1. Can you prove the ROI?

Once all is said and done, what will the return be to the organisation (in money terms)?

Before you invest in that tech; and those consultants; and all the hardware, can you come up with a water tight business case that will stand up to the Financial Directors scrutiny?  If not, look for another way to achieve the same thing. Could you run the automation as a service, using someone else’s equipment and licenses, without incurring all the risk? Work on proving the value of the use case before committing the organisation to the dollar-based licensing costs and developers.

 

2. Measure!

What will the key performance metrics be that measure the success of this project.

If you are automating a paper-based process and creating a digital workflow, how will you decide if it is successful or not?  Start with the why – why are you automating this?  What do you hope to achieve?  How will you measure that?

Take a measurement before you automate, so that when you need to report back the results you have “that was then, this is now” stats to show your success.  This measurement may have been taken when for the business case, but if you weren’t the one making these buying decisions and have just become the project owner, take measurements before implementation to show the improvement.

 

3. Don’t skimp on documentation

Make sure the as is / to be processes are well documented, and that the interactions between people and systems are clearly represented.  The inputs and deliverables need to be clearly defined and agreed so that there can be no unintentional scope creep or moving of the goal posts which will affect the ROI and KPI measurement at the end of the project.  Decide what will be tested and how, and formalise a test plan that is agreed and followed when the time comes. Don’t fall into the trap of making it up as you go along – he who fails to plan, plans to fail.

 

4. One bite at a time

Look for a way to rapidly deploy and measure the success in small increments.  We have all heard the horror stories of the mammoth project that got shelved after 36 months that was supposed to be completed in 12.  We suggest eating the elephant one bite at a time – automating small steps, “automation islands”, that can be linked up later, minimising your exposure and giving you the chance to take the off ramp if you decide the technology is wrong for you or the partner isn’t working out. To this end you need to choose the solution that gives you the most flexibility so that you have these options available to you.