IBM/COGNOS Pain

[Jeroen Eynikel]

Two reasons, IT policy and IBM not being obliged to help you with any issues. IT policy is often a big driver for this.
Cheers

Ok. So there is not a good reason. Only political reasons

[bihints.com]

sorry, I just went through our IBM/Cognos quotes again after some of the feedback here.
I didn't see the lines in small characters at the bottom of our last year fees quote claiming it is VAT free . typical.
So the increase is "just" 10%.

Though the fact remains we are still pushed to upgrade to 9.4 with all the issues that implies.

[Jeroen Eynikel]

Though the fact remains we are still pushed to upgrade to 9.4 with all the issues that implies.

I think you can only partially blame IBM for that, i.e. I agree that any new release should be officially supported for a longer period than is currently the case. The other part of the blame should be on stupid internal IT policy.

[bihints.com]

I think you can only partially blame IBM for that, i.e. I agree that any new release should be officially supported for a longer period than is currently the case. The other part of the blame should be on stupid internal IT policy.

well we can't say IBM is making it easy.
.they do not keep 9.0 support alive until 9.4 is "ripe" (i.e. all 9.0 functionalities are present and behave like in 9.0)
.they do not supply a tool to "downgrade" .cub files to their previous format so we can revert to pre-9.4 systems if 9.4 proves unsatisfactory in our live environment. I can live with strings being cut back to a max 255 characters.

[mattgoff]

We recently received the yearly quote for support fees, these increased by 26.50% vs last year, without any additional products or licenses.
And that is for the privilege to be forced to upgrade to 9.4. We are still on 9.0.3, so we have to skip directly to 9.4, as 9.0 support ends in December 2009 and 9.1 support ends in December 2010.
Moving to 9.1 would just land us in the current situation where we fork a load of cash to be told to upgrade to a newer version as 9.1 probably won't get anymore fixpacks next year.

So is it just us, or are you seeing a similar hike in the yearly support fees?

Not that bad, but still. 10% year on year increase in maintenance seems to be the standard over here. A lower maintenance increase is possible in some specific cases though, but far from guaranteed. Lowest I have seen so far was 5% year on year increase but that was the exception rather than the rule. (i.e. they really really wanted to get the deal and the client made a very big issue of the standard 10% increase (i.e. dealbreaker)).

Our original sale contract with Applix stipulated the maintenance fee increases would be tied to the RPI. After receiving an invoice with a large renewal fee this year, I was eventually able to get them to comply to the agreement and significantly reduce our support contract fee. You all might want to check your contract to see if you have a similar term in there as well. AFAIK, this was a standard part of the Applix UK contract, not something we asked to have inserted.

Matt

[Alan Kirk]

but a giant corporation like IBM moves slowly.. you just have to accept that.

But that's been the plaintive cry of dinosaurs since the dawn of evolution; "You have to make allowances for us because we have too much bulk being powered by too little brain". Unfortunately for them the planet (or in this case the market) doesn't. They may rule the planet for an epoch give or take, but extinction is still their inevitable fate as smaller, more adaptable life-forms (hello, Palo) evolve.

In this particular thread there have been several examples:
- "Support" consisting of being told to "upgrade to 9.4". (When was the last time a fix pack for the "supported" 9.0 OR 9.1 was released, given that both still have known bugs?);
- Substantial upgrades to maintenance fees; and
- The absurd new enhancement request procedure which serves two purposes; first, to put more bureaucratic inertia in the customer's way to reduce the number of enhancements that they need to look at that they weren't planning to do anyway, and second to provide a triumph of process over end results.

Both purposes, it may be noted, are the antithesis of "customer focus", though at least the latter would have made Sir Humphrey Appleby proud.

