All case studies

Do quick answers help turn website visits into enquiries?

Answer common questions on the page and give customers a clear route to someone who can help with the rest.

Read the key findings

TL;DR

  • Answer repeated questions where customers choose. A railway’s agency reported 9% more dining bookings after it added an FAQ drawn from real customer calls.

    See the FAQ booking result
  • Set a response target for new enquiries. In a study of six organisations, leads called sooner were more likely to be reached or qualified. Record how quickly your own team sends a useful reply.

    See what the lead study actually measured
  • How a bot handles uncertainty matters. In a 2025 chatbot experiment, button-based interaction scored better than free text on user-rated outcomes; an apologetic error message modestly improved satisfaction.

    See the chatbot experiment
Average satisfaction rating after a chatbot error, on a 7-point scale
  • Apology plus an explanation5.18 / 7
  • Neutral response4.90 / 7

Klein, Papies and Utz measured how participants rated two ways of handling a chatbot error. The comparison is about satisfaction in the experiment.

See how it was measured

Reduce uncertainty at the moment it matters

A visitor may need to know whether you cover their suburb, what happens after a quote request, or whether a date is available. Clear on-page answers can remove avoidable uncertainty; a timely human reply can help with a question the page cannot resolve.

These studies cover three moments in that process. A booking page can answer questions before a customer contacts you, a person can follow up quickly once an enquiry arrives, and a chatbot can make it easier to recover when an automatic answer misses the question.

Agency-reported booking page test

Bookings rose when common questions were answered on the page

Conversion Rate Experts reported 9% more dining bookings after adding answers to common questions on a railway’s dining listing page.

How the team turned repeated calls into answers on the page
  1. Review the questions customers call about
  2. Write answers with the support team
  3. Place answers on the page people use to choose
  4. Measure completed bookings

The questions came from call logs. The agency wrote answers in the same language the support team used on the phone and put them where customers compared the available experiences. The reported result covers this page and booking journey.

Review the enquiries you already receive and pick one repeated question a customer needs to answer before choosing your service. Add a clear answer where they make that decision, then track completed enquiries or bookings.

Conversion Rate Experts, Churnet Valley Railway case study. Read the dining FAQ test

Retrospective lead-response records, six organisations

In one older lead dataset, waiting was linked to lower odds of contact and qualification

The study reported that odds of qualifying a lead were 21 times higher when the call came within five minutes than when it came at 30 minutes. It also reported 100 times higher odds of contact at five minutes than at 30 minutes.

Relative odds reported in the study at five minutes compared with 30 minutes
  • Lead qualification21×
  • Lead contact100×

These are relative odds in a retrospective study of six organisations’ telephone follow-up records, presented in 2007. How they staffed follow-up and qualified leads could have influenced the comparison.

Record when requests arrive, when they receive a useful reply and how many lead to a worthwhile conversation. Use those measurements to set a response target your team can meet.

Dave Elkington (InsideSales.com) and James Oldroyd (MIT Sloan), Lead Response Management Study, presented 16 October 2007; summary PDF dated 17 July 2008. Read the Lead Response Management Study

Online chatbot experiment, 416 participants

Give customers quick, clear options and a route to a person

Participants rated button-based chatbot interaction better than free-text interaction on five outcomes. Satisfaction averaged 5.18 out of 7 for an apologetic error response that disclosed the bot was still learning, compared with 4.90 for a neutral response.

Average satisfaction rating after a chatbot error, on a 7-point scale
  • Apology plus an explanation5.18 / 7
  • Neutral response4.90 / 7

The researchers measured perceptions such as competence, satisfaction and intention to use the chatbot again. Give customers a way to get help when the automated answer misses their question.

Start with the questions customers repeatedly ask: where you work, what services you provide, what information a quote needs, and what happens next. Put reliable answers on the relevant page. If you add chat, offer simple topic choices, say clearly when a person will respond, and provide a way to leave a message or book. Review unanswered questions and handoffs so the tool helps the customer reach an outcome.

Stefanie H. Klein, Dominik Papies and Sonja Utz, “How Interaction Mechanism and Error Responses Influence Users’ Responses to Customer Service Chatbots,” International Journal of Human–Computer Interaction, 41(7), 2025. Read the chatbot interaction study

Make the next answer easy to find and act on.

A website can answer common questions immediately; a person can handle the exceptions. Track whether visitors get a useful answer and complete the next step, rather than assuming a chat bubble is the result.

  1. List the questions that stall enquiries, such as service area, price basis, availability and what to prepare.
  2. Answer the common questions on the service or booking page. Give any chat or form a clear response-time expectation and human handoff.
  3. Measure time to first human reply, qualified enquiries and completed bookings for a defined period, then improve the point customers still get stuck.

Start with the question that most often delays an enquiry.

Explore website conversation agents