Ingredients for the perfect failure of your enterprise IT project

• 1 utopian cost and time estimate
  An imaginative budget of, say, €20 million and 24 months to modernize a core banking system — silence the few experts who want to know that this is nonsense, right now that AI is getting better and better, it can't be a problem
  ‍
• 1 Hydra as client
  Preferably a team with representatives from all key business areas with conflicting priorities — cutting costs, ensuring compliance and data protection, and innovating at the same time. Without a clear direction, of course, because everything is really important
  ‍
• 1 generous pinch of methodological and certification fanaticism
  Choose Scrum, SAFe or Waterfall and religiously believe that this will solve your problems. A project manager who has no deep understanding of IT, architecture or technical connections, but who can demonstrate all sorts of certifications, can work wonders here
  ‍
• Vague or infinite deadlines
  Preferably no real content milestones in a production environment for as long as possible — implemented story points and velocity indicators are completely sufficient
  ‍
• 1-5 Weak Product Owner
  A puppet without authority who is afraid of making quick decisions and forwards every question to different teams or experts, it is of course even better to install several product owners who, if possible, do not follow a common goal
  ‍
• 1 overhyped external partner
  Choose a fancy consulting company that promises 100 Spartans but delivers a colorful team of beginners led by 2 veterans.
  ‍
• 1 development team with no clueless content
  Ideal when developers don't understand the goal — even better if they've never seen a core banking system or never heard of actual customer and compliance processes
  ‍
• Restricted data access
  Block access to real legacy data, no one is familiar with the old Cobol, CICS, VSAM, DL1 or DB2 junk anyway — instead, there are synthetic test cases and rough user stories that hide the true complexity right to the end, and you can also elegantly avoid tedious discussions with the security department
  ‍

Preparation

1. ‍Start with a fantasy plan
   Start with an absurd estimate: 20 million and 24 months to modernize the old core banking system.
   
   Build timelines and budgets on them and ignore the few voices who know that this is an illusion.
   
   What if experts warn? Suffocate with optimism or bureaucracy.
   ‍
2. ‍Sending the team into chaos
   Give developers the mandate to “modernize” without explaining what this means for SWIFT, credit processes or regulatory requirements.
   ‍
   ‍Let them be iterative.
   ‍
3. ‍Paralyze the product owner
   ‍Use someone who answers every question with “I'll sort this out with the appropriate specialist department.”
   
   Let the decision making simmer until everyone goes in circles or the whole topic is forgotten
   ‍
4. Deny data access — Security and GDPR First
   Block access to real legacy data
   
   Feed the team synthetic test cases that hide any issues.
   
   It's only in the production environment that it really crashes
   ‍
5. ‍Eliminate deadlines
   ‍Focus exclusively on story points, velocity and burndown charts for as long as possible
   
   Stay away from technical, professional arguments that could spread bad vibes here — content is overrated
   ‍
6. Let the certificate hero do
   Let go of the project manager based on fantasy planning, crammed full of certificates, but with no idea of the actual scope or banking, legacy systems or software development, he can do real wonders
   
   ‍Outcome: Process beats reality.
   ‍
7. Unleash the cult of methods
   ‍ Enforce strict Scrum, SAFe, or Waterfall, regardless of whether it fits. Make meetings and methodologies more important than progress.
   
   ‍If everyone here does everything correctly & correctly, everything will certainly be fine in the end!
   ‍
8. ‍Cultivate bottleneck blindness
   Make sure deployments don't become too easy due to compliance, checklists, audits, and sign-off hurdles
   
   Ignore as long as possible that you will ultimately need a brute force approach with hundreds of deployments due to “high-level user stories,” old legacy data and “hidden” requirements
   
   ‍Two-week cycles for deployments to an integration environment are perfect!
   ‍
9. Unleash the client Hydra
   ‍Especially now that there is finally another major organizational project, all areas, departments and staff units should really be able to submit their wishes and requirements on an equal footing
   
   Each department has collected requirements within the department, compliance requires audits, Finance saves on developer notebooks, IT operations insists on all old and new checklists, data protection sees the GDPR gold medal within reach
   
   ‍Trust all this to your super-certified, content-free project manager — he knows what to do with it
   ‍
10. ‍Use the Spartan trick
    Get the highly acclaimed external partner who can supposedly solve your problem
    
    His few, really expensive experts disappear quickly, while the junior troupe takes over with cheaper hourly rates
    
    Lean back and relax when you see that project managers and implementation partners are in good harmony with the progress reports — soon it will be Point of no return achieved - once you've spent 15 million and no one could actually know that everything has somehow become much more complicated, you stick with it — and end up accepting 40 million
    ‍
11. Bake until burnt
    ‍ Push the project into a high-pressure enterprise oven — ideally with a banking conference where the board sits on the podium as deadline pressure.
    
    Ignore the stench of skyrocketing milestones, cancellations, and tripling costs.
    
    Bake until a real disaster comes out, which also becomes too much for the steering committee

Serving suggestion

Present the failed project with a fancy PowerPoint slide full of excuses:

“Unpredictable complexity”
“Issues with external partners”
“Scope Creep”

‍As a side dish: A bitter glass of disappointment — no one is surprised, but everyone is frustrated because everyone has done everything right in their field.

Finally, a small selection of failed IT projects in recent years in the areas of banking, trade, industry and administration, so that you can see that you are not alone: