Why we built Org360
Org360 started with a 360 of my own. Here's what we learned from it, and the 6 decisions that shaped the tool.
Where it started
My first 360 was at a leadership retreat, back when I ran an organization. It had about 70 questions and felt long. It was also hard to read, and it changed how I worked: I built a management team around me and learned to delegate. A coach then helped me see where I belonged next, and that led to Nonprofit Builder.
A few years later my coach ran another 360 for me. It had 5 questions, all open text. My colleagues were frank, and I got practically the same depth as from the 70-question version. Short surveys work when people feel free to be honest.
Out of curiosity I gave those raw comments to Claude. It organized them well and picked out 3 clear things for me to work on. That was the moment I saw what AI could do for coaches: sort a pile of feedback in minutes, and then write a report in whatever format and voice you ask for.
The second idea came from the tools coaches had to choose from. The well-researched 360s are locked behind certification courses, and you can't change their questions. The free survey tools let you change anything, but they don't handle rater groups, anonymity or scoring. Coaches who had their own method were stuck in spreadsheets. We built Org360 for them.
1The coach decides the method
Our surveys are starting points. Coaches change the questions, add their own and save versions for different clients, and the report follows whatever structure they use. We look at the changes people make and use the best ones to improve the starter forms, so the library gets better as more coaches use it.
2The person who asked for feedback owns it
Whoever launches a 360 controls the report. When a coach runs it for a client, the coach and client decide who sees it.
We built it this way because people soften their answers when a report travels up the line to HR or a manager. They give the honest version over coffee and the polite version in the form. A report that goes to one person, who asked for it, gets the honest version.
Read more: Who should own 360 feedback?3Raters need to feel safe
Comments are where raters are easiest to recognize, so the coach chooses how much protection each survey gets. With No confidentiality, names show and the report opens after 1 reply. In Standard, it opens after 3 replies and comments are shuffled and shown without names. In Highest, it opens after 3 replies and the comments are replaced by an AI summary. In every mode, names are removed before any AI reads the text, and raters can change or delete their answers until the survey closes.
Read more: Is 360 feedback anonymous?4The questions are open
Our questions are published under a Creative Commons licence (CC BY 4.0). Anyone can read them, change them and use them, in Org360 or in Google Forms. They draw on published research in leadership and organizational behaviour.
We'd like Org360 to be a place where coaches and HR people improve feedback tools together. You can browse every question in our 360 question bank and our engagement questions.
Read more: Why are Org360's surveys open source?5AI reads, the coach judges
The AI does the slow work: combining scores and comments by topic, flagging an outlier view an average would bury, and drafting a summary with 3 priorities. Every score stays in the report. The coach decides what it means and what the client does next.
The client report itself is built in the coach's own AI, through our connector. That's where the coach's voice, materials and method already live, so the report can sound like them. Org360 doesn't keep a copy of anyone's report style.
Our own summaries use Claude. Names of people and organizations are removed before any text reaches it, and survey data isn't used to train AI models. Data is hosted in Germany.
Read more: How does Org360 use AI?6Short surveys by default
My 2 360s taught me that length adds little once people are frank. The research says the same.
Drop-off rises once a survey passes 7 to 8 minutes. And answers late in a long survey get faster, shorter and more uniform, open text included (Galesic and Bosnjak, 2009). At 12 questions, a request from a colleague is easy to say yes to.
So our ready-made surveys have 12 or 24 questions: 3 per area, which is enough to score an area reliably, plus an eNPS question and 2 open-text questions while raters are still fresh. They take 5 to 10 minutes. If your method needs more, build a longer one; we'd just suggest keeping only what you'll use.
Read more: How long should a survey be?Trust is the thread through all 6 decisions. Short surveys, a report that goes to one person, and protection for every comment: together they make raters willing to be honest, and honest answers are the whole point of a 360.
Daniele D'Esposito, founder
Try it on our sample 360 first
Your first real survey is free.