Re: Wild speculations about the "other" factors
- From: "Simon Kissel" <kissel@xxxxxxxxxxxxxx>
- Date: Tue, 17 Oct 2006 11:39:50 +0200
Michael,
first of all, thank you for your extensive response. As this
thread was about unfunded speculation, I didn't really expect
anyone from Borland to step it, really giving insights, that's
a nice suprise :)
[...]1.) Enterprise customers - maybe there are a few VERY
big enterprise customers that buy thousands worth of
Delphi copies through a direct channel at Borland.
You just described the largest market in the software development industry[...]
by far. Also the segment with the largest portion of Delphi sales and
revenue every release since v1 client/server edition.
Delphi's sweet spot is the small to medium enterprise, up to 50 developers
or the department within a larger IT organiztion. Companies typically use
Delphi because they can bootstrap a working project very quickly with a
smaller number of developers than their fully built out VS or Java teams.
Delphi teams are typically the tactical "special ops" team in a
organization. The VS teams are like the Army. You won't find as many jobs,
and the team sizes are smaller, thats partly a result of Delphi's
"advantage". However I think we can do more from Borland's (DTG's) side to
improve the awareness and availability of jobs though. Anyhow it's good
point, in fact we keep tabs on Dice and Monster data as well.
Ok, so you are saying that the number of enterprise customers with up
to 50 Delphi seats makes up a bigger fraction of your Delphi revenue
than those that can be tracked by public data sources.
What are your findings on what these enterprise customers mostly use?
Are they in total contrast to what the customers represented by the
"public samples" say? Have they shifted from maintaining Win32 projects
to .NET? Or are they mostly doing ASP.NET stuff with Delphi?
Well, in any case, we now have a clear response about what those
mysterious "other" factors are - a big number of enterprise customers
that use Delphi for inhouse development, thus not leaving behind
traces in the samples taken. Thanks, that makes seeing the big picture
possible for the first time for me.
However, it also means (at least if those enterprise customers mainly
do Delphi.NET) that you've got two big camps of customers with very
conflicting interests, and whatever side you focus on (for the last
releases, the enterprise customers, obviously), the other side will
be unhappy. Not a nice situation I guess.
There is another thing: While stock market stuff
isn't exactly my cup of tea, from what I've read the
IDE products in total right now make up about 7% of
Borland's 300M yearly revenue (attention, very rough
numbers, if anyone with better ones could please step in,
thanks). So probably the revenue generated per Delphi
version is under 20M, and as it includes Java, possibly
much smaller. That's not much. That's not much at
way way way off. From Borland 2005 Annual Report "Revenue from the IDE
products were approximately $84 million, or approximately 30%, of total
revenue in the year ended December 31, 2005". I can't say anything about our
2006 mix, but that your numbers are way way way off.
Well, I said that I didn't research those properly. Actually
I only did a quick google search on "Borland earnings", and what
I got was that Borland currently makes about $75M per quarter.
So, I multiplied this by 4, and got $300M. Then I found an
article about Borland selling its IDE business, and that said
that IDE products now only make up 7% of the revenue, dropping
from previously 14%. Some of those numbers may be wrong or
highly inaccurate. As you are saying it's rather 30%, this might
have been "hooray, Borland drops IDEs and focusses on ALM"
propaganda I've fallen for.
What happens when data is bad? All the data that extrapolates from it is
even worse. Accurate data and info is critical to analysis. BDS has sold
very well. This guess just isn't even in the ballpark at all.
I know, that's why I said so in my post that those numbers probably
are off and the unit counts seem to be far too low.
If anyone cares and has the time, he may invest more research to
find out, but it's not that important to me anymore now that
you've given your thoughts.
I don't want to come off sounding like individuals and ISVs aren't
important -
You don't. If it's really the case that the enterprise shops
are a majority, and if it's also the case that their needs
are very much focussed on .NET, your roadmap makes much more
sense to me than it did before. It's still a highly unfortunate
situation, as there obviously is a high degree of frustration
with you mid-size/small-size/individual developer customers,
who have other focusses. But it's understandable.
2.) Warning, extremely unfunded wild thought without
any base whatsoever here:
What if it's an enterprise customer who is simply buying the
DTG right now so they can dictate their needs to them?
That indeed would be an "other" factor and a good reason
to no longer care about us customers.
Lol. Would be an interesting story... but way way off. Big points for
creative thinking though! :)
Good to hear that this theory is out ;)
I just want to set some of the perception vs reality straight because you've
put a lot of time and thought into this. If fact with somewhat accurate data
I would love to see an alternate roadmap v2 :o)
Well, obviously I can't do this - I would need data from sources that
are completely locked away from the public, your enterprise customers.
FWIW, I'm also still not too sure you really got much up-to-date data
about that at your hands yourself. After all, the Delphi sales numbers
from recent years don't give a basis to derive what parts / languages /
platforms are used by your enterprise customers.
I guess (and I've seen indicators that this is coming) that a public
survey directed at all your customers asking about their crucial needs
for today, 2007 and 2008 would help a lot to create a much better
roadmap than both the current one and the one I've written.
Though, one really critical thing that is missing from your alternate
roadmap is real development costs and time estimates which are are as
important or more so than the requirements because they really determine
your ability, capacity, and timing. There are some things you mention that
might seem trivial or on equal footing from a development perspective but
are actually many times more costly (time and engineers) to develop and
maintain. This can be the difference between something making sound business
sense or bankrupting a product or company. Knowing what things cost and how
long they take makes all the difference.
Frankly, I don't see that. OK, of course the roadmap makes a few assumptions
that may be wrong - if not even research and preparations for a Win64
compiler have been started, getting a preview command line compiler out
for Highlander of course is undoable with the resources at hand.
But I definately think that dropping that whole "Delphi for Vista" release
out of the roadmap would enable you to do at least one native-code
*focussed* release again after those years.
I also think that integrating more of the passioned users into the process
would help a lot. After all, many (including me) are offering their help
for free instead of only being vocal about their needs.
Anyhow I do want to say thank you very much for the input and rest assured
it was time well spent and appreciated. Even if your data and assumptions
aren't always in the ballpark, some of your conclusions and ideas are really
good. Customer input is valuable and critical, but input such as what you
prepared is more than valuable and it shows just how incredibly special and
great Delphi developers are. Thank you. Looking forward to more input and
insight from you and others.
Michael, again thank you very much for taking your time to give a
very detailed view about how things look from your side. This has been
the most helpfull post on this topic by far.
Simon
.
- Follow-Ups:
- Re: Wild speculations about the "other" factors
- From: Shloma Baum
- Re: Wild speculations about the "other" factors
- From: Wayne Niddery [TeamB]
- Re: Wild speculations about the "other" factors
- References:
- Wild speculations about the "other" factors
- From: Simon Kissel
- Re: Wild speculations about the "other" factors
- From: Michael Swindell \(Borland\)
- Wild speculations about the "other" factors
- Prev by Date: Re: Can we achieve better statistical insight into the Delphicustomerbase?
- Next by Date: Re: Can we achieve better statistical insight into the Delphi customerbase?
- Previous by thread: Re: Wild speculations about the "other" factors
- Next by thread: Re: Wild speculations about the "other" factors
- Index(es):
Relevant Pages
|
