From the final questionnaire to feedback during your event with EVA

On the Monday after the conference, your client asks: “How did the workshops go?” You have photos, an attendance figure and some responses to the final survey. One person says there were not enough practical cases, but you do not know which workshop they mean. Another rates the whole day. You can prepare a summary from that; deciding what to change in each session is harder.

With EVA, you can start collecting that feedback while the event is taking place. You define what you need to learn and the AI assistant introduces questions in conversations when the context supports them. Answers are stored for analysis and the report.

The change has two parts: asking for an opinion close to the experience you want to learn about, and turning it into information you can use. Both are worth preparing before the event.

Decide what you want to explain afterwards

A feedback objective should relate to a decision the organiser needs to make. “Find out whether people liked it” leaves too many questions open: what they liked, who liked it and how that information will help you.

Consider an event in Madrid with a workshop on sustainable purchasing. The agency and client want to decide whether to repeat the format at the next edition. They need to understand how useful the workshop was, which content was missing and whether its length gave people time to work through the cases. Those needs can become objectives for EVA.

To prepare them, specify what you want to measure, which people the question applies to and what kind of answer you need. A rating from one to five makes responses easier to compare; a choice between options helps identify a preference. An open question helps explain the reason or reveal something you had not considered.

Decision you need to make	Question you can prepare	What the answer provides
Repeat the workshop format	How useful was it, from one to five?	A distribution of ratings, with the number of responses
Expand a topic	Which part would you like us to cover in more detail?	Themes in attendees’ explanations
Review the length	Was the workshop too short, about right or too long?	Preferences you can relate to other comments
Improve a transfer	What could we change about the return transfer?	Specific problems with an experience separate from the session

Each row needs its own context. Attending the conference does not mean someone used the transfer or took part in that workshop.

EVA asks when the conversation provides an opening

The objective guides EVA to obtain specific information within a relevant conversation. It does not require the same question to be sent to everyone at the same moment.

Clara leaves the workshop and writes to find the next session. The exchange might look like this:

Clara: I have just come out of the sustainable purchasing workshop. Is the next talk still in the main room?

EVA: Yes, it starts in the main room at 12:30. Thinking about the workshop you have just attended, how useful was it, from one to five?

Clara: Four.

EVA: What would you have liked us to cover in more detail?

Clara: Comparing suppliers. The examples were good, but we ran out of time to discuss how to choose between them.

EVA has answered Clara’s immediate question first. It then collects a rating and an explanation related to the workshop. The organiser can retain both without asking Clara to open a separate questionnaire for that exchange.

The conversation provides a basis for asking: Clara has said she took part. In other situations, the information may come from a connected source holding the relevant data. If you only know she registered, you cannot yet assume she attended.

A connection to registration or chat does not mean EVA knows where each person is or what they have just experienced. It needs enough information to ask a relevant question. If that is missing, you can confirm participation first or leave the rating for another moment.

Configure the objective with enough detail

The organiser’s work is to define what EVA needs to learn and how to collect it. You do not have to script every conversation in advance, but you do need to specify the question and the expected result.

For the workshop in our example, a working instruction could include:

Objective: understand the usefulness of the sustainable purchasing workshop and which content could be expanded.

Who to ask: people whose participation in the workshop is confirmed by the conversation or available information.

Data to collect: a rating from one to five and an explanation of any missing content.

Approach: introduce the question after resolving the attendee’s enquiry, when it fits the conversation; respect their choice not to answer.

That approach helps separate feedback collection from interrupting any conversation. If someone is trying to resolve a problem with their ticket, their immediate need is to get in.

Avoid questions that steer the answer. “Did you enjoy our excellent workshop?” combines a rating with the organiser’s judgement. Asking about usefulness and the reason leaves more room for what the person actually wants to say.

Before activating the objective, test several situations: someone who attended, someone who did not, and someone who mentions the workshop without saying they were there. Review the question each receives and how the answer is recorded.

From answers to the report your client needs

Answers collected by EVA can be brought together in a report with ratings, the distribution of selected options and themes from comments. For that report to be useful, readers need to understand where each result comes from.

For a rating from one to five, a bar chart can show the number of answers at each level. If you ask about length, another chart can separate “too short”, “about right” and “too long”. Open comments can be grouped by theme, retaining examples that explain what lies behind them.

A word cloud can give an initial impression, but “suppliers” alone does not tell you whether people liked the examples or wanted more information. In the workshop report, distinguish “more supplier comparison cases” from “good supplier examples”. The same term can appear in different opinions.

Alongside the chart, include what you asked, when answers were collected and how many people responded. If one person answered more than one question, do not present the total number of answers as though they were different attendees.

Do not calculate participation using an inappropriate denominator, either. To assess a session, you need to know which group you are considering: confirmed participants, people asked the question or the whole event. If that information is unavailable, show the response count and explain the scope you do know.

Present a decision alongside each result

The report provides more value when it helps the client decide what to keep and what to change. A chart on its own leaves that work to the reader.

For the sustainable purchasing workshop, you can combine the distribution of usefulness ratings with requests for more examples. A working conclusion might be to keep the format and review whether a dedicated supplier comparison section would help. If responses are limited or conflicting, leave the decision open and identify the missing information.

A simple structure for each part of the report is:

• What you wanted to learn and which experience it concerns.
• The answers collected, with their count and context.
• Themes or differences that deserve attention.
• The change you propose investigating for the next edition.

The agency can then explain why it recommends an adjustment. Within a company, the events lead can present the findings to management with a basis that can be reviewed.

Not every opinion needs to lead to an automatic decision. A comment about a transfer may flag an incident worth investigating; it does not by itself prove that all the logistics went badly.

The final survey still has a role

Conversational feedback can collect details about specific moments. The final survey can ask for an overall assessment, compare aspects through a common questionnaire or reach people who did not speak to EVA during the event.

Preparing them together helps avoid repeated questions and reserve the closing stage for what is still missing. You might collect workshop usefulness when it comes up and ask about the overall event afterwards.

Neither method guarantees that the whole audience will respond. People who speak to EVA may differ from those who do not. Asking close to the experience improves an answer’s context, but does not automatically make it representative.

In EVA, objectives connect what you want to learn with event conversations and subsequent analysis. For your next edition, start with a specific decision you currently struggle to justify: which session to repeat, which activity to expand or which part of the service to review. Then prepare the question, context and report that will help you make it.

Frequently asked questions

Do I have to send a survey to collect feedback with EVA?

You can collect answers within conversations through objectives defined for the event. A survey linked at the end remains useful if you need another kind of assessment or a common questionnaire for attendees.

Does EVA know someone has just left a session?

Only if it has information that establishes that. It may come from what the person says or a connected source with relevant data. Registration for a session alone does not confirm participation.

Can I combine rating scales and open comments?

Yes. Scales and options make answers easier to group; comments help explain the reason. Define in advance how each piece of data will be collected and retain its context when preparing the report.

Do the responses replace impact measurement?

They express what respondents say about their experience. To measure attendance, sales or other results, you need the relevant data. A positive rating does not by itself demonstrate a causal effect of the event.