AI Conversation: Getting started guide (HR)

You already know what your engagement scores are. AI Conversation tells you why. It is an anonymous, AI-led conversation that gives every employee the space to explain the context behind their feedback. The AI finds the themes, summarises them into team- and organisation-level insights, and hands managers recommended actions. Individual answers are never shared.

This guide is for HR leaders who want to run AI Conversation for the first time, get the most out of the 30-day trial, and then bring managers along.

💡 30-day free trial

Try AI Conversation with full access for 30 days, available on all plans. Everything in this guide works during the trial.

💬 Chat📨 Create and sendout📊 Survey results

Quick links

 

Why HR teams choose AI Conversation

Turn feedback into action

Understand why scores change, so leaders can take meaningful action with confidence instead of guessing. AI Conversation has a 95% happiness score among customers.

Equip managers with clear direction

AI-generated insights and recommended actions help managers focus on what matters most, without spending hours reading comments.

Build trust through anonymity

Employee feedback stays anonymous and is only shared as aggregated insights once the participation threshold is met.

Scale listening across the organisation

Run the same conversation across teams, departments, or the whole organisation, and compare what comes back.

How it works

  1. Spot a trend. You or a manager sees a shift in the temperature data and wants to understand the why.
  2. Invite the team. Employees are invited to a confidential, 10–20 minute chat with the AI as a follow-up to the pulse survey.
  3. The AI listens. It asks open follow-up questions based on each answer. Every reply stays anonymous.
  4. Insights land. When the conversation closes, managers get an anonymised synthesis with key themes, shared concerns, and positive highlights.
  5. Action follows. The AI recommends actions, so managers move from insight to action faster.

AI Conversation is built on the same science-based framework as Winningtemp's pulse surveys, so the qualitative layer and your quantitative data speak the same language.

Make your 30 days count

The trial is short, so here is a plan that gets you from activation to a decision with real results in hand.

Week 1: Activate and choose your focus

  • Activate AI Conversation and select the teams that should have access. Start with teams that already work actively with their engagement results.
  • Pick one question category that connects to a live priority. AI at Work is a strong first choice right now: most organisations are rolling out AI tools and few know how adoption is really going.
  • Tell the managers of the selected teams what is coming and why. Point them to the AI Conversation: Manager guide.

Week 2: Launch

  • Create and send the AI Conversation. You can run it yourself for the first round so managers can watch how it works before they own it.
  • Send yourself a test email first to see exactly what employees will see.
  • Give it a response window of a week or so, and extend the closing date if participation is still climbing.

Week 3: Read the results, together

  • Once the window closes and the anonymity threshold is met, generate the synthesis and the report.
  • Walk through the key themes and recommended actions with each manager. Pick one or two actions per team, not ten.
  • Share the themes back with employees so they see their input landed.

Week 4: Decide and plan the next conversation

  • Compare what you learned with what the scores alone told you. That gap is the value of AI Conversation.
  • Choose the next topic using the annual cycle below, and let the managers who saw results first become the ones who run the next round.

Pick a topic that matters right now

AI Conversation goes deep on one topic at a time. There are 25 categories to choose from, so you can match the conversation to what is happening in your organisation.

New: four trending topics

  • AI at Work. How employees are adapting to AI in their work: where it helps, what gets in the way, and whether they have the skills and support to use it well. Use it when you are rolling out AI tools or shaping an AI strategy.
  • Change Readiness (Part 1). How prepared employees feel for a change that is planned or just starting. Use it before a reorganisation, a new system, or a new process.
  • Change Impact (Part 2). How a change is actually playing out in day-to-day work. Use it once the change is in place, and pair it with Part 1 to compare expectations with reality.
  • Digital / Hybrid Work Environment. Whether tools, systems, and hybrid ways of working make work easier or add friction and overload. Use it when you hear about meeting and notification overload.

 

Not sure which category to choose?

If you want to understand Try
Growth and career opportunities Personal Development, Person-Job Fit
Why people are disengaged or leaving Commitment, Job Satisfaction, eNPS
High stress or sick leave Work Situation, Subjective Well-Being
How a manager or leadership team is perceived Leadership, Trust
Whether people dare to speak up Psychological Safety, Participation
How a change will land, or is landing Change Readiness (Part 1), Change Impact (Part 2)
How AI adoption is going AI at Work
Friction from tools, meetings or hybrid work Digital / Hybrid Work Environment
Silos between teams Cross-Functional Collaboration, Team Spirit

 

Plan your year: which category, when?

