Main use:
Using an open-ended question, participants can freely type answers in their own words (as opposed to a pre-determined list in Choice questions), which can be in lengthy essays or short texts.
A. For inca platform users, you access inca SmartProbe Studio as you build your open-ended questions in the Question Builder thru the Project Directory (see Section A, below)
B. For inca SmartProbe API users, you can create SmartProbe instances in the SmartProbe Directory, and access SmartProbe Studio directly through any of your SmartProbe instances (See section B, below)
Section A. How to set up using the platform (inca platform users):
1. In the Project Directory, click New Project. Once you are in the Questionnaire Builder page, add a new question OR insert a new question by hovering over an existing question and clicking the green (+) button (either above or below the existing question)
2. Select Open Ended from the popup menu that appears.
3. Write your question number or label in the upper box and write your question in the Question Text Box.
4. When you want the AI to probe participant answers and prompt them for answers that could have more value/insight derived from a follow-up, enable the probing by turning on “SmartProbe” toggle in Question Settings. Wait for “SmartProbe Configuration” to load.
5. Review the simulated conversations or try it yourself. You can let inca generate more examples for you if you’d like to see more. If they look good you can just leave it and continue to build your questions.
Simulated Conversations
Try It Yourself
6. You may decide how many times you’d like inca to probe. If probing more than once, enable the toggle and specify the maximum number of rounds probing you require.
Best Practice on selecting number of probing:
- One to two turns works well for a single focused topic; three to five turns suits questions with multiple distinct themes, certain probing techniques or qualitative frameworks where depth really matters.
- If a question covers three distinct themes — like emotional benefits, utility, and sensory cues in laddering— plan for at least three to four turns to do them justice.
7. Users have the option of adding a brief with specific objectives, context or any specific content to help inca generate probing questions that reflect your brief.
Once you've added the brief, click Regenerate
8. Users also have the option of going to “inca SmartProbe Studio” if you think the SmartProbe can be further optimized.
Click “Go to SmartProbe Studio”
You will see your question information is captured in the “Your Input” section and the Advanced Settings should be toggled on
“Advanced Settings” shows how inca’s AI is reasoning and allows that reasoning to be refined through a structured template. It gives you complete transparency and control.
In this mode, your original “brief” is no longer editable. You will edit it directly in the “AI's understanding of your objectives” section. All the edits will be reflected in simulated conversations after you hit “Regenerate”
To learn more about the Advanced Settings, please go to the section inca SmartProbe Studio (Advanced Settings) by scrolling down further below.
Section B. How to set up using the inca Smartprobe Directory (API users)
1. API users need to create an inca account (through an invitation from their inca Customer Success Manager or main contact person). Once on the inca platform and logged into their account, API users will have access to the SmartProbe Directory, where they can create new SmartProbe instances by clicking + New SmartProbe.
Note: For existing API users who want to use inca SmartProbe Studio but do not yet have an inca platform account, please inform your Customer Success Manager, and we will set up your account (and, in some cases, update your MSA). Similarly, once you are logged into your account, you will have access to the inca SmartProbe Directory, where you can create new inca SmartProbe instances.
2. Input required details for your inca SmartProbe instance.
Project ID - Project ID supports multiple entries if you're reusing this instance across related projects. When project ID(s) are selected, only instances associated with those projects are shown.
Client ID - Client is a single-select field. Users can only see/edit inca SmartProbe instances associated to clients which user has access to.
SmartProbe name - Give your instance a clear name — something that makes it easy to identify in the directory. This could simply be your question label like Reason for Appeal, Barriers to Trial etc.
Language - inca SmartProbe probes in 90 languages natively. Users can select multiple languages for a multi-country study, but you'll need to provide question text in each language you select but the AI always works in English internally.
Question Wording - This is the survey question the AI will be probing on. Type in your main question here.
Tool Tip: If your study uses specific terminology in another language — local brand names, culturally specific concepts — include those terms in your brief or topic descriptions.
3. You may decide how many times you’d like inca to probe. If probing more than once, enable the toggle and specify the maximum number of rounds probing you require.
Best Practice on selecting number of probing:
- One to two turns works well for a single focused topic; three to five turns suits questions with multiple distinct themes, certain probing techniques or qualitative frameworks where depth really matters.
- If a question covers three distinct themes — like emotional benefits, utility, and sensory cues in laddering— plan for at least three to four turns to do them justice.
Important Note on Multi-turn Probing:
Multi-turn probing is fully supported on most platforms, however there may be limitations or more advanced configuration depending on the platform you are integrating with. If you have already setup single-turn integration, you may need to update the integration code. Please click the Export button to provide your programmer with more detailed information.
