricing simply outruns the fixed cost of a build. By 30K per month you are not in the same conversation. By 100K per month you are looking at platform bills that exceed the cost of an in-house engineering team.
Aisera's build-vs-buy analysis lands on a similar crossover, noting that organizations with predictable, high-volume use cases consistently get better unit economics from owned systems than from per-conversation pricing.
The kicker most cost models miss: even below the crossover, custom gives you an asset. Platform spending is rent. Custom spending builds equity in software your business owns.
Hidden Costs Nobody Includes But Should
Both sides of this debate have hidden costs that don't show up in the sales pitch. Honest comparison requires naming them.
Hidden costs of buying Sierra, Ada, Decagon, Intercom Fin, or Agentforce
- Volume overages during demand spikes. Holiday seasons, product launches, outages, PR crises - exactly the moments when your support system has to work, and exactly the moments when per-conversation pricing punishes you. Some contracts cap overages; many don't.
- Escalation charges. When the AI can't resolve a ticket and hands off to a human agent, several platforms still charge for the AI portion of the interaction. You pay for the failure.
- Integration costs. Connecting Sierra or Ada to your CRM, billing, order management, knowledge base, and internal tools is usually a separate professional services line item, often $25K-$100K depending on complexity. Agentforce hides this cost in mandatory Service Cloud seats.
- Migration costs. If you switch from Sierra to Ada (or off Agentforce entirely) in Year 3, you start from zero. No data portability, no workflow portability, no model improvements come with you. The Quickchat AI 2026 buyer's guide flags this as the single most underestimated cost in platform selection.
- Roadmap dependency. Your support strategy is constrained by what your vendor decides to build next. If your business needs a feature their roadmap doesn't include, you wait or work around it.
Hidden costs of building
Custom builds are not free of tradeoffs. The honest version:
- Longer time-to-value. Platforms can be live in days. A well-built custom agent takes weeks for an MVP and months to reach production-grade reliability. If you need automation running by next Tuesday, custom is not the answer.
- Partner dependency. A custom build requires a partner who actually understands your business. That relationship has to work, and the partner has to stay engaged for ongoing optimization. This is real risk - choose carefully.
- Ongoing maintenance is not free. LLMs change. APIs change. Your business changes. Someone has to keep the agent current. The retainer model handles this, but it is a real cost to budget.
- Less plug-and-play. Turning on a platform feature is a checkbox. Building a new capability in a custom system requires actual engineering work.
The honest framing: platforms trade long-term cost for short-term speed. Custom trades short-term speed for long-term cost advantage and ownership. The question is which tradeoff matches your business.
What Doesn't Show Up in a Spreadsheet
Cost is the easiest part of this decision to model and the least interesting. The real strategic differences are harder to quantify.
Ownership as exit value. Custom AI software is an asset on your balance sheet. Platform subscriptions are an operating expense. If you sell the company, the acquirer pays for what you own; what you rent disappears at renewal. This matters more than most operators realize until they are in due diligence.
Competitive differentiation. Every company in your industry can buy the same Sierra, Ada, or Decagon deployment. Nobody else has support software built specifically for your operations, your product, your customers, your edge cases. This is the same logic that drove early SaaS buyers to eventually build custom internal tools - and it applies harder to AI customer support, where the system is making decisions on your behalf in front of customers.
Vendor independence. Platform pricing changes. Vendors get acquired. Roadmaps shift. As Simon Willison has documented in his work on agent architectures, the underlying primitives of agent systems - tool use, retrieval, planning loops - are increasingly commodified. Paying enterprise pricing for primitives that are becoming cheap is a strategic error.
Flexibility across use cases. A platform built for customer support chat doesn't extend to voice support, internal operations, compliance review, or revenue ops. Sierra doesn't do AP automation. Ada doesn't do RFP response. Each new use case means another vendor, another contract, another integration. A custom architecture extends to new use cases at marginal cost.
