In our brand new console dedicated to import and advanced analysis. we bring a full set of practical tools to dig and dive into the data, making the best of our unique semantic engine and simple but powerful searching, filtering and counting features. So that our users can generate insights from larger crowds and save even more time!
When magic brings frustration
When we designed FluidityApp, we were following our vision: deliver practical and meaningful collective intelligence for remote teams.
So we built a tool to enable collaborative, efficient and simple asking, and allow our users, project managers, transformation leaders, to make sense of answers and generate insights 10x faster.
How?: by providing participants with a simple engaging interface, and by empowering our customers, the organizers, with a semantic A.I. in real time, in 50 languages.
And today, running a remote workshop with FluidityApp comes with plenty of unique good surprises:
- Much more productive idea or feedback generation, thanks to anonymity and UI simplicity
- Automated clustering of answering into topics, based on semantic similarity
- Interactive dataviz giving access to insights in real time
The truth is: our initial target were project teams, of regular size, 10, 25, 50 people, likely to share feedback, ideas, suggestions. And we designed our app, its UI and its performance accordingly.
But as our customers were experiencing the magic of semantic A.I., well... they quickly spotted the opportunity. "What if I would ask 100, 1 000, 10 000 people ? Sure I will receive tons of feedback … but now I can handle it, thanks to A.I. "
As a consequence, over the last 12 months, the average size of groups working on FluidityApp has been multiplied by 7! Our organizers are now dealing regularly with dataset of several hundred of answers.
Our customers now work with crowds 7 times larger than 12 months ago
This trend surfaced three challenges:
- Adapt our server infrastructure to larger computing load to maintain the ‘real time ‘promise
- Rethink our User Interface to deliver a great analysis experience, even with large or very large datasets
- Refine our A.I. algorithms to cope with corpuses of different sizes and natures
A significant milestone in FluidityApp journey
While this is a never-ending journey to customer satisfaction, we are passing a significant milestone in releasing the Import and Advanced Analysis console.
Both the Dev and the Data Science team have made miracles to optimize the performance:
- faster response time of our A.I., migrated on a full GPU infrastructure
- optimized front and back-end performance of the web app
- Leaner import and export process
User experience is much smoother, even with dataset of several hundreds of rows… and much less dependent on whether the user has a recent PC with a decent CPU!
We learned that, contrary to workshop facilitators, organizers with large dataset to handle do not value as much "magic all-automatic" processing than having a full set of practical tools to dig and dive into the data.
A full set of practical tools to dig and dive into the data
- Richer tags: keywords or expressions, but also locations, persons, organizations
- Advanced filtering, allowing savvy combinations of keywords, topics, segment or date information
- One-click filtering: click on one tag to view all matching answers, and track the count of answers by topic
- A.I-assisted split: click one composite answer and the app suggests a relevant (adjustable) split
- Topics merger
Based on observation and interviews with our customers, we are now giving our users more control over A.I.
Of course, one can still ‘feel lucky’ and ask the app to suggest topics, but with more focus (eg limit the discovery to subtopics of one large topic) and more precision (you can guide the A.I. classification towards target topics by entering sample answers).
More importantly, we also bring more security with the long awaited Versioning feature, enabling to keep track of different stages of analysis, or different angles to it, and be able to switch from one version to another.
Designed for import
In our journey of empowering the people who need to extract insights from textual feedback, we met many people who actually ... did not need a tool to ask. They had already plenty. But they were still desperate to find help and support to accelerate their analytical work.
Some of our customers had already a tool to collect. So we built our console for import too.
These people had often one thing in common: one large ugly excel spreadsheet with survey results! Extracted from an online survey, from a satisfaction questionnaire, from a Miro or Klaxoon sessions, or just interviews notes.
We made it ultra simple to import these ugly spreadsheet into our smooth nice-looking console. But we felt it was important to keep as much precious information as possible when importing. So our console allows importing results:
- from plain open questions
- also from standard ratings (How satisfied are you, why?)
- from NPS scores (how much would you recommend and why?)
- from multiple choice questions (In Favor/Against or A/B/C/D and the reason why)
So now with FluidityApp, our users are covered.
For those who need interactive brainstorming powered by real-time semantic magic, our legacy app fits them perfectly.
For those who need more depth of analysis, more control for larger dataset, collected via FluidityApp or any other survey/facilitation tool , there is now the Import and Advanced Analysis console
Reading time: 3 min