Thursday, December 10, 2009

What is the $$$ cost of migrating Notes applications?

I've just started talking to a customer who is looking at migrating away from their R6.5 Notes environment with the help of their current Business Partner. They have the usual half dozen CRM + Helpdesk + Asset Manager apps and " ...will migrate them when the email migration is finished."

OK, we've all heard it before, but what I need is a White Paper which exposes the pain and the cost of rewriting Notes applications in other environments. Does anyone have anything beyond anecdotal evidence which can be presented to a CIO?
.

14 comments:

David Price said...

Graham,

In these cases I ask customers if they have the money to update the UI of the existing applications with 8.5.1, XPages, etc.. Often the answer is no. I then ask how they will find the money to migrate if they don't have the money to refresh/update.

Keith Brooks said...

No White paper I know of but he can talk to a CIO of a company I know that spent so much money, to migrate one app and failed that they won't do it again. Skype me to discuss it.

David, your perspective may be wrong. They have the money to move, not to stay, your question will just get you fired.

What you need to do is remind them that the apps MUST be migrated first. Why? Because usually they inherently use doclinks which would need to be recoded to work properly as URL links for gmail or outlook.

They never think about how the apps work or why, once that is raised, the migration takes on a very different approach, again, sometimes horribly wrong as well(ok, then screw it toss it all) or goes the way David proposes which is to upgrade the apps. But remember, they had budgeted BIG/HUGE money to migrate, so the costs to update/upgrade Domino should be easily covered, but don't do it ont he cheap is all I am saying.

Pierre said...

We have a client where we are in charge on the evolution of the Lotus Notes Apps. At this place it used to be an Microsoft Access Application that is now rewritten in Java with Oracle Back-end.

This Access Db is the size of a commond Lotus Notes Apps.

This Acces Db cost around 75 to 100 days to develop and is maintained by one part-time developer.

Now the actual development (only) cost of the rewriting to Java with Oracle is over 1500 days.

They have more than 30 running Lotus Notes Apps over there. If they want to migrate to Java, just do the math...

This is a real case btw. Hope this helps.

Ian Randall said...

I have a client that is a huge multi-national with a globally recognised brand. This multi-billion $ organisation has over 80 major locations in the US, and another 60 major locations spread across Australia, the UK and New Zealand.

They migrated from Lotus Notes mail to Outlook/Exchange over 6 years ago and have also installed SharePoint for collaborative applications (mainly used for web based file sharing). They are also an Oracle shop.

They have one major Lotus Notes based application that is still operating in-house. They have been attempting to migrate this one application to other 'strategic' platforms for over 6 years, and they recently needed to upgrade their Domino servers from 5 to 8.5 because they needed vendor support from IBM.

I don't have an estimate of the costs currently incurred for this one migration effort, but it is unlikely that they would have recently upgraded their Domino licence if the migration was due to be completed any time in the near future.

They also run one of our Lotus Notes applications on an outsourced Domino server, which I understand is their only globally accessible application that staff in all sites from around the world regularly access via Web browsers, along with a handfull of Notes Clients used by power users (who run Notes clients in a Remote Desktop Client environment).

Sorry, that I can't mention names or quote dollars at this time, but this example proves that sometimes migrations from Lotus Notes can be a huge investment of both time and resources.

If Lotus Notes was such a crappy development environment and other system platforms were are really so much better, they why does it take over half a decade and costs a huge investment to migrate one Lotus Notes application? I also understand that it didn't take much effort or time to develop this application in Lotus Notes in the first place.

In case you are wondering, this one "legacy" application is what this organisation currently uses to generate the majority of their revenue, so simply turning it off is not an option.

Thomas Schulte said...

Hi Graham.

I have got another example here where a pure Notes application was evaluated to be moved into a SQL Database backend, JBOSS type of application.

We had a consultant here who tried for half of a year to get around the functions built into that database in the eight years before this attempt was made. The evaluation was closed without a conclusion when the estimated costs for rebuilding the system reached the million euro border.
And the consultants evaluation costs reached the 100.000 euro line.

That was five years ago and the system has been developed further. Guess what the numbers would perhaps be today?

And yes this is a mission critical application that is developed and maintained part time.

Henning Heinz said...

Interesting comments. If you want to migrate applications away from Domino you have to hire developers that have an in-depth Domino knowledge. Many companies hire people that are only experts on the new platform that then try to redevelop applications. This often fails badly and cost a lot of money.

But your question is not simple. Much of the $$ depend on the target platform and if it is about a web or a Notes client application (or both).
Often you do not know how many days it took to develop the original Notes application and it's overall complexity (and hooks into other applications).

Finally trying to answer your question. I would try to guess the development time of the original application, add 50% for a migration scenario (as long as the customer does not say that they can start from scratch) and another 50% because I can. And for sure there are applications where I would suggest to keep them because it is not worth the effort.
As external billable hours sometimes cost more than internal resources the bill can become quite high. Pure domino web applications may cost less.

By the way
"They have the money to move, not to stay, your question will just get you fired."
This is a real gem and so true.
Unfortunately it also means that budgets for migrations often are not the biggest problem.

Wayne said...

I have a general question for everyone who responded with a story to tell. Are these migrations away from Notes Client apps or Domino Apps?
The reason I ask is that while IBM/Lotus have certainly made headway with XPages and Composite Apps, they have neglected the client side of things when it comes to user interaction.
I really believe that much of the perception of Notes being old stems from this.

Darren Duke said...

Wouldn't it be more pertinent to move email last? I agree with Keith, Doclinks, etc. Users don't realize until the email is moved HOW MUCH they relied on email for workflow. The customer would/should not want to jeopardize key business critical applications and workflow at the beginning of a project of unknown length (and it is unknown and the customer should make the other BP do it fixed bid and hold 50% until complete, watch them squeal on that hook).

Yes, you can co-exist but that can be difficult and costly.

Darren Duke said...

Oh, and the probable reason there are no white papers it more than lightly due to no migrations ever really being completed. See the comments above for everyone......

It is pretty common to start, very uncommon to complete.

Nathan T. Freeman said...

One could read between the lines on this post...

http://www.accenture.com/Global/Accenture_Blogs/Accenture_Blog_for_Internal_IT/Our+Migration+from+Lotus+Notes+to+Microsoft.htm

Anonymous said...

Let's look around. ExxonMobil is in the process of switching, after tens of millions spent, they gave up and are keeping Notes and using Exchange mail. Shell, which switched "Email" about 1997, still has Notes and it's been growing! They never got rid of their Notes Apps. A Toyota dealer in Texas, switched about 1998, still has Notes applications! There seems to be a trend here.

Stephan H. Wissel said...

You can do the math relatively easily. Let us for one moment presume, that the target environment is an as efficient platform as Domino, so you don't need to write *more* code. You just need to find out how many lines of code are inside the Notes application. To make things simpler lets count forms, fields, columns etc. as line of code. You get a number which you then can feed into a COCOMO II analysis here: http://www.cms4site.ru/utility.php?utility=cocomoii
The DXLMagic tool helps you to get the line-of-code equivalent from a Notes/Domino app

Stephan H. Wissel said...

That little tool will then give you the number of hours and cost. To reflect reality you have 2 axis: complexity of the application and experience of your team.
Since you have to do some data migration your project is probably on the higher complexity.

Benoit Dubuc said...

I have done a conversion project that is almost completed by now (over a year). You can check on my blog, I have a series of articles that relate my story (links are located top left).

If you have questions, I'll be glad to answer them!!! Just send me an email.