I'm not for a moment suggesting that IBM needs to act on every enhancement request that they receive. For a start, some are just plain daft, and some might be good for some users but bad for others. However if a company was truly customer focused, what they might do is, oh, I dunno, random thought that people wearing blue suits and ties wouldn't be allowed to have... maybe have a page where enhancement requests can be submitted. A page which made such requests visible to all users. (Hey, they could use our Enhancements forum, at least people can find things on this site! ) A page where there can be discussions about the pros and cons of the requests, where such requests could be voted upon, and which could help drive the direction of the product.

But that clearly couldn't be a good idea (to them) because I suspect that there are still too many people in Armonk who see the IBM of 2009 as being the IBM of 1969; able to dictate the terms of the market and the way that (they think) products should be. Very little has been learned from the way the market handed them various pieces of their anatomy on a platter after the PS/2 and OS/2 debacles.

Microsoft, for all their faults, used to exhibit such flexibility before they too became moribund and started having waitresses deciding to inflict new GUIs on unsuspecing users without a classic fallback one. The MS Wish program, for example, was a useful way to get feedback to the development teams before MS axed it. It was things like that that helped MS cut IBM off at the knees in many markets, but that was then, and this is now, and the subject isn't MS but Big Blue.

Why should I care? I don't, not about IBM, since I'm not an IBM stockholder. But I don't want to be one of the putzes left holding the next incarnation of OS/2 (potentially, TM1) when the Apatosaurus finally gets boinked on the skull by an asteroid because it's been focusing on pretty dashboards rather than the raw product functionality. Especially if in the meantime the rest of the world has moved on to a more nimble life-form which survived because it made its priority... the customer. (That's actual priority, not just lip-service priority, and by "priority" I don't mean "milking as much as they can get out of them".)

Such an extinction won't happen overnight unless another asteroid hits the Yucatan Peninsula, in which case we'll have bigger problems than whether SaveDataAll still hangs a server in 9.1. And indeed it may never happen, but having a focus on selling a product isn't the same as having a focus on supporting it. (And on the latter point, that includes not overcharging; see also "goose, golden".)

And let us be honest: Applix support or roadmaps was always more a question of knowing the right people. If you had to deal with unknown helpdesk person X it was a nightmare as well.

Not only do I not disagree with you, I pre-emptively agreed earlier. You couldn't be more right there.

Applix could well have done with a site like the one mentioned above too instead of wasting resources on that waste of space Applix On Demand. There were many, many posts in the old Forum about just how hard it was to get good information on what was getting fixed and when.

(Edit: Upon that point of nostalgia, I stumbled across this ancient post from Garry Cook on the old forum while I was looking for something else:

I'm sure you know my personal feelings (along with a significant proportion of other forumites) about the lack of feedback on issues such as this - I was hoping that since Jason had taken such an active role in setting up that forum that posting there might actually mean he came back with a response but it's now been over a fortnight with no response. All I want is for him to recognise that there's a whole host of features/minor bug fixes there that he could easily hand to development to make some big improvements quickly to the system.

"That forum" refers to the late and unlamented Applix On demand (AOD). Plus ça change, plus c'est la même chose.)

Certainly the release notes these days are an order of magnitude better on that point.

IBM will do what benefits IBM

I allllmost agree with you there; I'd add just one qualification.

"IBM will do what it thinks benefits IBM"

Unfortunately when a corporation gets too large, it sometimes can't see what's truly in its own best interests. Information about what's happening in the market takes too long to filter through to the decision making levels, and even when that does happen there are so many people who won't want to rock the boat (and potentially their own careers) that a decision to respond to the problem can't get made, or can't get made in time. And thus you have... hit it:

(e.g., shoddy support, new features to promote marketing over stability, website) and IBM is mostly IBM Global Business Services. The products will go in the direction that will have the most benefit to the GBS side of the business. 'Tis folly to hope for anything else.

Esattamente

(If you want history on how well IBM does with commercial software, you may want to consider Lotus Notes as an example)

Yup, or any other Lotus product, or Via Voice, or...

[Martin Ryan]

However if a company was truly customer focused, what they might do is, oh, I dunno, random thought that people wearing blue suits and ties wouldn't be allowed to have... maybe have a page where enhancement requests can be submitted. A page which made such requests visible to all users. (Hey, they could use our Enhancements forum, at least people can find things on this site! ) A page where there can be discussions about the pros and cons of the requests, where such requests could be voted upon, and which could help drive the direction of the product.

The problem with such an approach would be that it would be like the old form of democracy where however shouted the loudest or spoke the most emotively would get things their way.

Now that I'm working at a large company with the afore mentioned buckets of money, I've been finding IBM pretty responsive. Senior IBM managers were even issuing apologies when some IBM laptops showed up to training an hour late as they were very concerned it tarnished their professional image.

They definitely have a strategy, its just not a strategy that seems to be particularly concerned with the medium sized companies that used to be Applix's target. Now they are only interested in the bohemeths. The problem is, TM1 is still not an industrial strength product. E.g. promoting objects to a live environment that requires 24/7 up time requires you to rebuild the object (unless you have Calumo's slick tool, which we can't get it due to policy. Though building objects manually on a live instance is also, strictly, against policy, but needs must), maintaining versions of dimensions across instances is a nightmare and you can't replicate different objects at different times.

If they don't manage to produce the industrial strength tool, they're going to find they've alienated their old customer basis and do not have a tool suitable for the group they want to target. But I suspect they will, and they'll get their bohemeths, and will just the little people prop up their cashflow.

Martin

[Alan Kirk]

However if a company was truly customer focused, what they might do is, oh, I dunno, random thought that people wearing blue suits and ties wouldn't be allowed to have... maybe have a page where enhancement requests can be submitted. A page which made such requests visible to all users. (Hey, they could use our Enhancements forum, at least people can find things on this site! ) A page where there can be discussions about the pros and cons of the requests, where such requests could be voted upon, and which could help drive the direction of the product.

The problem with such an approach would be that it would be like the old form of democracy where however shouted the loudest or spoke the most emotively would get things their way.

That would be where the voting bit comes in. If it's a bad idea, no amount of loud shouting is going to overcome 500 votes to 10 against. Conversely, if it's a good idea with a good outcome, who cares how it gets there, as long as it gets there. I'd rather have one Julius Caesar who's capable of seeing an issue and speaking passionately on it (and cutting and thrusting if necessary) than 50 UN representative wannabes sitting around in overstuffed chairs, oh so politely and eruditely discussing all 47 sides of an issue and never reaching an accord because everyone is too concerned about the possibility of giving offence.

Of course, both are better than a supplier which takes the approach of "you'll take what we give you and like it".

Now that I'm working at a large company with the afore mentioned buckets of money,

There are any left in the current environment? Even former behemoths need to be watching costs very closely indeed, with the possible exception of BHP which can write off a few $billion on a nickel plant and not even feel it.

I've been finding IBM pretty responsive. Senior IBM managers were even issuing apologies when some IBM laptops showed up to training an hour late as they were very concerned it tarnished their professional image.

Well they're plenty responsive... where sales are concerned. Or other cashey money things like training. Once they HAVE the cashey money, though...

They definitely have a strategy, its just not a strategy that seems to be particularly concerned with the medium sized companies that used to be Applix's target. Now they are only interested in the bohemeths. The problem is, TM1 is still not an industrial strength product. E.g. promoting objects to a live environment that requires 24/7 up time requires you to rebuild the object (unless you have Calumo's slick tool, which we can't get it due to policy.

I think you're referring to Vizier; that's Cubewise, not Calumo.

Though building objects manually on a live instance is also, strictly, against policy, but needs must), maintaining versions of dimensions across instances is a nightmare and you can't replicate different objects at different times.

If they don't manage to produce the industrial strength tool, they're going to find they've alienated their old customer basis and do not have a tool suitable for the group they want to target. But I suspect they will, and they'll get their bohemeths, and will just the little people prop up their cashflow.

