🧵 threads.net/@morsla

I hope you're well, amongst all the chaos. I mostly use this 𝕏 account to talk about #swarmconf & community management. Conference season is approaching, but I'm planning to post Swarm content over on 🧵 Threads this year - you can find me at the same username over there.

Then, you can compare CDI by location/language/tenure/type/participation level; identify areas of weakness, and adapt your tactics to address these. Good community metrics are the ones that help you to do your job better, diagnosing problems. #swarmconf

How can you include engagement in this? Take the community-driven impact score (-5 to +5) and multiply it by the number of unique visitors. That gives a "total" community impact score you can track over time and benchmark against. #swarmconf

Needs to be more specific than the generic NPS/CSAT "would you recommend this product or brand" - people correlate their community experience with the product experience. For our purposes, we need something specific to the impact the community has had on members. #swarmconf

Ask: "How has the community impacted your likelihood to do X," on a sliding scale. Simple, looks for causation, easy to survey members for. That gives you an impact score that can sit at the heart of the community strategy. #swarmconf

Set up the right questions (look at brand perception examples, etc) - then re-survey a year later to see the impact during the time you could influence those metrics. Compare between different cohorts, and see what changes over time. #swarmconf

If you want to prove the impact of a community, you MUST collect attitude data when members join. Member satisfaction, brand attitude/perception/preference, NPS - whatever the measure, you need it from the point of joining. #swarmconf

Rich Millington - 'You should be spending 20% of your budget on quantifying the community's value' #swarmconf

We don't set aside anywhere near enough time and resources to analysing and explaining the value of our community. We're all "too busy." But if you DON'T do that, you're community will remain that busy and starved of resources, forever. #swarmconf

Blunt correlation tools aren't a good fit for measuring complexity like "how valuable is this part of a community." But we keep using them, because we don't have a ready alternative, and people want to see some kind of handy numbers on a graph to infer results/value. #swarmconf

This kind of community data is still relevant, but it's better for helping to set priorities: relative to all the other things you could be doing, which one is important to do first? #swarmconf

Where do call deflection ROI models break down? Figuring out how many searches were "successful" at connecting customers to answers. How is that defined? Are you just hand-waving something that will become a key part of downstream calculations? #swarmconf

Community ROI metrics are, to put it very politely, rather suspect. The methodology used for them is often pretty rudimentary, and they don't really hold up to the kind of scrutiny given to ROI for other channels. #swarmconf

Last session for the day at #swarmconf, and it's going to be another big one: Unpacking the new community-driven impact model from Feverbee, with the excellent @RichMillington.

@evanhamilton It's Swarm week, which is always an excellent part of my year - though I'm really missing the in-person connections that usually make up half the value of attending an event. Less serendipity and stumbling into great conversations you weren't looking for.

I was out at camp for a week and presumably the algorithm has deprioritized me. Please help me by liking this picture of a cat and replying to tell me how your week was.

@evanhamilton *waves across the Pacific Ocean* Hope you've had a good break!

"What should I measure for my community" is usually the wrong question to start with - your metrics should come from your goals. Then, work out the best way to track those in a way that helps monitor success (or failure) as you implement change. #swarmconf

Also on that, Microsoft was used as an example: curating their "featured member" spots not for the uber-users with 50k+ posts, but for upcoming early-tenure superstars. They'll benefit more from that exposure, and they're far closer to the majority of other users. #swarmconf

Great point on recognition: "Communities should be really good at recognising contributions, and the people making them." Often they aren't... so find examples where this is being done well, and look at the impact it has. #swarmconf

Another interesting point from Ness on innovation under constraints - in Australia at present, community teams very rarely have much budget. That can lead people to really hone their skills at getting (and showing) results while running VERY lean. #swarmconf

Interesting thread running through a few of the #swarmconf presentations - successful community teams generally include some degree of investigative analytics, as no software platform is really going show you the value (to your specific organisation) generated by a community.