4. After deciding the required number of probes, click Generate
5. Users will see sample generated probes in the Simulated Conversations section. If you would like more examples, click + Example Conversation. You can also try answering the question by toggling Try It Yourself.
6. Users also have the option to brief the AI — and for most projects, we recommend it. This is where you can share your research objectives, any specific topics you want to explore or avoid, and any context that would help the AI probe more purposefully. Add your question objective and click Regenerate.
7. If the generated probes meet your expectations, you can click Export to generate the configuration for your developers. Then, copy the URL and share it with your programmers to finalize the integration in your survey platform.
8. If the generated probes are still not right with the current configuration — or if you're optimizing a complex or high-stakes question — enable Advanced Settings.
To learn more about the Advanced Settings, please go to the section inca SmartProbe Studio (Advanced Settings) right down below.
inca SmartProbe Studio (Advanced Settings)
When optimizing the probing of an open-ended question, you may select any combination of the following settings in the inca SmartProbe Studio. Optimizing probes thru these Advanced Settings works the same whether your are using the inca platform or the API.
1 AI's understanding of your objectives
This is like your moderator plays back the objectives you’ve shared with them, so take a look at AI’s take on the objectives and refine as needed.
AI may pick up cues in your brief about the tone/personality you’d like it to adopt. In most cases the brief doesn’t provide any specific instructions. Now that you see this “Persona” section, you are welcome to include instruction to ask AI to take on a particular persona
There is no longer a word limit for briefs, but users should aim to be clear and concise about their goals. A good brief should include any of the following:
1. Objectives: What they want to achieve and the topics to cover, including any rules
2. Context: Background information the AI should be aware of
3. Persona: How they want the AI to ask questions (i.e. The survey participants are Teens, therefore the conversation should be appropriate for this age segment and use friendly conversational tone)
4. Specific requirements: Clear guidance on what to do and what not to do. For example, if probes should not mention other brands, they can state: “Important: Do NOT mention other brands or competitors.”
2 Topics
In this section, the AI has generated Topics or keywords phrase or name (including people/brands/products/etc.) that are of interest to your study, which a participant might mention in their open-ended response.
Now that you see the topics, you may be reminded of other topics you’d like to specify but you didn’t think of when you wrote the brief. You can add topics.
Users can set/refine topics for the AI and confirm if they are aligned with your thinking.
Users can set also set/refine the Action for each topic:
Topic Action:
Probe always: triggers to probe on this topic always, even if other topics are mentioned.
Probe on detection: triggers to probe on this topic if the topic is detected. In other words, SmartProbe will probe for more details when a participant has mentioned this topic.
Probe on absence: triggers to probe on this topic if the topic is NOT detected. In other words, SmartProbe will acknowledge the participant’s answer but then divert the question so that it focuses on this topic.
Capture only: triggers to stop probing and move on to the next question in the survey. The topic is captured or counted as a variable that can be used for other questions.
Avoid: ensures that it will NOT generate any probe related to this topic. If a participant mentions the topic, then the probe will divert the conversation away from that topic to something else that is still relevant to the research objectives.
Terminate: triggers to end the entire survey when this topic is mentioned
Users can also use Canned Questions when the action is use any "Probe always” / “Probe on detection” / “Probe on absence".
On the new text field, user can type in their canned text or the specific probing question.
Priority
Since each request only produces one probe question, if multiple topics are triggered to probe, then SmartProbe will randomly choose one of them to focus on. This randomization behaviour is the default, but you can override by adding a priority field to the topic (a number, 1-10, with 1 indicating “highest priority” and 10 indicating “lowest priority”). Each topic must have a unique priority number. The order of the topics is not important: only the priority field indicates a preference between topics
After making changes to all of the Advanced Setting, click Regenerate. Once you are satisfied with the Simulated Conversations, make sure that it has the status "Changes Saved" and then click "Back to Question Editor" to continue editing other questions in your survey.
3 Behavior optimization
When this is enabled, users can train AI on the fly by identifying examples of good and bad probes and giving reasons why.
An example below would be a good use case on when to use Behavior Optimization.
Question:
Why do you say Guinness is different to other beer brands?
Brief:
We want to probe into how the points of difference people mention about Guinness make them feel when drinking Guinness. Focus on the emotions people feel.
After setting up the questions and brief, and without Behavior Optimization you will see the Simulated Conversations below:
In this case, you also have a specific objective to understand how the points of difference people mention about Guinness ladder to the emotions they feel when drinking Guinness.
