Chatbot ROI: a simple calculator you fill in with your own numbers
Chatbot ROI is (value gained minus monthly cost) divided by monthly cost, where value gained comes from three sources you measure yourself: enquiries you previously missed and now capture, the conversion rate on those enquiries, and support hours saved on repeat questions. There is no universal percentage — plug in your own business's numbers.
- ROI only means something with your own numbers — never trust a generic 'X% more sales' claim from a vendor.
- Track three inputs for two weeks before switching on a bot: enquiry volume, current response time, and how many enquiries go unanswered.
- Value gained has two parts — recovered enquiries that convert, and staff hours saved on repeat questions.
- A flat monthly platform fee makes the ROI formula predictable; a per-message markup means your cost line moves as you grow.
- Recalculate every quarter — your enquiry volume and conversion rate will change as the bot's flows mature.
The formula
Chatbot ROI is straightforward once you separate cost from value:
ROI (%) = ((Value gained per month − Monthly cost) ÷ Monthly cost) × 100
The monthly cost is easy, it is your subscription plus any usage costs (WhatsApp per-message charges, AI tokens, telephony minutes) at your provider's rates. The value gained is the part you need to measure honestly, from your own data, not from a benchmark someone else quotes you.
Measuring value gained: three inputs
| Input | How to measure it |
|---|---|
| Recovered enquiries | Count enquiries (calls, DMs, form fills) that went unanswered or were answered late over a two-week window before automation. |
| Conversion rate on those enquiries | Of enquiries you do respond to today, what share becomes a paying customer? Use your own CRM or sales register. |
| Average order or deal value | Your own average, do not use an industry figure, since ticket sizes vary hugely by business. |
Value gained from sales ≈ Recovered enquiries per month × your conversion rate × your average order value. Add to this any hours saved on repeat support questions, valued at what an hour of staff time actually costs your business (salary plus overhead, divided by working hours).
A worked template (fill in your own numbers)
| Line item | Your number |
|---|---|
| Enquiries missed or delayed per month (before automation) | ___ |
| Your current conversion rate on responded enquiries | ___% |
| Average order/deal value | ₹___ |
| = Estimated recovered revenue | ₹___ |
| Support hours saved per month × cost per hour | ₹___ |
| = Total value gained | ₹___ |
| Monthly subscription + usage cost | ₹___ |
| ROI = (Value gained − Cost) ÷ Cost × 100 | ___% |
Leave the recovered-revenue estimate conservative in your first pass, count only enquiries you are confident you were genuinely missing, not every enquiry that came in.
Segmenting ROI by channel
Different channels rarely convert the same way, so lumping every enquiry into one ROI number hides more than it reveals. A WhatsApp enquiry that arrives after hours, a phone call that rings out at lunchtime, and a website form filled in at midnight each have a different chance of converting and a different cost to recover.
Calculate value gained separately per channel where you can:
| Channel | Recovered enquiries/month | Your conversion rate | Value gained |
|---|---|---|---|
| ___ | ___% | ₹___ | |
| Voice calls | ___ | ___% | ₹___ |
| Website chat/forms | ___ | ___% | ₹___ |
Tagging each lead's source in your CRM pipeline makes this split possible without extra spreadsheet work, a lead card already carries the channel it arrived on, so you can filter by channel when you review conversion.
Payback period: when does the bot pay for itself
ROI as a monthly percentage is useful, but a single payback period figure, how many months until the bot has paid for its own setup, is often what actually decides whether a business proceeds. The formula:
Payback period (months) = One-time setup cost ÷ (Monthly value gained − Monthly subscription and usage cost)
Setup cost includes the hours spent building flows, writing the knowledge base and testing, value that time at what your own hour is worth, or what you would pay someone else to do it.
A quick illustrative example, with placeholder numbers only, not a benchmark for your business: suppose setup took 15 hours valued at ₹500/hour (₹7,500 total), monthly value gained works out to ₹40,000, and the monthly plan plus usage cost is ₹5,000. Net monthly value is ₹35,000, so payback period = ₹7,500 ÷ ₹35,000 ≈ 0.2 months, well under a month. Replace every figure with your own before drawing any conclusion, the point is the formula, not this specific outcome.
Doing this on Convo360
Your monthly cost line should be easy to forecast, or the ROI number becomes unreliable month to month. This is where the pricing model matters: a flat platform fee stays fixed regardless of how many conversations you handle, while a per-message markup model means your cost rises with usage, sometimes faster than your value gained, especially during a busy season. See per-message markup vs flat fee for a full comparison.
Convo360's Analytics and CRM pipeline give you the raw numbers to fill this template honestly: conversation volume, lead stages, and handoff rate from the dashboard, so 'enquiries recovered' and 'conversion rate' are not guesses after the first month. Pricing is flat and published: Starter at ₹999/month, Pro at ₹2,999/month, Business at ₹7,999/month, excluding GST, so your cost line in the formula never has hidden usage markups from the platform itself. Because it is BYOK, the only variable costs are what you pay Meta, your AI provider and your telephony provider directly, at their own published rates. The prepaid wallet with automatic GST tax invoices also makes this cost line easy to reconcile at tax time, so the denominator in your ROI formula is a real, auditable figure rather than an estimate.
Common mistakes in ROI math
- Counting every conversation as a converted sale. Use your actual conversion rate, not 100%.
- Ignoring the setup time. Building flows, training the knowledge base and testing takes hours in month one, factor that into a fair comparison, even though it will not repeat every month.
- Comparing against a benchmark instead of your own baseline. The only fair comparison is your business before automation versus your business after, over the same season.
Example conversation
Try Convo360 free for 14 days
Start free in minutes — no code, no credit card. Bring your own AI keys and pay no per-message markup.
Frequently asked questions
Is there a standard chatbot ROI percentage I should expect?
No — ROI depends entirely on your enquiry volume, conversion rate and order value, all of which vary by business. Treat any vendor-quoted percentage as marketing, and calculate your own using your baseline numbers.
How long should I measure before calculating ROI?
Measure your pre-automation baseline for at least two weeks to capture normal variation, then recalculate ROI monthly for the first quarter as your flows and knowledge base improve.
Does chatbot ROI include the cost of my own time building flows?
It is worth tracking separately, especially in month one. A no-code flow builder reduces this cost significantly compared to custom development, but initial setup still takes real hours.
Should I include WhatsApp conversation fees in the cost line?
Yes, if you are on a BYOK platform, WhatsApp per-message charges from Meta are a real monthly cost and belong in your denominator alongside the platform subscription.
What if my conversion rate is different for missed enquiries versus responded ones?
Use the lower, more conservative rate for missed enquiries since urgency and intent can differ from customers who already got a fast reply. Adjust as you gather real data after automation.
Related reading
Available across India — see WhatsApp & AI calling automation by state and city.