Conversational Mode turns your product finder into a natural, dynamic conversation. You control the matching and conversation structure, while Conversational Mode delivers a personalized, more human experience. In short: you keep control over the advice logic, while Conversational Mode makes the advice faster to deploy, easier to maintain and future-proof.
💡 Conversational Mode is not available by default. Contact [email protected] to explore your options.
Why Conversational Mode?
Conversational advice is becoming increasingly common through tools like ChatGPT, Gemini, and service chatbots. With LLMs, it is potentially possible to provide better and more personalized advice. However, there is a risk: LLMs are unpredictable and difficult to control. For example, they do not take your commercial preferences into account and often fail to provide reliable advice, because the responsibility for asking the right questions is left to the customer.
With Conversational Mode, we take a different approach.
The product knowledge - and with that, the advice logic - remains completely under your control. The LLM is only used to make the conversation dynamic and smooth. This way, you get the best of both worlds: a natural conversation and consistently reliable advice.
How does Conversational Mode work?
Curious to see this in action? Try out the demo Conversational Headphone Expert here.
Conversational Mode is not a chatbot. Unlike a standard chat experience where users type in their own responses, in Conversational Mode customers select their answers by clicking. This is a deliberate choice: clicking is faster than typing, keeps the advice reliable, and ensures a consistent, predictable experience.
Conversational Mode guides the shopper step by step, introduces each question in a logical way, and confirms the customer's answers. After that, they receive the same advice they would get through a classic product finder.
How does Conversational Mode determine the content of the conversation?
Conversational Mode follows the advice logic you set up in Aiden. The question order, answer options, and matching are applied exactly the way you configured them.
The texts in the product finder - your questions and information texts - are used by the LLM to generate a logical, dynamic conversation.
In short: you focus on the logic; Conversational Mode ensures a smooth conversation.
Conversational Mode or Scripted Mode?
Alongside Conversational Mode, our classic product finder (Scripted Mode) continues to exist. Which interaction is best for your product finder depends, among other things, on your target audience and product category.
Conversational Mode offers a personal, natural experience, guiding the user step by step through the conversation. This can work particularly well when nuances or preferences are an important part of the advice - for example, when choosing headphones. On the other hand, Conversational Mode provides less space for images or structured information. If your product finder relies heavily on visual or detailed (technical) explanations, Scripted Mode might be a better fit.
We have tested Conversational Mode and Scripted Mode side by side. The results show that Conversational Mode and Scripted Mode have comparable completion rates, click-through rates, and conversion rates. However, more users answered the first question in Conversational Mode, suggesting that shoppers are more inclined to start the interaction with the chat advisor.
Ultimately, it’s all about results. Conversational Mode is easy to set up, so be sure to (A/B) test it to find out what works best for your product finder!
Good to know
The LLM that powers Conversational Mode uses your existing product finder as its foundation; it’s therefore important that this foundation is set up correctly. Before deploying Conversational Mode, pay extra attention to the following:
Order of questions
Conversational Mode follows the exact order of your questions, so make sure this order is logical. Our best practice: start with easy questions to warm up the customer and prevent drop-off, then move on to slightly more complex or technical questions, and finish with practical questions such as color preferences or budget.
💡 Curious about more best practices for setting up your product finder? Click here!
Texts in your product finder
Conversational Mode uses an LLM to phrase your questions and information texts so they fit naturally into the conversation. This means you only need to think about which information to include, rather than the exact wording. This makes setting up a product finder much faster!
Note: Do not use texts that refer to specific Scripted Mode interactions, such as “Click on the previous question.” These interactions do not exist in Conversational Mode and can lead to confusing instructions for customers.
Use of infotexts
Remember that the infotexts in your product finder are used to provide extra explanation in Conversational Mode. The more details you include in these texts, the more helpful your advisor becomes.
Infotexts for answers vs questions
Answer-level infotexts do not appear directly in the Conversational advisor’s question. In Conversational Mode, customers only see the answer title; the LLM uses the infotext in the following text bubble to confirm the customer’s choice.
Question-level info texts, on the other hand, are used to explain the purpose or reasoning behind the question and are displayed with the question.
If your product finder has a lot of answer-level info texts, it’s a good idea to test Conversational Mode carefully in Preview. If you think customers will lack context to answer some of the questions, you can place the relevant information in the infotext of the question instead of the answer.
Infopages in Conversational Mode
If you want to share important information that the customer must see, use an infopage. The content of infopages is not altered by Conversational Mode and is displayed as a static information block to the user.
Start with Conversational Mode
Ready to take your product finder to the next level and make it feel like a real conversation? We got you! There are three steps to follow.
1 - App settings
Configure your AI advisor in Personality.
Avatar
Avatar image: Upload an icon or image to make the avatar more personalized.
Name: The name of your AI advisor. Keep it simple; just a first name is OK.
Role description: The text that appears under the name. Note: according to the EU AI Act, you must indicate that this is AI.
Tone of voice
Customize the tone of voice of Conversational Mode to make sure it aligns with your desired communication style.
Change the setting to 'Use custom tone of voice instructions' to be able to change the instructions and give examples.
Edit the suggested instructions or add your own instructions describing how AI Mode should communicate. Keep each instruction focused on one aspect (e.g., formality, sentence length, personality). This can be done in English or the app language.
Optional: Paste up to 3 example texts that capture your brand voice. Make sure to use the language of the app for the example text.
Save your settings.
Try Conversational Mode in preview to check if the tone of voice is correct. If not, try making the instructions more specific or add an example.
Don't forget to publish your app!
Note: These instructions cannot be used to change the content of the responses, they only influence the communication style.
Start message
You configure your Start message - the first message that appears when the product finder is opened - in Conversation in the 🏁 Start block. If left blank, the LLM will automatically generate a message like: “Hi, I’m happy to help you find ….” Providing a start message is optional.
Below, you can configure the Start message for product check: the first message that appears when people do a Product Check on the PDP. It’s smart to reference the specific product here, e.g.: “Hi! Let’s see if {{product}} is the best choice for you.”
2 - Look & Feel
In your existing Look & Feel templates, there is now an extra tab: Conversational Mode. Here you can adjust the appearance of your advisor, such as font and bubble colors.
💡 Happy with your Conversational Mode settings? Testing would be your next step. Go to Preview → Conversational Mode and thoroughly test your advice by having several conversations with your advisor.
3 - Embed
Embedding your product finder in Conversational Mode works the same way as in Scripted Mode: go to your finder → Embed → Yes, I’ve set up the script. You can now choose between Conversational Mode and Scripted Mode. Click Conversational Mode, select Sidebar app, and the embed code will be ready for you.
Here's an example of a code:
<!-- Do not use this example code, copy the code from the Embed page in Aiden -->
<button
data-advisor-id="00000000-0000-0000-0000-000000000000"
data-aiden-mode="ai-sidebar"
style="
border: none;
background: #F5C402;
border-radius: 0.25rem;
padding: 0.75rem 1.25rem;
color: #000000;
font-size: inherit;
font-weight: bold;
cursor: pointer;
">
Start Product Finder
</button>
Completed all steps? You're ready to launch your first product finder in Conversational Mode! 🎉