Skip to main content

[Closed Beta] How to Use the Scanner in Studio

  • Updated

  1. Context

  2. Step-by-step guide to running and analyzing the Scanner Diagnosis

  3. Details by specific flow (Repeatable standard)

  4. Tips and executive reading of priorities (FAQ)

 

Context

The Scanner is a feature that analyzes existing customer service flows on the Blip Platform, identifying deviations, abandonment risks, and automation opportunities. It generates a detailed report with general metrics, comparisons between flows, and targeted recommendations to enhance the efficiency of intelligent contacts without compromising the customer experience.

Before you start

To use the Scanner, you must meet the following requirements:

  • Have customer service flows already structured and operational on the Blip platform to generate data.

  • Have administrator access to Studio.

 

Step-by-step guide to running and analyzing the Scanner Diagnosis

1. Start the diagnosis

  1. On the entry screen (landing page), you will see the main message "Discover how to increase efficiency with one click."

  2. Click the Start diagnosis button.

  3. Wait and monitor the diagnosis execution status.

  4. After the diagnosis has been fully processed, you will be able to view your Scanner.

2. Monitor processing execution

When you click the action button, the Running diagnosis of your flows modal will be displayed. Processing takes place sequentially in 5 stages and may take some time:

  1. Loading interaction logs

  2. Mapping message volume

  3. Detecting conversation loops and deviations...

  4. Calculating friction and unexpected inputs…

  5. Systematizing recommendations and priorities...

  • If you need to interrupt the diagnosis before completion, click the Cancel button.

  • The Go to report button only becomes available and active after all 5 stages have been completed.

3. Analyze the overall result (Conversational health metrics)

After the diagnosis is processed, the system will display the Your diagnosis is ready! screen.


1. Check the report's time window. The diagnosis uses a fixed analysis window (for example, the last 15 days of operation).

2. Evaluate the three consolidated indicator cards for Conversational health metrics:

  • Interaction volume: total accumulated during the period (e.g., 10.3M).

  • Unexpected inputs: percentage of interactions outside the expected flow (e.g., 27.2%).

  • Completion: percentage of users who reached the final goal (e.g., 2.4%).

4. Evaluate volume and quality composition by flow

Scroll down to the general comparison charts:


  1. Flows with the highest volume: consult the list sorted in descending order to prioritize flows with the greatest operational impact.

  2. Interaction quality composition by flow: examine the stacked bar chart to identify user behavior in the flows.

5. View the overall distribution of interest by product

1. Analyze the relative volume of interactions among the various topics and subproducts.

This view consolidates users' actual demand by product/topic, regardless of which flow they started. It serves as a strategic map for identifying priorities for new automations and improvements.

 

Details by specific flow (Repeatable standard)

To conduct a deeper analysis, you can access the detailed view of any selected flow. The system applies a standardized structure consisting of 3 views in sequence:

View 1: Flow health metrics and goal proximity analysis

  • Interaction volume: number of unique accesses to the flow.

  • Abandonment: percentage of users who stopped navigating before completion.

  • Completion: percentage of users who successfully reached the final goal.


The Goal Proximity Analysis chart distributes volume according to the user's level of progress in the conversation:

  • Low

  • Medium

  • High

  • Completed

This first view allows you to directly measure journey efficiency, highlighting the exact conversion percentage versus the dropout rate.
 

View 2: Message profile and journey stages

1. Consult the User message profile and format chart to check the types of media sent:

  • Numbers only

  • Short text

  • Medium-length text

  • Image

  • Audio

2. Examine the complementary Goal Proximity Analysis chart, which reorganizes the journey by functional stages:

  • Identification

  • Execution

  • Completion

  • Contextualization

Classifying messages by profile helps you understand how the user interacts (whether they use more audio, text, or send documents/images), while the breakdown by stages reveals the precise phase where the user encounters bottlenecks and abandons the service.

View 3: Distribution of interest by product and automatic insights

1. Observe the Distribution of interest by product chart within the selected flow. In the Registration flow, for example, subproducts such as Change of account holder, Consumer Unit (CU), CU inquiry, Email registration, Contract termination, Patient registration, Recover account, Change registration, Energy account, and Self-reading will be listed.

2. Analyze the Automatic Insights panel, which always consists of three structured pillars:

  • What we identified: descriptive diagnosis of the problem observed in the data.

  • Why it matters: quantified business impact (e.g., a severe abandonment rate of 28%).

  • What we recommend: action plan and technical guidance (e.g., create an integrated financial conversational agent).

Automatic insights are synthesized from interaction data and serve as immediate input for outlining opportunities to improve intelligent contacts. They can even be integrated with Brain to automate guidelines.
 

Tips and executive reading of priorities (FAQ)

Based on the Scanner processing, the consolidated data generates clear action priorities for the operation:

  • Priority 1 — Customer Service (Focus on Bot Understanding): If the diagnosis shows a high rate of unexpected inputs (e.g., 57.7%) and low completion (e.g., 0.5%), it means that the AI/NLU is not understanding customer inputs. Action: redefine and intensify model curation to reduce unexpected inputs to acceptable levels.

  • Priority 2 — Registration (Focus on Conversion): When a flow has low bot friction/failure (e.g., 18%) but extremely low final conversion (e.g., 3.3%), the problem lies in the length of the journey. Action: restructure and simplify the requested steps or fields to increase the completion rate.

  • Priority 3 — Financial (Focus on Transaction Retention): High abandonment (e.g., 38.4%) in critical revenue or billing journeys requires immediate intervention. Action: adjust the conversational design and simplify the delivery of bills/financial data to increase retention.

  • Priority 4 — Technical Services (Focus on UX and Operational Bottlenecks): When the flow has high initial user intent (e.g., 92.6% adherence) but falls apart at the end (e.g., 2.3% completion), the reason is usually technical. Action: conduct a qualitative investigation into integration instability, lengthy validations, or failures in external APIs.



Need more help? Explore our content at Blip Academy or Blip Community, watch tutorials on our YouTube channel or get your questions answered through our support channel 😃

Related articles