Why enterprise chatbot contracts are shaped for the wrong buyer

By Mike Berris · April 22, 2026 · 7 min read

Enterprise AI chatbot platforms — Cognigy, NICE, Kore.ai, similar — charge $75,000 to $300,000+ per year for a license, billed LLM inference separately on top, plus $50,000 to $100,000+ in professional services to get the thing into production. Total year-one cost: commonly north of $200,000.

That math makes sense for exactly one kind of customer: a Fortune 500 contact center running multi-channel voice + SMS + WhatsApp + web chat + in-app messaging + email at massive scale, with regulatory requirements that demand SOC 2 + ISO 27001 + dedicated account management and a procurement cycle measured in quarters.

For everyone else — the mid-market company with 10,000 website visitors a month, the service business with three locations, the B2B SaaS startup running a lead funnel — those contracts are sized for a use case they don't have and a budget they don't have.

The shape of the enterprise chatbot contract tells you who it was designed for. Let's walk through the tells.

Tell #1: LLM inference is billed separately

Enterprise platforms don't include AI inference in the license. You pay the platform fee, and then you pay per-token to whichever model vendor (OpenAI, Anthropic, Google) the platform is using.

This is fine if:

• Your finance team wants to see the cost breakdown by vendor.
• You have a dedicated FinOps person who optimizes model spend month-over-month.
• You're running enough volume to negotiate direct enterprise contracts with model vendors.

It's terrible if:

• You want a predictable monthly bill.
• You want to focus on business outcomes, not on model-spend optimization.
• A traffic spike shouldn't be a budget event.

For SMB and mid-market buyers, "AI included" isn't a product feature — it's a financial product. It's the difference between a predictable cost center and a mystery bill.

Tell #2: The platform is a canvas, not a product

Enterprise chatbot platforms are flow builders. You log in and see a blank canvas. You drag nodes, configure branches, connect APIs, write fallback logic, handle edge cases, configure multi-language support, design escalation paths, wire in your CRM.

This is fine if:

• You have a team of 3-10 people whose full-time job is building and maintaining conversational flows.
• You have a professional services partner who's done this before for similar use cases.
• You have 3-6 months of runway before the project needs to show ROI.

It's terrible if:

• You're a marketing team of 2 trying to get a lead bot live this quarter.
• You don't know yet what the right flow looks like — you want to start from a sensible default and tune.
• You can't afford a 6-month implementation.

The canvas model assumes you know exactly what you want to build. Most SMB and mid-market buyers don't. They want a starting point that works out of the box, which they can tune over time.

Tell #3: Implementation is a separate line item

Enterprise contracts include a "professional services" component that's 30-50% of the license cost. This is where the platform partner (or the vendor's own services team) does the actual work of building the bot.

On a $150,000/yr license, that's another $45,000-$75,000 in services just to get to a working bot. After that, you're still paying the license every year.

This is fine if:

• You have the budget for it.
• Your procurement process is designed to bundle professional services into SaaS contracts.
• You think of the chatbot as a 3-year capital project, not a this-quarter tool.

It's terrible if:

• You want to try the product before committing six figures.
• You want the price on the pricing page to be the price you pay.
• You want to self-serve into production.

Tell #4: The platform is multi-channel by default

Enterprise platforms support 25+ channels — voice IVR, SMS, WhatsApp, Telegram, Instagram, Microsoft Teams, Google Business Messages, Apple Business Chat, and web chat somewhere in the mix.

If you need omnichannel coverage, this is great. You build once, deploy everywhere.

If you only need web chat? You're paying for 24 channels you'll never use, and the complexity of a platform built for omnichannel works against you when your job is "get a chatbot on the marketing site."

Most SMB and mid-market buyers are web-only, with SMS and WhatsApp as "nice to have someday" channels. Paying enterprise omnichannel pricing to use 1-2 channels is a tax on the shape of the platform, not a feature you're buying.

Tell #5: Procurement is a multi-month process

Enterprise platforms expect an RFP. A security review. An MSA with negotiated terms. A dedicated success manager. Quarterly business reviews. A procurement committee that meets once a month.

This is fine if you're a Fortune 500 with a procurement function designed for six-figure SaaS contracts. It's the only way those organizations can safely buy anything.

If you're a 50-person company whose "procurement" is your ops lead swiping a credit card, the enterprise procurement cycle is a multi-month delay with nothing on the other end except permission to start implementation.

What fits the non-enterprise buyer

If you're not Fortune 500 — and 99% of the businesses reading this aren't — your chatbot platform should:

1. Include AI inference in the license. Predictable bills.
2. Ship a working product out of the box, not a canvas. Auto-setup that gives you a sensible default you can tune.
3. Have implementation baked in, at zero professional services cost. If a 5-person team can't set it up in a day, it's not built for you.
4. Focus on the channels you actually use. Web + maybe SMS. Not 25 channels you'll never touch.
5. Let you self-serve into production. A 14-day trial. A credit card. Public pricing. No RFP.

That shape — which is what Wengrow is built for — doesn't replace enterprise platforms. It serves a market they're not trying to serve.

Both markets are real. Most buyers are in the second one.

What we charge

Wengrow is $49 to $7,499 per month. AI inference is included. No implementation services are needed. You self-serve into a 14-day trial. All features are on the pricing page.

If you need SOC 2 today, we're not your vendor (yet). If you need voice bots or WhatsApp, not yet. If you need a dedicated success manager, that's a Business+ conversation.

For everything else — which is most of the market — we're built for you. See pricing →