Part 2 · Article 7 of 14
How often should you run a 360 or engagement survey?
In short
- Run a survey only as often as you can act on the results.
- 360 feedback: a baseline at the start, a second round around month 6 to see what changed, then once a year. A short pulse or eNPS check can fill the gaps.
- Engagement surveys: every few months. Fix 3 priorities, then survey again.
- Client surveys: keep them running all the time and watch the trend.
Org360 defines a pulse survey as a very short survey, often just an eNPS question and 1 or 2 open-text questions, run between full surveys to track how things are moving.
Why does frequency depend on action?
A survey creates an expectation that something will happen. When nothing does, people stop answering, or stop answering honestly.
Culture Amp's standard engagement template asks employees whether they believe action will follow the survey and whether they've seen positive changes from the last one. Those questions exist because belief in action drives participation.
So the right rhythm is the one that leaves enough time to act. Each round, pick 3 priorities, work on them, and survey again once people have had time to notice the change.
How often should you run a 360?
- At the start: a baseline, ideally when someone takes on a new role or starts work with a coach.
- Around month 6: a second round with the same questions, to see whether the 3 priorities moved.
- Then once a year, with a short pulse or eNPS check in between if you want a running read.
A 360 more often than every 6 months rarely shows real change, because behavior takes time to shift and colleagues take time to notice.
How often should you run an engagement survey?
Every few months, for example once a quarter. Each round, the leadership team picks 3 priorities, acts on them, and tells the team what it did. The next survey shows whether those changes landed and surfaces the next 3.
A short survey is what makes that rhythm possible. 12 questions every quarter is easier for a team than 57 questions once a year, and it gives you 4 readings instead of 1.
How often should you run a client survey?
Continuously. Client surveys work best as an always-on survey: send it after each project, delivery or service visit, and read the trend. Org360's ongoing reports show the NPS trend over time, the volume of responses by month, and an AI summary you can refresh as new answers come in.
Should 360s and engagement surveys run at the same time?
Better not. Stagger them by a few weeks, so nobody faces a pile of surveys in the same month. Article 1 explains why survey load matters.
What the research says
- What you do with feedback matters most. Walker and Smither (1999) followed more than 250 managers who received upward feedback every year for 5 years. Managers who discussed their feedback with their team improved more, and they improved more in the years they held those discussions. The study's title sums it up: "what managers do with their results matters."
- Expect steady, modest change. Smither, London and Reilly's (2005) meta-analysis of 24 longitudinal studies found that improvement after 360 feedback is generally small. Improvement was most likely when people saw a need to change, believed change was possible, set goals and took action. That's the case for repeating the cycle instead of expecting one survey to transform anyone.
Key readings: Walker and Smither (1999); Smither, London and Reilly (2005). Full references are in the research reading list.
Org360's view
Org360 recommends short, regular surveys over long, rare ones. Each round should end in 3 priorities that people can see being worked on. That's how a survey becomes a habit people trust, and how leaders always know where they stand.
Frequently asked questions
Is one engagement survey a year enough?
It gives you 1 reading a year, and a lot can change in 12 months. Org360 recommends every few months with a short survey.
How soon can I repeat a 360?
Around 6 months after the first one. Sooner than that, there's rarely enough change to measure.
What is a pulse survey?
A very short survey, often an eNPS question plus 1 or 2 open-text questions, run between full surveys.
What if we can't act on the results yet?
Wait to run the survey until you can. Asking without acting does more harm than not asking.