The probes generated above does not consistently address your specific objective. Thus, using Behavior Optimization can help with this. Click "Go to Smartprobe Studio" and then enable Behavior Optimization. Once enabled, you will see the Simulated Conversations on the right side.
Simulation 1 is considered a good/expected probe to your specific objective, so you can save this as a "good" example to train the AI (click Save as Example).
Simulation 3 is considered a probe that does not exactly follow your specific objective, so you can save this as a "bad" example for the to AI avoid (click Save as Counterexample).
When adding a Counterexample, users can then add the reason why they found this probe to not useful and/or adding your ideal probe.
After making the adding all the Examples and Counter examples, click Regenerate.
Once you are satisfied with the Simulated Conversations, make sure that it has the status "Changes Saved" and then click "Back to Question Editor" to continue editing other questions in your survey.
Section A - part 2: Quality Controls (for Question Builder for inca platform users only)
The quality control system is a built-in feature on the inca platform, which automatically evaluates and may disqualify participants during surveys. A participant’s quality score is determined based on demerit points that they may accumulate, and whether they are disqualified or not depends on the disqualification control you set on your project. To learn more about the disqualification control, click on this article here.
When adding an open-ended question, you may select (click the drop down and select On) any combination of the following checks through the Quality Control tab:
Gibberish Detection
Select on/enable when you want to prompt another answer to the participant for a gibberish response (e.g. dasbjlhflsdbfsjkd)
Toogle off/disable this feature when asking for brand/person names. Please note that some brand names and person names may look like gibberish thus it's not recommended to use when answers are short brand names/person.
Uninformative Answer Detection
Select on/enable when you want to prompt for more details when the participant is deemed uninformative or generic (e.g. ok. good, don't know, yes, no etc.)
Not recommended when low-information responses (e.g. “nothing”) may be considered a valid response.
Duplicate Answer Detection
Select on/enable when you want to check for answers that match answers given in other open-ended questions that have this check enabled.
Demerit AI
Leverages SmartProbe’s demerit confidence capability to intelligently analyze the semantics of a participant’s response to a question. It goes beyond detecting gibberish answers or uninformative answers, it potentially can identify when people provide a good but off-topic response. For example, if people answer "I like going to Tesco for grocery shopping as it's close to me" to the question "What do you like about this ad?".
Questions with any of these checks enabled will have a maximum demerit point influence which can be controlled on the Quality Control tab: Strong (10 points) Normal (5 points) Weak (2 points)
Note: If any of the enabled checks are triggered, then the participant will be allocated demerit points. For open-ended questions without probing, this will equal the maximum demerit point influence. When probing is enabled, the maximum demerit point influence is only allocated if all of the responses trigger an enabled check; otherwise, each response contributes a diminishing amount of points, with the prime response having the most influence and each subsequent probe response having less.
Section B - part 2. inca Demerit AI (API users)
Demerit AI is the built-in quality control check for open end responses.
1. Toggle on "Analyze Conversational Effort with Demerit AI" to activate it. Once enabled, a demerit score between 0 and 1 will be included in your data export for each conversation — where 0 means no quality issues detected, and 1 indicates very low conversational effort.
Use Case without Demerit Instructions
Take the question: "If you were the CEO of this hotel, what's the one thing you'd do differently to better serve customers like yourself?" A one-word answer like "nothing" would likely score close to 1 — flagged as very low effort. Whereas a thoughtful, specific response like "I'd make sure premium coffee was available in the lobby every morning, served in proper cups — the kind of small touch that makes guests feel genuinely looked after" would score close to 0.
2. Users can optionally provide Demerit Instructions to give the AI additional context or criteria for evaluating responses.
Use Case with Demerit Instructions
Take the same question — but this time, it's a closing question where we're simply inviting any final thoughts. We can brief the AI accordingly: "We are just asking for final thoughts and suggestions. A response of 'no' or 'nothing' is perfectly reasonable and expected — please do not treat these as low effort." With that instruction in place, the same one-word answer of "nothing" would now score close to 0.
Voice+ Features
If the survey is generated on the inca platform, participants now have the option to answer open-ended questions using voice or video, leading to longer, more detailed answers. Then, in real-time, inca returns a probe that participants can also answer by voice or video. To learn more about this feature, please go to the Voice+ article here.
Voice+ is not available among API users. The entire survey must be programmed in inca to use the Voice+ features.
Other Useful Question Settings
Allow Copy Paste
Allow participants to copy and paste text from different windows or sources. We recommend disabling it when you don't want them to repeat their stock answers without thinking.
Allow User Corrections
By default (Yes), participants are allowed to change their answer in this question type