A Decision Matrix You Can Hand to Your CFO
If you read nothing else, use this:
| Your Situation | Best Model |
|---|---|
| Low support volume (under 2K tickets/month), need to launch this week | Buy: Intercom Fin, Ada, or Sierra |
| Already deep in Salesforce Service Cloud or Zendesk Suite, basic AI deflection | Buy-In: Agentforce or Zendesk AI |
| High support volume (over 10K tickets/month), cost-sensitive | Build: Custom |
| Support workflows spanning CRM, billing, order management, inventory | Build: Custom |
| Regulated industry (healthcare, financial services, legal) | Build: Custom |
| Building IP or planning a future acquisition | Build: Custom |
| Need agents beyond customer support (voice, SDR, finance ops, RFP) | Build: Custom |
| Small team, no engineering, single FAQ deflection use case | Buy: Intercom Fin or Ada |
Platforms win on speed and simplicity. Custom wins on volume, complexity, ownership, and strategic flexibility. Most mid-market and up companies eventually land in the custom column - the only question is whether they spend three years funding a vendor's growth before they figure that out.
Don't start the conversation with "which platform should I buy?" Start with "Do I want to rent automation or own it?" If you handle significant volume, your workflows are specific to your business, and ownership matters to your long-term strategy, the math almost always points to build.
How OpenNash CX Builds Custom AI Customer Support for Less Than Sierra or Ada
The objection most buyers raise about building is that it sounds expensive and slow. That assumption is based on a market that no longer exists. The right partner can deliver production-ready custom AI customer support at a price that competes directly with - and usually beats - Sierra, Ada, Decagon, Intercom Fin, and Salesforce Agentforce, on a timeline measured in weeks instead of quarters.
This is exactly what OpenNash CX does:
- Flat $7,995/month retainer. No per-conversation fees. No per-resolution fees. No overage charges. No surprise renewals. Unlimited requests, 48-hour average turnaround, pause anytime.
- OpenNash CX included. Inbound support, voice agents (calls, scheduling, intake), and SMS/email triage built on our AI-native customer support stack come with the retainer. That alone replaces a Sierra, Ada, or Decagon subscription that would run $40K-$80K/month at 30K interactions.
- Production-ready, not proof-of-concept. We deploy what other agencies only demo. Audit your support workflows, design with guardrails, build, deploy in your infrastructure - the same four-step path for every engagement.
- You own everything. Full documentation, source code, infrastructure access. No vendor lock-in. If you ever stop working with us, the support agents keep running.
- 1,000+ native integrations. Salesforce Service Cloud, Zendesk, HubSpot, Intercom, Stripe, Shopify, Snowflake, Workday, Notion, Google Cloud, Slack - whatever helpdesk and CRM stack you already run, we plug into it.
- 14-day risk-free pilot. Every engagement starts with one workflow, scoped together and shipped in 14 days. If it doesn't pass the test cases you sign off on, you don't pay.
The retainer covers more than customer support. The same engagement can ship SDR and outbound agents, finance ops automation (AP/AR, expense auditing, month-end close), revenue ops agents, RFP response agents, compliance monitoring, document processing, and AI copilots for your team. One flat fee, the full custom AI agent stack - not just a chat widget.
For a mid-market team handling 30K support interactions per month, the math is straightforward: $7,995/month with OpenNash CX, or $50K-$80K/month with Sierra, Ada, or Agentforce that you will never own. Over three years, that is the difference between $306K and $1.85M - and at the end of three years, one of those paths leaves you with software, the other leaves you with a renewal quote.
Want to see what build-vs-buy looks like for your specific support volume and workflows? Book a 30-minute call with OpenNash CX and we will walk through your numbers against Sierra, Ada, Decagon, Intercom Fin, or Agentforce side by side. If buying is the right answer for you, we will tell you that too.
Sierra, Ada, Decagon, and Agentforce all know the three-year math. It is why they don't publish it. Now you have it.