GAE Pricing Changes - Sucker Punching the Development Community

[zdravko]

Hi folks,

With regard to the newly proposed GAE pricing changes and other
related issues ...

IMHO, A this point the minimum honorable thing that GOOG could do is
to take a gradual phasing in approach along these lines:

a) For the first 3 months, continue charging the old way, while in
parallel showing what the costs would amount to under the new scheme.
This would give everyone a chance to see what they will be facing and
a bit of time to do at least some initial tinkering and refactoring of
their apps. I would be surprised to learn that they have not already
been doing this sort of comparison. Otherwise, it would mean that
they did bulk analysis of resources vs revenues and that they
themselves have no idea how many customers will end up being
negatively impacted - in which case, it would be just as valuable for
them as it would be for their clients.

b) Thereafter, they should start billing according to both schemes,
where in the first month the customer pays 90% of old & 10% of new
charges. The next month it would be 80/20 and so on. Furthermore,
they should ensure that for a full year, no customer ends up paying
more than twice the cost of what it would have been the old way. The
reason why they should do at least this much is to make up for *sucker
punching* everyone - which I will explain.

GAE is not the only way by which GOOG has *sucker punched* the
developer community because they have already done the same thing with
the Translation API and just about all other "data retrieval" API.
The reason why I believe that they have *sucker punched* everyone is
because it turns out that they have done some quite amazing things,
either by design or with total disregard to its developer community
and/or with total disregard to some very basic realities.

It is an economic reality that there was never a way for them to make
any money with any of the data retrieval or data "transformation"
API's such as Translation API. So, what were they thinking or were
they thinking at all, when they offered such services in the first
place? How did they ever hope to make such services economically
viable when such API services do not provide means for things such as
ad insertions? While GOOG has the right to make a profit with
everything that it does, it has *NO* right (not in the past, not now
and never in the future) to offer developers what amounted to a "free
lunch" because too many developers ended up investing their time and
effort, based on that silly "free lunch" premise. While many of us
developers are not too savvy when it comes to issues such as having
economically viable revenue models, GOOG on the other hand is lot more
sophisticated and it should have known better from the very beginning.

When Translation API user community cried foul, GOOG knee-jerk reacted
with a promise that they will offer a paid subscription. Great !?!
NOT REALLY ! The problem is just that, in that it was a knee-jerk
reaction because most developers will not be able to generate enough
revenue to pay for such services - no matter what they price it at.
There is a solution that apparently they have not even considered.
They could have helped those developers in ways by which the
developers could have displayed advertising within their apps and with
that ad revenue, maybe they could end up covering at least their
costs. However, even with that there would be a huge problem because
GOOG seems to do nothing that does not scale well without requiring
lots of human intervention such as reviewing apps for compliance, etc.

Here is a really fantastic article which deals with some of these
"automated scalability" issues. http://www.wired.com/magazine/2011/03/mf_larrypage/all/1

When all is said and done, has GOOG even bothered to come out and
state how much more revenue are they trying to generate with the new
pricing scheme. Is it 10% or 50% or 100% or 500% or more? If they
stated that much and if they gave us 3 months of new billing data
(before new billing kicks in) then everyone would be able to see if
they are edge case exceptions or whether they fall in-line to that
stated average revenue increase goals.

Now, while they have either been totally irresponsible or perhaps they
just made a callous revenue projection mistakes, the problem is that
they are evidently continuing to do more of the same with the "free"
quota offerings. The fact of life is that while there is no "free
lunch", GOOG continues to make believe the developer community that
there is. It would be interesting to know just how much of the
overall resources are being eaten by their "free" offerings. In other
words, how much of that "free lunch" is factored into the new pricing
and with that, how much of GOOG's research and development (yes,
research into what types of apps can be economically viable) and how
much of their overall business development is being funded by the
proposed pricing increases?