One problem with that is that behemoths themselves are an endangered species for the same reasons described earlier, except perhaps in industries where there are high barriers to competitor entry. Also, if a company narrows its focus excessively, their products become irrelevant. A classic example is DB2. As of 2006, the last figures that I saw, DB2 still held an estimated 21% of the DBMS market by revenue. (Oracle was so far out in front on that score that it wasn't funny.) However SQL Server was closing the gap rapidly, and part of the reason that it was in 3rd place rather than 2nd was that Microsoft wasn't doing huge price gouges. It was making a decent profit for a decent product, and I'll betcha that in installed systems, SQL Server is by now streets ahead of DB2. Seriously, who even contemplates DB2 for a new installation any more unless they're running a mainframe? It's usually Oracle or SQL Server, unless something like My SQL can serve the purpose. DB2 has become so irrelevant that one IBMer from the UK even reflected on whether it might go open source. That 21% is history, and history doesn't always repeat.

And when you get to that stage, then sure, IBM can express great joy when they get their behemoth contracts, but reality may get in the way. "Yyyyyay, party, we got General Motors as a customer!!!" {Picks up paper, reads the business section...} "Uh-oh. Well, at least we still have Citibank." {Reads futher...} "Uh-oh..."

[Steve Vincent]

...If I had to write a business case for paying the maintenance invoice given the possible future benefits of development, how they apply to our application of TM1 and the level of support that is recieved, well I'm not sure where I would start....

Thats exactly where i am right now. A possible large increase in users and astonomical (or is that astrocomical?) cost for doing so, and if i'm really honest i cannot say the result justifies the cost. So much so that the business case is likely to need studies done on alternative products - cue interest in Palo.

As for why pay for support - right now i really don't know why we do. IBM refuse to deal with us direct (we're a 3rd party in relation to the contract) and our IT supplier knows as much about TM1 as i can write on a pinhead. If i could navigate round their site then maybe i could at least say we get "free" upgrades. I don't get any warm happy feeling from IBM so far and as a result i find it very hard to justify the cash we are paying for it

[Martin Ryan]

Now that I'm working at a large company with the afore mentioned buckets of money, I've been finding IBM pretty responsive.

I take it back. Been over a month waiting for a fairly simple query. It's so long that I'd forgotten about it when I made that last post. Trawling through emails I realised it was still outstanding.

Martin

[Jeroen Eynikel]

- Substantial upgrades to maintenance fees;

- The maintenance % went up yes - but TM1 on a whole became substantially cheaper and the licensing model became easier. Gone are the days when you had to pay extra to use TI, the web, 64 Bit. Gone are the days were you had to pay for several server licenses because TM1s almighty administrator was entitled to look at all data otherwise. I very well realize that this doesn't matter to you if you already had everything you needed, but even then you may want to get a quote from IBM on buying new licensens. Even replacing the current licenses with new ones without buying anything extra is often cheaper in the long run.

Now your mileage may vary but I am living in a country where previously a single Applix partner had an exclusive distributor partnership with Applix. Maybe their prices were always inflated vs the prices the rest of the world had to pay but I do not get the impression that TM1 became more expensive over here - rather the contrary.

[Alan Kirk]

- Substantial upgrades to maintenance fees;

- The maintenance % went up yes - but TM1 on a whole became substantially cheaper and the licensing model became easier. Gone are the days when you had to pay extra to use TI, the web, 64 Bit. Gone are the days were you had to pay for several server licenses because TM1s almighty administrator was entitled to look at all data otherwise. I very well realize that this doesn't matter to you if you already had everything you needed, but even then you may want to get a quote from IBM on buying new licensens. Even replacing the current licenses with new ones without buying anything extra is often cheaper in the long run.

{Cough, choke...} You may perhaps want to explain that to my Iboglix rep who has cited me a 6 digit sum should we want to go to 64 bit.