The HR year has recurring periods with different focus areas. Run an AI Conversation a few weeks before a key activity, such as the salary review or development conversations, and managers walk in with the context they need.

Period What is happening in the HR year Run an AI Conversation on
Spring (Mar–May) Motivation and engagement, salary review, performance and goal-setting conversations Engagement, Job Satisfaction, Self-Leadership, Leadership
Early summer (June) Check-in before the summer break, focus on recovery Work Situation, Subjective Well-Being
Autumn (Aug–Oct) Growth and personal development, development conversations, skills planning Personal Development, Person-Job Fit, Innovation
Winter (Nov–Feb) Budgeting and planning for next year, new goals and priorities Transparency, Trust, eNPS
Ongoing Reorganisations, new systems, AI adoption, and other changes Change Readiness, Change Impact, AI at Work

 

See every category with a short description in Question categories in AI Conversation.

 

Getting started in Winningtemp

Step 1: Activate (System administrators and HR)

  1. Go to System & templates.
  2. Click AI Conversation.
  3. Select which teams should have access to AI Conversation.
  4. Select a question category. The chosen category applies to all selected teams.

After activation, the managers and HR leaders responsible for the selected teams can create AI Conversations. Full details: Activate AI Conversation.

Step 2: Create and send (HR and managers)

Go to Employee surveys, click Create survey, and choose Get started under AI Conversation. Set the name, dates, category, and recipients, give the AI a sentence of context about the team, review, and launch. Employees are invited by email and in the app. Full details: Create an AI Conversation.

Step 3: Review results and act (HR and managers)

Go to Results & analysis and open Survey results. Once the window has closed and the threshold is met, generate the synthesis and a shareable report. Full details: View your AI Conversation results.

How to prepare your organisation

  • Explain the purpose and expected outcomes. Tell employees what you want to learn and what will happen with the results. Participation follows clarity.
  • Communicate how anonymity is protected. Say it plainly: managers never see individual answers, and nothing is shown until at least five people have taken part.
  • Prepare managers to act on the results. Agree up front that every conversation ends with at least one visible action. Share the Manager guide before launch.
  • Follow up with visible actions. Close the loop within a few weeks. The second conversation always gets better participation when the first one led somewhere.

For employees, point them to the AI Conversation: Employee guide.

From HR-led to manager-led

Most organisations start with HR running the first AI Conversation. That is the right call: you choose the topic, you see the results first, and you learn what a good follow-up looks like. The goal after the trial is for managers to run their own.

  • Show, don't tell. Share the first synthesis and report with your leadership group. Real themes from real teams are more convincing than any pitch.
  • Start with the willing. Give access to the managers who already engage with their temperature data. Their results become the internal case.
  • Tie it to the calendar. Use the annual cycle above so managers know when a conversation helps them most, such as before development conversations.
  • Keep the bar low. One topic, one team, one action. Managers who get a win in the first round come back on their own.

How anonymity is protected

  • No one sees individual responses. All input is aggregated before anything is shown.
  • Identifying details are removed. The AI is designed to strip personal, identifying, or concrete details before the text is processed.
  • There is an anonymity threshold. By default, a synthesis is only generated once at least 5 unique team members have participated. Teams below the threshold cannot be selected.

Read more about the AI model in AI-powered features in Winningtemp.

 

Frequently asked questions

Who can activate AI Conversation?

System administrators and HR, under System & templates. Once activated, the managers and HR leaders responsible for the selected teams can create conversations.

Which category should we start with?

Pick the one closest to a decision you need to make in the next quarter. If nothing stands out, AI at Work or eNPS are good first conversations: both are relevant to almost every team and give you themes you can act on immediately.

Does HR have to send it, or can managers?

Both can. Running the first round from HR is a good way to learn the flow. After that, managers with access can create their own conversations for their teams.

Can we run it on more than one team at a time?

Yes. Activate access for several teams and run the same category across them. Insights are always reported at team level, so you can see how the same topic plays out in different parts of the organisation.

How long does it take employees?

Each conversation takes around 10 minutes.

How reliable are the insights?

AI Conversation uses the same science-based framework as Winningtemp's pulse surveys, and the synthesis is built only from your employees' actual conversations.

Next steps

 

 

Was this article helpful?
2 out of 2 found this helpful
Was this article helpful?
2 out of 2 found this helpful