If I did not know any better or if I did not have too much faith in
GOOG thus far, I might be inclined to think that most of this mess was
by design - by which they used the developer community to do just that
- to flush out and even to poach great ideas that had no hope on their
own but core of which would indeed be viable within GOOG itself.
Please tell me that I am totally wrong, before real disillusionment
sets in.

Sincerely,
zdravko

[Kaan Soral]

I couldn't read all of it but I am giving you a +1

And about the wired magazine link, I thought I was the only one who
insult people/companies with pretty urls, but theirs is very very
harsh

> "automated scalability" issues.http://www.wired.com/magazine/2011/03/mf_larrypage/all/1

[zdravko]

> I couldn't read all of it but I am giving you a +1

That's an oxymoron - good enough for a +1 but not good enough to be
read in it's entirety. ;)

[Joops]

Hello, just my thoughts on google and beta products for developers,

Google has historically used an approach of rapid early releases and
then rapid improvements.

Their strength is the speed at which they can innovate, because they
don't need to manage client software.
(Compare MS release cycle with Googles).

However, when people are building technologies on beta releases, there
is a serious tension between getting something in peoples hands, and
making it clear what the offering is. At the beginning they don't
really know what the offering is.

Google is like many tech companies, some of its lofty ideas don't do
as well in the market. (Hi Wave and Buzz!). So it must be difficult
when they release an early tech as a beta, not sure of exactly how it
will be used by the web at large. Even more so when this is a
technology that developers will use, and build upon.

I personally am ok with how Google has approached this, but I am doing
a free time project, and I joined the party quite late. Judging by the
various voices on this list, it would appear that there is a
significant sized group that agrees with you two.

J.

ps.. This post is mainly a procrastination excercise to avoid the
coding I know I should be doing.

[Branko Vukelic]

The part (s)he read was good enough for +1. :)

Anyway, I agree with 1), but not with 2) (the part where it gradually
increases to 100% new pricing. The reason I don't agree with 2) is
that it invalidates the data you have in your dashboard. What's the
point of knowing the 100% new pricing, if you're not going to be
charged that after the first three months, but some other value. It's
even harder to plan that way. But I'm sure 3 months head start is good
for finding out whether your real costs will be bearable after the
three months have passed.

--
Branko Vukelić
bra...@herdhound.com

Lead Developer
Herd Hound (tm) - Travel that doesn't bite
www.herdhound.com

Love coffee? You might love Loveffee, too.
loveffee.appspot.com

[zdravko]

> What's the point of knowing the 100% new pricing,
> if you're not going to be charged that after the first three months,
> but some other value. It's even harder to plan that way.

The point is to know what the new bills will look like, before they
kick in - so that one can plan, scramble, run or whatever.

Pozdrav :)

[stevep]

There are significant moral hazard issues with Google's current
latency = instance step-cost pricing scheme. I don't see anything Greg
talking about which address these.

I'm not sure they "sucker punched" anyone, but I do wonder how they
can so completely ignore these areas that demand transparency. Do they
think the Google brand is that trust worthy??

This is a combined "elephant / 800lb Gorilla in the room" for me, yet
it is never mentioned by Google.

cheers,
stevep

[Branko Vukelic]

On Tue, Jul 5, 2011 at 1:23 AM, stevep <pros...@gmail.com> wrote:
> This is a combined "elephant / 800lb Gorilla in the room" for me, yet
> it is never mentioned by Google.

Yeah, maybe because there were already double-elephant-giraffe combo,
and two tasmanian devils, discussed in other threads, so nobody is
surprised anymore.

[zdravko]

>. I'm not sure they "sucker punched" anyone,

I honestly believe that they sucker punched themselves as much as they
did their developer community. There is not much that they could have
learned over the past 3 years that they could not have anticipated
from the very beginning. At the time when Amazon was charging back
based on resources "hogged" (regardless if they are used or not) GOOG
tried to outsmart everyone by charging for what is being used -
without apparently (or so they NOW tells us) having any solution in
place by which resources are indeed being shared and never hogged when
not being shared. Even then, where I personally fault them the most
is that they waited until they ran into (what at least seems) like a
brick wall. Surely, they could have projected the inevitable after
the first 3 months of real usage patterns analysis. So, Yes. In the
process of sucker punching their developer community, they also sucker
punched themselves because most of this fall out and damage is
something that even their MONEY can not buy back and make right.

