MEASUREMENT

Measure chatbot lead generation beyond the chat count.

Use a small, clearly defined funnel to understand quality, cost and follow-through.

By the uchat.bot team · 3 min read

Define each stage before building a dashboard.

A chat open, a message and a qualified enquiry are different events. Agree on definitions with the team before reporting a conversion rate. A practical sequence is eligible visits, started conversations, completed enquiries, sales-accepted leads and completed next steps.

Define whether a “started conversation” requires a visitor message rather than an automatically opened widget. Decide how repeated sessions and existing contacts are counted. Consistent definitions make a modest dataset more useful than an impressive chart with ambiguous totals.

Keep the denominator visible.

A rate needs a clear numerator and denominator from the same population and time period. Conversation-to-enquiry rate is completed enquiries divided by started conversations. Acceptance rate is sales-accepted leads divided by the enquiries reviewed under your agreed process.

For illustration, 100 started conversations and 25 completed enquiries give a 25% conversation-to-enquiry rate. If 10 of those 25 are accepted, acceptance is 40%. These are example calculations, not industry benchmarks or results from uchat.bot customers.

Separate volume, quality and cost.

Track the cost per accepted lead alongside the number of accepted leads. Define which costs are included, such as advertising spend, and state exclusions. Do not present an advertising-only calculation as the total cost of running the programme.

Add practical quality signals: missing details, duplicates, incorrect routing and time to the first human response. These reveal whether an apparent increase in leads is helping the business or simply creating more sorting work for the team.

Compare like with like.

Keep channel, page intent and campaign context visible. A visitor arriving from an educational article is not equivalent to someone responding to a quote campaign. Compare similar groups, and note changes in traffic quality, spend or staffing that might explain the result.

Avoid declaring a winner after a handful of conversations. Choose an evaluation period and success measure before changing the flow. Review qualitative examples alongside the numbers, especially when a small number of unusual enquiries can shift a rate significantly.

Turn the report into one improvement.

If starts are healthy but completions are weak, inspect questions where visitors leave. If completions are high but acceptance is low, review the offer and qualification criteria. If accepted leads are not contacted, fix ownership and follow-up before buying more traffic.

A useful weekly review ends with a specific change, an owner and a date to check the effect. Keep the original definition of success visible. The purpose of measurement is to make customer conversations and team follow-through better, not to collect the most flattering number.