Open Stack

All,

I'm sorry about the delay in getting this out. This is a follow-up to the 
discussion at the personas meeting last Wednesday. To be honest, I am 
swamped and writing this down forced me to think through a lot of details.

Earlier this year, I proposed using companies and later ecosystems (which 
can contain multiple companies) to provide context to our personas. I 
proposed this because the goals of the personas are going to dependent on 
their job responsibilities. A cloud administrator at a large bank running 
a production environment is going to do different things than a cloud 
administrator at a cloud service provider or a small test and development 
cloud or a research cloud. In an enterprise, much of what users do is 
determined by their job, IT policies and management, and in a research 
cloud, their policies will likely be much less formal.

So does this mean we need to interview enough customers to create 25+ 
personas? No.

The goal is to make our assumptions clear about the kinds of companies 
that are out there and have a context to place our interviewees. We can 
fill in the ones that we already have data for and use them as a way to 
focus recruiting, but we would only explicitly recruit for the ones that 
either we as a group or our companies think are important. And for data we 
could use what we have, what we have from our own proprietary work and 
target new interviews with our own or new customers. We would then need to 
come together with other teams who share the same users and create 
personas that make sense (for example, IBM, HP & Red Hat all probably have 
many customers that look like CNBB Securities and we could work on those 
personas together). For the proprietary data, we would each be responsible 
for scrubbing it of anything confidential or that should not be otherwise 
made public. 

What we have as personas and ecosystems are not set in stone. I came up 
with a set of these derived from the personas and what I know about the 
industry and we've expanded these based on discussions with the other team 
members. We can add to or combine these as needed and they will change 
over time (for example, we would hope that CNBB Securities will be using 
OpenStack broadly in a year or two).

I think everyone was OK with this general approach.

So I think the next steps would be:

Short term

1. Fill in the persona, company and ecosystem placeholders with whatever 
you know based on your interviews-even one datapoint is something that we 
should share-and I think as long as we are honest about the datasources, 
then I think that is OK. 

2. Look at the ecosystems I created and the roles. Are we missing any key 
ones that you are aware of? Add them and make a note by clicking the Make 
a comment link.

3. Choose ecosystems where you have some need or expertise: Large Cloud 
Service Provider sounds like Rackspace and Aminah has agreed to help flesh 
out TBD Company (Enterprise using OpenStack). I suspect Red Hat, HP and 
IBM all have customers like CNBB securities, and probably the Medium-Sized 
IT service provider providing Cloud Services. Hopefully, somebody has some 
more info about the university one or maybe we can get that from the 
community. Put what you know about your ecosystems and work with the other 
team members on fleshing it out. We need to get rid of all the TBDs before 
we start making this more widely publicized. 

When we do our next wave of interviews

1. Recruit from those ecosystems. We can focus these on whichever roles we 
choose as a team. And different team members can choose whatever 
ecosystems make sense. We could get data internally, from our own 
customers or the community.

2. Use the companies as part of the personas interview and validate those 
with the interviewees. We could also ask interviewees if the personas that 
we have sound right for their environment.

Longer term

1. Target the Summit after Paris to have something we want to present the 
community.

2. Intersect with any efforts trying to define "the user" and work with 
them to validate, expand and adopt our users.

3. In the meantime, try to instill as a best practice including personas 
or ecosystems as part of the Blueprints process. As part of presenting to 
the broader community, it would be great to have some proof points about 
the value of personas.

If this sounds good, then let's get to work. If not, please send your 
feedback to the mailing list. We can also discuss at the next meeting.

Jeff

Jeff Calcaterra
IBM Cloud Manager with OpenStack User Experience Design and Research
STG UI Patent Board Chair
email: jacalcat at us.ibm.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-personas/attachments/20141012/b371755e/attachment.html>