[Stephen]

On Tue, Jul 5, 2011 at 9:40 AM, zdravko <email.w...@gmail.com> wrote:
>
> Surely, they could have projected the inevitable after
> the first 3 months of real usage patterns analysis.

You would think so. Is it credible that they didn't?

History shows that they waited 10 months before introducing billing,
tweaking quotas, adjusting measurement techniques, and reaffirming
their desire to offer about 5 million free page views/month:

http://googleappengine.blogspot.com/2009/02/new-grow-your-app-beyond-free-quotas.html

The situation is that 'App Engine for Business' has been renamed 'App
Engine' and 'App Engine' as we knew it does not exist any more.
They've been pretty clear about that.

[zdravko]

> They've been pretty clear about that.

If I am following (which would not be the first time that I am not)
and if we are talking about the same things then when is it that they
started to indicate that their whole resources measurements and
billing model is broken and that it needs to be rethought ? My main
point is that they waited too long and that way less damage would have
been cause much earlier - before too many developers invested
themselves to heavily on a "free" and/or "very inexpensive lunch"
platform. My whole point is that if there is anyone in the world who
can figure it out and who has empirical data with which to figure it
all out then its certainly GOOG. Perhaps they tried to be too
innovative while even to this day they appear not to have the
technology that enables true resource sharing so that it is not being
hogged - be it cpu or memory. So, if you can not make something
shareable and reusable then what ever made them think that they could
ever be able to sustain the burden that is caused by the fact that
resources that were being hogged while not being used were indeed
being given away. So, they swung the pendulum from that silly notion
all the way into the new scheme of where they have 15 minute penalty
for a resource that they much earlier grab back and reuse elsewhere.
With that 15 minute period in mind, just how many times could the same
"memory" end up being used and concurrently charged. That is not only
a draconian change but it also reeks of unfairness - even dishonest
unfairness.

[Stephen]

On Tue, Jul 5, 2011 at 10:48 AM, zdravko <email.w...@gmail.com> wrote:
>
>  My main point is that they waited too long and that way less damage
> would have been cause much earlier - before too many developers
> invested themselves to heavily on a "free" and/or "very inexpensive lunch"
> platform.

Goals have changed, and therefore prices have changed. App Engine is
an Enterprise product now and price has no relation to cost.

You've already explained that Google could and should have known that
it was undercharging, if it was, in less than 3 years. So, maybe they
weren't? Goals have changed.

Personally, I think the goal of selling to the Enterprise at the
expense of the wider web is bad goal. After all, Google's mission is
to organize the world's information and make it universally accessible
and useful, and this is not well served by cutting off access to the
world's most cost effective and scalable serving infrastructure to a
large fraction of the people who can best make use of it.

Greg D makes Moore's Law run backwards -- a good start to a collection
of Chuck Norris style jokes, but not something to be proud of.

[zdravko]

Stephen, thanks for basically repeating what you already said because
the first time I totally missed the notion (which I must admit that I
never considered) is that their goals might have indeed changed - in
which case, it just means that they DOUBLE, TRIPLE ***SUCKER
PUNCHED*** the developer community by using them like cheap dirt in
order to get the platform working (with suitable developer sucker in
pricing) in order to take the kinks out of the platform, in order to
make it usable by the big boys.

[Gopal Patel]

and in the meanwhile aws and ms both making incoming data completely free... ( lets see how google behave when their own game of "free" played against the. )

--
You received this message because you are subscribed to the Google Groups "Google App Engine" group.
To post to this group, send email to google-a...@googlegroups.com.
To unsubscribe from this group, send email to google-appengi...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.

[Drew Spencer]

Flippin' eck, this is a heavy thread.

