Thunderbird market segments
dascher at mozillamessaging.com
Mon Apr 26 18:05:48 UTC 2010
Kent, great questions.
My analysis on this market segmentation question is fairly coarse, but
maybe useful. This is mostly derived from anecdotal information rather
than great data, because solid analytics are really hard to come by.
I believe the two largest segments for current users of Thunderbird are:
1) "Institutional" or "enterprise" deployments, such as universities and
governments, who deploy Thunderbird primarily because it's open source,
and gives them a traditional email client (rather than webmail) which
does not tie them to a proprietary vendor. It's hard to tell how many
large businesses use Thunderbird -- I suspect most of our corporate
users are outside of North America, due to the different relationships
with Microsoft in different geographies. (Note that for most
universities I know about, desktop clients are primarily allocated to
faculty & staff, and students, more mobile, use webmail).
2) SOHO/SMB-type users for whom Thunderbird is the preferred choice
because of a combination of: a) history (many Thunderbird users picked
Netscape years ago), b) customization options, including add-ons, myriad
preferences, c) an affinity with Mozilla and the values of openness,
user control, etc.
There are other groups in my mind (and I realize my breakdown of the
market is similar but slightly different than yours), but the above two,
I believe, account for most of our users. This is a very gross
generalization, and indeed many educational users also like the Mozilla
mission, and there are people using it as a "home email client" as a
complement to their Outlook usage at work, etc.
Still, the dichotomy above is I think quite interesting, and portentous
(which is a word I wouldn't normally use except how often does one get a
chance?). In particular, I see the above two segments wanting different
things out of an evolving product.
Institutional admins have a job that I remember well (I used to be the
mail admin for brown.edu, dealing with about 6k users). In many ways,
success comes from establishing enough control so that users don't cause
too much havoc, because there's always way more users than there are
hours to help them. In particular, those deployments of desktop software
need things like remote management, more or less locked-down setups,
centralized configuration, multi-year roadmaps, SLAs, second-tier
support arrangements, interoperability with systems like Exchange,
blackberries, etc. The systems they want Thunderbird to integrate with
tend to be internal systems like inside-the-firewall calendar servers,
LDAP servers, CRMs, Exchange servers, etc. (Educational institutions in
some ways have it even worse than corporate deployments, because their
user populations resent central control, and their budgets are often
tight). These deployments tend to prefer slower product evolution.
SOHO/SMB users tend to be more like "consumers". They get software
straight from the "vendor" (or from their Linux distro), do security
updates as prompted, relish the ability to be "in charge" of their
software life. They tend to want to integrate with the internet rather
than the intranet, and many are exploring varieties of communication
systems from Facebook to Twitter to Skype, etc. They tend to live more
and more on the web, and expect more and more all of their software to
integrate with the web. These users range in their attitude towards
change and risk, ranging from early adopters to more conservative users
(most of us actually have different attitudes towards different things
-- e.g. I like my cars and clothes old, my web toys shiny and new).
A couple of additional meta-points:
- Both market segments periodically evaluate their use of desktop
software in a world where the web is becoming more potent daily, but
where their feeling of control over the software they use on the web is
being challenged as well. I think we need to build tomorrow's software
with as full an understanding of these factors as possible. I can talk
forever about that, but it's a bit off-topic.
- Mozilla's culture, at least since the inception of Firefox, and as
described by the Mozilla Manifesto, is user-centric, not
business-centric. When having to choose between something that is better
for people or something that is better for organizations, we routinely
choose the former. I'm proud of that and I don't see that changing.
- I believe it is easier to get unpaid contributors to projects that
are aimed at individual users; I believe it is easier to build
businesses making software for businesses. All other things being equal,
of course (_Ceteris paribus_ as philosophy teachers say).
- Mozilla's greatest successes to date center on making the web
accessible, fun, useful. I don't think that's an accident. I think the
web was built to be open, and as a result, a small group of people can
have a large impact with it. I don't know anyone who'se been able to
have similar scale impact in the world of traditional messaging stacks,
for a variety of reasons.
My thinking about where we should go from here is evolving, but my
current thoughts are:
- Mozilla Messaging will be exploring possible new features for
Thunderbird primarily through add-ons, so that we can iterate fast, and
take risks which we can't take with trunk. As an example of the kinds of
things I'm thinking about, see Shane Caraveo's blog:
http://shane.caraveo.com/ -- he's been getting familiar with the
codebase, and spinning up a few experiments. Aspects that I'd like us to
explore include making Thunderbird more competitive for the
"SOHO/SMB/consumer" market, including integration with web-based
services, including Weave for sync, etc.
- We'll also keep working on the "platform" aspect of our codebase, so
that both we and others can experiment more easily. That includes things
like keeping up with the work that the Jetpack team is working on, and
likely building more bits of componentry into the platform.
- I don't think it would serve our mission to set up Mozilla Messaging
as an enterprise-focused business, and I don't think we have the DNA for
it. However, maybe others can fill in the gap. To explore that, we've
been thinking about setting up an email list where admins of large
Thunderbird deployments can drill into the issues which get in their
way, and, I'm hoping, find companies, add-on authors, and contributors
who can help them. The conversation about autoconfig in universities was
interesting, but I think we can have maybe less virulent versions of
those discussions in a better defined forum, and where I can point
people who might be able to help.
- As an aside, I'm really interested in your work on EWS. I think
there are lots of opportunities there, especially for someone who's
interested in building a business around it. I also think it's
potentially a very large project, and has definite technical risk, which
is why I like the fact that you're poking at it to understand those
areas of risk.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the tb-planning