How to Price a Chatbot Development Project for a Small Business Client

A small business owner emails you: "Can you build us a chatbot? How much?" You quote $500 because it feels fair for a "simple bot." Three weeks later you're still building custom logic, answering support requests at midnight, and realizing you've made about $4 an hour.

Underpricing chatbot projects is the single most common reason freelancers and small agencies quit building them. Not because the work is bad — because the pricing was never structured to survive real scope.

How much should you charge for a small business chatbot?

Most small business chatbot projects fall between $1,000 and $8,000 for a one-time build, plus $50–$300/month for hosting, maintenance, and minor updates. A basic FAQ/booking bot on a platform like Manychat or a Telegram bot framework sits at the low end. A bot that integrates with a CRM, handles payments, or does multi-step qualification logic sits at the high end. Anything quoted under $500 almost always ends up being scope creep in disguise.

The range is wide on purpose — chatbot pricing isn't about the bot, it's about what the bot has to connect to. A bot that just answers "what are your hours" is a different animal than one that books appointments, checks inventory, and pushes leads into a sales pipeline.

What actually drives the price up or down?

Before you quote anything, walk the client through these cost drivers. This is also your best tool for managing expectations before the project starts.

• Number of conversation flows — a bot with 3 flows (hours, location, contact) costs a fraction of one with 15 flows (booking, FAQ, order status, escalation, upsell).

• Integrations required — connecting to a CRM, payment processor, calendar, or inventory system each adds development and testing time, typically $300–$1,200 per integration.

• Platform choice — website widget, WhatsApp, Telegram, Instagram DM, or multi-channel all have different build and maintenance overhead.

• AI vs. rule-based logic — a scripted decision-tree bot is cheap to build. A bot using an LLM for open-ended answers needs prompt engineering, testing, and ongoing tuning — expect 30-50% more.

• Handoff to a human — if the bot needs to escalate to a live agent seamlessly, that logic (and the CRM behind it) adds real scope.

• Content and copywriting — if you're writing all the bot's responses from scratch instead of the client supplying them, add hours accordingly.

Ask about every one of these in the discovery call. Clients rarely volunteer "oh and it needs to sync with our CRM" — they assume it's implied.

Should you charge a flat fee, hourly, or a retainer?

Flat fee for the build, retainer for the maintenance. That's the model that keeps both sides happy long-term.

• Flat project fee for the initial build — clients trust this more than hourly, and it forces you to scope tightly upfront.

• Fixed monthly retainer ($75–$400/month depending on complexity) for hosting, monitoring, minor flow edits, and platform fee pass-through.

• Hourly rate only for out-of-scope requests — this is your defense against scope creep, and you should say this explicitly in the proposal.

• Avoid pure hourly billing for the build itself — small business clients get anxious watching hours rack up and will nickel-and-dime every line item.

Put the retainer in writing as a separate line item from day one. Bots aren't "set and forget" — platforms change APIs, businesses change hours, and someone has to own that.

How do you scope the project so it doesn't balloon?

Scope creep is the real profit killer, not the initial quote. A tight scope document is worth more than a clever pricing formula.

• List every conversation flow by name in the proposal — "booking flow," "FAQ flow," "order status flow" — so "add one more thing" is visibly a change order.

• Cap the number of revision rounds (2-3 is standard) before extra rounds bill hourly.

• Define what counts as a bug fix vs. a new feature in writing — this single clause prevents 90% of scope disputes.

• Get sign-off on sample conversations before you start building — a written script the client approves is your scope contract.

• Quote integrations separately — never bundle "and also connect it to our CRM" into a flat fee without itemizing it.

What should a sample pricing tier structure look like?

Giving the client three tiers instead of one number closes deals faster — people compare, they don't decide in a vacuum.

• Starter — $800–$1,500: single platform, up to 5 conversation flows, no integrations, rule-based logic, 30 days of support included.

• Growth — $2,000–$4,500: multi-flow bot, 1-2 integrations (CRM or booking calendar), basic lead capture into a pipeline, 90 days of support.

• Pro — $5,000–$8,000+: AI-assisted responses, multiple integrations, human handoff logic, analytics dashboard, ongoing monthly retainer required.

Anchor the middle tier as the "recommended" option — most clients pick it, and it's usually where your margin is best.

Where does CRMChat fit if the client wants leads, not just answers?

A lot of "chatbot" requests are really lead generation requests in disguise — the client doesn't just want an FAQ bot, they want the bot to capture a prospect and get them into a sales conversation. That's a different deliverable than a scripted Q&A widget, and it should be priced as one.

If the client's real goal is turning chatbot conversations into tracked leads and follow-up sequences, a client reporting dashboard becomes part of the value you're selling, not an afterthought. CRMChat automates the outreach and follow-up side of this — it handles the sequences, tagging, and pipeline tracking once a chatbot conversation hands off a lead, so you're not manually copying names into a spreadsheet after every conversation. For agencies running this for several clients at once, CRMChat also includes isolated client workspaces, so each account, campaign, and result set stays separate and reportable — which matters a lot when you're pricing "per client" instead of "per bot." If you're quoting a package that includes ongoing outreach on top of the bot itself, that's the difference between an $800 one-off and a $300/month retainer client.

What mistakes make small business chatbot projects unprofitable?

• Quoting off gut feeling instead of counting flows and integrations — always scope first, price second.

• Not charging for content creation — writing conversational copy for 10 flows can take longer than the actual build.

• Skipping the maintenance retainer — a bot with no owner breaks silently and reflects badly on you, even after the invoice is paid.

• Underestimating platform risk — if the client's channel (Telegram, WhatsApp, Instagram) changes its API or rate limits, someone has to fix it. Build that into your terms. See what Telegram API flood limiting means if you're building on Telegram specifically.

• Selling "AI chatbot" pricing for a rule-based bot — be honest about which one you're building; clients pay more for genuine AI logic and you shouldn't undercharge if that's what you delivered.

Price the scope, not the hours you hope it takes. Small business clients aren't trying to lowball you on purpose — they genuinely don't know what's involved, which means your proposal is doing double duty as a sales document and an education document. Get that right and the pricing conversation gets a lot easier.