AI Chatbot Pricing for Small Business: What Will You Really Pay?
A current, source-backed guide to flat plans, seats, conversation allowances and per-resolution fees—built for solo founders and small Shopify teams.
Author: By Di - Founder of DocMind
Published: 11 March 2026
Updated: 23 September 2026
Reviewed sources: 23 September 2026
A sale weekend is going well until the support bill becomes another surprise. A solo founder may see a low monthly price, add an AI agent, invite a teammate, exceed a conversation allowance and discover that four different charging units sit behind one product page. Comparing only the headline price hides the decision that matters.
TL;DR: AI chatbot pricing for a small business is easiest to compare as a total monthly workload: platform, seats, AI outcomes, conversation limits, required integrations and review time. Flat plans offer budget certainty; usage pricing can align cost with results. Model normal and peak months, then compare cost per clean resolution.
What does an AI chatbot cost for a small business?
Small-business AI chatbot costs range from free pilots to fixed monthly plans and usage-based bills that grow with seats, conversations or successful AI outcomes. The correct comparison is not one sticker price. It is the complete cost of handling the same support workload, with the same channels, integrations, people and quality controls.
Five pricing models appear repeatedly. A vendor can also combine several of them, so read the unit beside the price before comparing totals.
| Model | How it works | Useful strength | Budget risk |
|---|---|---|---|
| Flat monthly plan | A known subscription with stated capacity limits | Forecastable spend; simple approval | Unused capacity or a forced upgrade when a limit is reached |
| Per seat | Each full support user adds a recurring charge | Maps neatly to staffed helpdesk access | Cost can rise before support volume does |
| Per outcome or resolution | Each qualifying AI-resolved conversation is billable | Spend follows successful automation | Success can increase the invoice during busy periods |
| Conversation allowance | A plan includes a fixed number of human or AI conversations | Clear starting quota | Different conversation types may use different buckets |
| Custom/API usage | Tokens, calls, hosting and engineering are costed separately | Maximum control for technical teams | Monitoring, retrieval, evaluation and maintenance become your responsibility |
“Unlimited” rarely means every capability is unlimited. Check the limits for AI replies, human conversations, sources, bots, tickets, channels, integrations and people. Also ask whether unused capacity rolls over, what triggers an overage and whether the advertised price assumes annual billing.
How did we compare current chatbot pricing?
We reviewed public vendor pricing on 23 September 2026 using six disclosed criteria: entry cost, charging unit, included capacity, required companion products, overage exposure and intended customer. This is a public-information comparison, not an account-level trial, negotiated quote or claim that every product offers identical features.
The comparison includes DocMind because this is the DocMind blog. That creates an obvious commercial interest, so the same fields are shown for every option, DocMind's principal limitation is stated, and conclusions are tied to the buyer's workload rather than a universal ranking.
| Option | Public price checked | Charging unit | Best fit | Main weakness |
|---|---|---|---|---|
| DocMind | Free; Starter US$9.90/mo; Personal US$29/mo; Standard US$59/mo | Monthly plan with published reply, message, source, bot and ticket limits | Solo founders and small stores that want answers from approved store content at a predictable base cost | Not a broad omnichannel enterprise helpdesk; team seats begin on Standard |
| Intercom + Fin | US$29/seat/mo Essential, plus US$0.99 per Fin outcome | Paid full seats plus usage; Fin can also run on another helpdesk with a minimum outcome commitment | Teams that want a mature inbox, ticketing and workflow platform with an AI agent | A small team must model both seats and successful AI outcomes, plus any optional add-ons |
| Gorgias Helpdesk + AI Agent | Helpdesk tier plus AI Agent; most AI plans US$0.90 per resolved interaction, Starter US$1 | Helpdesk ticket allowance and a separate AI resolved-interaction allowance | Established ecommerce teams that want a Shopify-focused helpdesk and AI in one workflow | Two linked pricing layers and overage exposure require a peak-month volume estimate |
| Tidio + Lyro | Starter US$24.17/mo annually; Lyro standalone from US$32.50/mo for 50 AI conversations | Billable human conversations, Lyro AI conversations and Flows visitors use separate quotas | Small teams combining live chat, ticketing and a limited AI conversation allowance | Several quota types make like-for-like cost modelling less immediate |
Prices above are public US-dollar figures where the vendor uses dollars. Taxes, exchange rates, annual-billing discounts, contracts and custom plans can change the amount paid. Recheck the linked vendor page before purchase.