I'm still a bit of a newbie to this appengine malarkey, and I've never even uploaded an app yet. I really hope that appengine doesn't turn out to be too expensive for smaller, non-enterprise developers.

I've thought from the start of my GAE journey that one reason for Google offering the service was to let small time developers create apps on their platform, making it easy for Google to buy up and integrate apps/companies into their network, so I would be surprised if they make it unfeasible for regular joes like us to use it.

Let's hope so, as i have invested a lot of time and love into it already. Goodness knows what some of you others have! Don't let us down Google - "don't be evil".

Drew

[zdravko]

>> I've thought from the start of my GAE journey that one reason for Google offering the service was to let small time developers create apps on their platform, making it easy for Google to buy up and integrate apps/companies into their network, so I would be surprised if they make it unfeasible for regular joes like us to use it.

Why would I buy you out and pay you anything if I can put you out of
business and make all that you have, free for the taking. Is that not
what Microsoft did and is GOOG not everyday behaving more and more
like the Microsoft of old?

[nickmilon]

Stephen +1

"The situation is that 'App Engine for Business' has been renamed
'App
Engine' and 'App Engine' as we knew it does not exist any more. "

This describes the situation we are in.
Problem though is: 'App Engine for Business' has not been really tried
in the battlefield.

On Jul 5, 12:12 pm, Stephen <sdeasey+gro...@gmail.com> wrote:

> On Tue, Jul 5, 2011 at 9:40 AM, zdravko <email.workbe...@gmail.com> wrote:
>
> > Surely, they could have projected the inevitable after
> > the first 3 months of real usage patterns analysis.
>
> You would think so. Is it credible that they didn't?
>
> History shows that they waited 10 months before introducing billing,
> tweaking quotas, adjusting measurement techniques, and reaffirming
> their desire to offer about 5 million free page views/month:
>

> http://googleappengine.blogspot.com/2009/02/new-grow-your-app-beyond-...

[Drew Spencer]

On Tuesday, 5 July 2011 21:40:34 UTC+1, zdravko wrote:

Why would I buy you out and pay you anything if I can put you out of 
business and make all that you have, free for the taking.  Is that not 
what Microsoft did and is GOOG not everyday behaving more and more 
like the Microsoft of old? 

 

Because you can't just put me out of business... that's not how business works. You can't just steal a customer base that easily.

http://en.wikipedia.org/wiki/List_of_acquisitions_by_Google

And no, I don't really see any similarity to Microsoft. Comparing apples with oranges.

[zdravko]

>> Because you can't just put me out of business... that's not how business works.

GOOG can certainly put you out of existence by suddenly pricing their
services out of your affordability range.

>> You can't just steal a customer base that easily.

Without you in business, your customer base will go to whoever is
providing the same service.

>> And no, I don't really see any similarity to Microsoft. Comparing apples with oranges.

Nevertheless, both apples and oranges are our dependencies on their
monopolies and self serving control.

[vivpuri]

+1

[Drew Spencer]

On Wednesday, 6 July 2011 16:45:23 UTC+1, zdravko wrote:

>> Because you can't just put me out of business... that's not how business works.

GOOG can certainly put you out of existence by suddenly pricing their
services out of your affordability range.

 

But if they want to charge me extortionate rates, they have to do it to the whole appengine customer base.

 

>> You can't just steal a customer base that easily.

Without you in business, your customer base will go to whoever is
providing the same service.

But I would just move my service to another host, keeping my customer base.

 

>> And no, I don't really see any similarity to Microsoft. Comparing apples with oranges.

Nevertheless, both apples and oranges are our dependencies on their
monopolies and self serving control.

I think you're being unfair. Google has never tried to exercise its near-monopoly on the market like Microsoft did and still does. Google is where it is because it's products are the best. It destroyed hotmail and yahoo mail with gmail by just making a better alternative and you can still get to hotmail through google search, can't you? Why would you though - it's crap.

[Geoffrey Spear]

Don't feed the trolls.

[zdravko]

Perhaps with a bit of an explanation you would like to address
something that is incorrect ;?)