Yes, granted it's "only" high 5 digits if we have only 4 cores on the machine, but you need to double it for any machine which has 8.

Now your mileage may vary

Not so much "mileage" as "light yearage", I think...

but I am living in a country where previously a single Applix partner had an exclusive distributor partnership with Applix. Maybe their prices were always inflated vs the prices the rest of the world had to pay but I do not get the impression that TM1 became more expensive over here - rather the contrary.

I've said it before and I'll say it again... the lack of a transparent pricing list means that the pricing policy is "milk the cow for as much as you can get out of it". See also "Goose, Golden".

[BigDSter]

Just had a renewal quote for a 10% increase which made us splutter out of biscuits!

We did have an agreement in our original contract that we inserted

"Maintenance renewal pricing is calculated at 18% of the then-current software component list price schedule at the time of maintenance renewal, but will not increase in any given year by more than a percentage equivalent to the cumulative increase (if any) shown by the CEL index published by Computer Economics Limited of Survey House, 51 Portland Road, Kingston Upon Thames, Surrey KT1 2SH England. This is a non-refundable service once purchased. The maintenance service includes new product releases and upgrades as they are available, as well as technical support. "

We wanted some pricing index linked to the computer industry, but because of the cumulative part of that we snapped their hand off when they came back with 5%.

Still seems like a lot and their justification of the original 10% (lots more support, lol) seemed a bit weak.

Has led us to look at the quote again, and in particular what we are paying for. Can someone translate as its not exactly clear what we are paying for

COGNOS TM1 LEGACY PROC ANNUAL SW S&S RNWL *4
COGNOS TM1 NON-PROD PROC ANNUAL SW S&S RNWL *2
COGNOS TM1 PERSPECTIVES AUTH USER ANNUAL SW S&S RNWL *3
COGNOS TM1 READ ONLY AUTH USER ANNUAL SW S&S RNWL *200
COGNOS TM1 READ/WRITE CONC USER ANNUAL SW S&S *13
COGNOS TM1 TURBO INTEGRATOR SERVER ANNUAL SW S&S RNWL *4
COGNOS TM1 WEB SERVER ANNUAL SW S&S RNWL *4

As far as I'm concerned we have 3 TM1 Servers, one of them being a development one,
200 Read Only Licenses split across the two Production servers.
TI on each of the 3 servers
5 Read/Write Licenses on each production server and 3 on the development server

Not sure what the perspectives bit is about, or why we have 4 servers/TI's or even
a web server which I'm sure we sold back to them years ago in return for some more
users

May just be me being dumb of course

[kpk]

IMHO:

COGNOS TM1 LEGACY PROC ANNUAL SW S&S RNWL *4 - These are your productive servers 4=2server*2processor
COGNOS TM1 NON-PROD PROC ANNUAL SW S&S RNWL *2 - These are your non productive servers 2=1server*2processor
COGNOS TM1 PERSPECTIVES AUTH USER ANNUAL SW S&S RNWL *3 - You need one administrator for each server
COGNOS TM1 READ ONLY AUTH USER ANNUAL SW S&S RNWL *200 - Your 100+100 read-only user licenses
COGNOS TM1 READ/WRITE CONC USER ANNUAL SW S&S *13 - Your 5+5+3 Read/write user licences
COGNOS TM1 TURBO INTEGRATOR SERVER ANNUAL SW S&S RNWL *4 - These are your productive TI licenses 4=2server*2processor
COGNOS TM1 WEB SERVER ANNUAL SW S&S RNWL *4 - These are your productive web server licenses 4=2server*2processor.

"As far as I'm concerned we have 3 TM1 Servers, one of them being a development one,
200 Read Only Licenses split across the two Production servers.
TI on each of the 3 servers
5 Read/Write Licenses on each production server and 3 on the development server"

[BigDSter]

Me being dumb then

Ta