Is flat pricing or usage-based pricing better?
Flat pricing is better when budget certainty matters and expected use fits the published limits; usage pricing is better when you want spend tied to actual activity or outcomes. Neither model wins automatically. A flat plan can waste unused capacity, while a usage plan can become expensive precisely when a campaign drives more customers and questions.
For example, Intercom publicly lists Essential at US$29 per full seat each month plus US$0.99 for each Fin outcome. At 500 qualifying outcomes, the outcome component alone is US$495; one Essential seat would make the simple public-price subtotal US$524 before optional add-ons, taxes or other usage.
Gorgias uses a different structure. Its official explanation says AI Agent is separate from the Helpdesk plan, most AI plans price a resolved interaction at US$0.90, Starter begins at US$1, and overages apply beyond the selected allowance. A human handoff is not billed as an AI-resolved interaction, according to that explanation.
Those examples are not quotes and do not compare complete feature parity. They show why every calculation needs a named billing unit. A “conversation”, “ticket”, “message”, “resolution” and “outcome” may represent different events.
How do you calculate the real monthly cost?
Calculate a normal month and a peak month using the same five inputs: human users, inbound conversations, clean AI resolutions, required channels and required integrations. Add setup and monthly review labour separately. Then divide the recurring total by clean resolutions—conversations solved without an avoidable repeat contact or unsafe answer.
Worked example: one-person Shopify store
Assume 600 monthly support conversations, 180 suitable for AI, one human user and a peak month with 360 suitable AI conversations. These are planning assumptions, not a DocMind customer result.
| Pricing structure | Normal-month calculation | Peak-month question |
|---|---|---|
| Flat plan | Use the tier that includes the required capacity and integration | Does peak usage exceed a plan limit or require an upgrade? |
| Seat + US$0.99/outcome | Seat subtotal + 180 × US$0.99 = seat subtotal + US$178.20 | Seat subtotal + 360 × US$0.99 = seat subtotal + US$356.40 |
| Helpdesk + US$0.90/resolution | Helpdesk tier + 180 × US$0.90 = helpdesk + US$162 | Does 360 exceed the selected allowance, and what is the overage rate? |
Next, add implementation hours, monthly source review, conversation review and exception handling. Use the chatbot ROI calculator to keep labour assumptions separate from software charges. If a vendor price changes with annual billing, show both the cash commitment and the monthly equivalent.
Which pricing model fits a small Shopify store?
A small Shopify store usually benefits from a low-risk pilot and predictable spend, but only if the plan supports the store questions that actually create work. Price the workflow for order status, shipping, returns, product details and handoff—not a generic message count that ignores whether the answer can use approved store information.
Start by tagging 100 recent questions. Mark which can be answered from a current product page, policy, FAQ or order-status source. Keep refunds, fraud, chargebacks, unusual damage claims and other judgment-heavy cases on a human path. That gives you a defendable first automation scope.
DocMind's current pricing page lists Free with 50 total AI replies, Starter at US$9.90 monthly, Personal at US$29 and Standard at US$59. Shopify order tracking is included on every paid plan; Standard adds team seats. Plans publish limits for replies, messages, sources, bots and tickets, with no per-resolution or per-seat overage charges inside those limits.
The main trade-off is scope. DocMind is designed for approved-content AI support and small-store workflows, not as a replacement for every omnichannel helpdesk function. A larger team that needs advanced inbox routing, many channels or complex workforce operations may reasonably choose a broader platform even if its pricing has more components.
Watch the workflow before pricing it
This DocMind walkthrough shows Shopify product, policy and order-question handling from approved store content. It demonstrates the workflow to evaluate; it is not evidence of a specific store's savings or a test of the other products in the comparison.
What hidden chatbot costs should you check?
The most common hidden costs are required base plans, paid seats, AI add-ons, overages, integration tiers, source cleanup and ongoing answer review. They are not necessarily unfair fees; many fund valuable capabilities. They become a problem when a buyer compares an incomplete entry price with another vendor's complete monthly plan.
- Companion product: Can the AI be bought alone, or does it require a helpdesk subscription?
- Billing definition: What makes an outcome, resolution, ticket or conversation chargeable?
- Annual commitment: Is the displayed monthly price available only when a year is paid or committed?
- Peak usage: Is there an overage, hard stop or automatic tier change?
- People: Which users need paid full seats, and which can use free collaborator access?
- Knowledge work: Who owns outdated policies, duplicate product facts and unanswered questions?
- Quality control: How much time will be spent reviewing weak answers, repeat contacts and escalations?
Tidio illustrates why definitions matter. Its public page separates billable human conversations, Lyro AI conversations and visitors reached with Flows. Starter is listed at US$24.17 monthly on annual billing with 100 billable conversations and 50 one-off Lyro conversations; standalone Lyro starts at US$32.50 for 50 AI conversations.
What should you ask before choosing a plan?
Ask each vendor to price the same normal and peak workload, then document what is included, excluded and billable. A strong buying decision should survive a busier month without a surprise invoice and should explain how the team will verify answers, update sources and hand off cases the AI should not resolve.
- Define the workload. Record monthly conversations, repetitive topics, channels and peak-season volume.
- Define a clean resolution. Exclude avoidable repeat contacts, unsafe answers and failed handoffs.
- Confirm the complete stack. Include helpdesk, AI agent, seats, integrations and support tier.
- Request billing definitions. Ask exactly when a conversation becomes a charged outcome or resolution.
- Test the knowledge path. Use current product, shipping, return and policy questions from your store.
- Stress-test the peak. Recalculate at twice the normal eligible volume and apply the stated overage.
- Assign ownership. Name who reviews conversations and updates approved source material.
For Shopify-specific capabilities, review the Shopify AI chatbot workflow. For source readiness, use the Shopify knowledge-base checklist. These checks prevent a cheap subscription from being used on content that cannot yet support reliable answers.
What else should buyers know about AI chatbot pricing?
Buyers should treat public pricing as a timestamped input, not a permanent quote. Vendors change packages, allowances and billing definitions. Save the source page and date used for approval, verify currency and tax treatment, and repeat the calculation whenever support volume, team size, channels or required integrations change materially.
How much does an AI chatbot cost for a small business?
A small business can start free, pay a flat monthly subscription, or pay by seat, conversation or resolved outcome. The useful number is the total monthly cost at your expected volume. Include the base plan, AI usage, human seats, required integrations, implementation time, review time and any peak-period overages.
What is the cheapest AI chatbot pricing model?
The cheapest model depends on real usage. A free or low flat plan is usually easiest for a narrow pilot, while outcome pricing can suit low or seasonal volume. Compare the same workload across vendors, including all required seats and add-ons, rather than treating the lowest advertised starting price as the final bill.
Is per-resolution pricing good for a small Shopify store?
Per-resolution pricing can be fair when it replaces costly work and volume is predictable. It becomes harder to budget when campaigns create a sudden rise in order, shipping and return questions. Model a normal month and a peak month, then confirm exactly what the vendor counts as a billable resolution.
Does DocMind charge per AI resolution or per support seat?
DocMind does not add per-resolution or per-seat overage charges within its published plan limits. Its pricing page lists Free, Starter, Personal and Standard plans with fixed monthly prices and stated capacities. A store should still choose the tier whose reply, message, source, bot and ticket limits match its expected workload.
How should a small business compare chatbot quotes?
Give every vendor the same assumptions: monthly conversations, expected clean AI resolutions, number of human users, channels, required integrations and peak volume. Ask for the full monthly total and the overage rule. Then compare cost per clean resolution, not messages sent, because repeat contacts and failed handoffs still consume support time.
Price your first support workflow
Compare DocMind's published limits with your real questions, then start free before choosing a paid plan.
Review DocMind pricingReviewed sources
Prices and billing descriptions were checked against official public pages on 23 September 2026. Vendor terms can change; follow the source before purchase.
- DocMind pricing — plan prices, currencies, capacities and Shopify order-tracking tier.
- Intercom pricing — full-seat prices, Fin outcome price, minimum standalone commitment and add-ons.
- Gorgias pricing — helpdesk ticket range, AI outcome model and add-on structure.
- Gorgias AI Agent pricing explained — resolved-interaction rates, allowances, handoff treatment and overages.
- Tidio pricing — Starter, Lyro, conversation and Flows allowances.
About the author
By Di - Founder of DocMind. Solo founder building DocMind from Melbourne, Australia. Di reviews DocMind's public pricing and product limits directly and separates published vendor information from illustrative cost scenarios.