Do I leave a failing project?

**LondonManc** · 1 March 2017, 15:21

You're the Head of IT?

Who is your representative at board level/hiring manager?

**malvolio** · 1 March 2017, 15:39

CYA time. IF you know how to sort it and they don't, or won't let you, give them a costed and timed plan that will work and offer to either implement that or leave. At least then your CV either has a success on it, or the production of a detailed recovery plan...

I've been there before*. You either bully your client into doing it your way or you leave them to it. They bought your expertise, if they're too dim to utilise it you have a loyalty to nobody other than yourself.

* as in two months to move a 300Tb datastore from one DR site to another with no direct connection, no accurate mapping of the existing datastore (including some 200 virtual C: drives...) and no compatibility between old and new architectures. That one I delivered - just...

**MrMarkyMark** · 1 March 2017, 15:42

Originally posted by malvolio View Post

CYA time. IF you know how to sort it and they don't, or won't let you, give them a costed and timed plan that will work and offer to either implement that or leave. At least then your CV either has a success on it, or the production of a detailed recovery plan...

I've been there before. You either bully your client into doing it your way or you leave them to it. They bought your expertise, if they're too dim to utilise it you have a loyalty to nobody other than yourself.

Absolutely this 100%.

Although I would also build some risk based stuff in as well. i.e. discussing the options and what they risk if they leave certain key steps out.

At your level this should be childs play, no?

**GB9** · 1 March 2017, 16:11

In situations like this I get to the point at which I no longer care if I'm there on not. Once that's reached I tell them what needs doing, often quite bluntly. They either let me get on or I leave.

If you feel like leaving, as opposed to just peed off, then say your piece and see what happens. If you get kicked out you've lost nothing.

**LondonManc** · 1 March 2017, 16:12

Originally posted by malvolio View Post

CYA time. IF you know how to sort it and they don't, or won't let you, give them a costed and timed plan that will work and offer to either implement that or leave. At least then your CV either has a success on it, or the production of a detailed recovery plan...

I've been there before*. You either bully your client into doing it your way or you leave them to it. They bought your expertise, if they're too dim to utilise it you have a loyalty to nobody other than yourself.

* as in two months to move a 300Tb datastore from one DR site to another with no direct connection, no accurate mapping of the existing datastore (including some 200 virtual C: drives...) and no compatibility between old and new architectures. That one I delivered - just...

Originally posted by GB9 View Post

In situations like this I get to the point at which I no longer care if I'm there on not. Once that's reached I tell them what needs doing, often quite bluntly. They either let me get on or I leave.

If you feel like leaving, as opposed to just peed off, then say your piece and see what happens. If you get kicked out you've lost nothing.

These two go together nicely.

In the position as interim Head of IT, you need to be firm and speak to your person on the board - CTO, or whoever it may be. Explain what you believe needs to be done and the cost of that. Explain the cost of not doing it.

**Bluenose** · 1 March 2017, 16:17

Originally posted by crawding View Post

I am looking for some advice really, I am currently in a contract as Head of IT in charge of a data centre migration project. I am having continuing problems with senior management due to their inexperience at the senior level and they are failing the project. I have made my recommendations but I don't think these recommendations are getting up to the board as they don't fit with the current management stance of nodding and agreeing.

these types of projects can still turn out OK as long as the correct action is taken early on.

First off, one of the killer aspects for businesses saying 'lets migrate it quickly' is that businesses can rarely facilitate the support and testing required that the IT migration team need, what you will find is that as your project team engages with the business and the schedule is created, the business appetite will soften. This is because major business events such as software releases, bank holiday maintenance, SME holidays for key resources, 'un-movable business events' yadda yadda will extend the timeline, I guarantee it.

In a larger business the Group PMO and Group Risk will audit the project to make sure the business are not trying to pull a fast one. If you don't have a PMO or risk team keeping things on a leash suggest you have an external 3rd party to offer impartial advice.

This is especially effective (in my experience) in a law firm or Insurance company as they will have business side experience of using such advice and its value.

Get engaged early.

For me, thats the key thing, make sure the business is informed correctly and given the risks to accept, its not for I.T to accept all the risk. If they are informed correctly and still wish to proceed then its their call. If you get thrown under the bus then really, there is not much else you could have done.

Set yourselves up correctly. I would not worry about zero budget and tech challenges as long as you (or your Project manager lead) pins actions, risks and issues onto the correct owners in a formal fashion and document it.

Make sure you have a project manager (1) a technical lead (2) and a schedule manager (3). If all these roles are you then highlight the risks of doing so (head of IT should be doing other things) and get the business to accept that risk that things will run slower.

Create an project exec board running for the project asap. If the business are on the board, do you think they will allow you to start migrations if the risk is too high? not in my experience.

mobilise well.

Get the schedule manager talking to the business directly via assigned contacts for each of the affected business areas. Get your schedule manager to create a draft schedule and get the business to sign off on the migration ordering for the apps.

what I have found is, once the risks and issues are laid out and they (the business) are made accountable, the tune starts to change.

What does this mean? Get some risks and issues laid out, scored and prioritised. Run an internal face-to-face meeting with your programme sponsor, not the business and agree which high-level ones need to be assigned and where.

Then run a risks and issues workshop only with people who will and are able to take action. Run two workshops if you need to, one technical and one business led.

How many applications are running on the 212 servers? If its one app then purely from a logistics perspective, it could be easy to accomplish if your migration team is big enough and the business can do all the testing in time. It could simply be an IP address change.

Do past DR tests for the 212 servers go well? i.e. are they easy to move around if they are then this is a good sign that the apps are not fragile and migrating them to a new DC will be easier.

does the 45 days include the application complexity assessments for the apps which run on the 212 servers? If it does then you need to place the ball back with the business by giving them lots of questions and asking them to turn them around in 7 days. From experience they won't be able to do this.

what is the size of the migration team? 212 servers is doable in a short space of time as migrating virtual servers is fairly straight-forward in 45 days.

get one of your storage guys to check the change rate of the data in terms of daily, weekly and monthly. This can be done by looking at the backup reports of the data. This gives you an indicator of churn rate. This will give you a business case for an uplift on the 100mbps line if you can calculate that the churn rate of the data is over a particular level.

If the physical box is an AIX server or some sort of customised IBM/SUN piece of kit, expect and plan for a specialist consultancy to help you plan the move, do not attempt to do this in house, even if your tech specialists are good as it de-risks the move.

**Abraham** · 1 March 2017, 16:37

It is amazing that the Senior Management "always" listen to those who are determined to fail the project. I have seen this happening in few projects, where the "politicians" bring in their favouites, who in turn work towards the goal of failing the project.
In this situation, the contractor is there to perform the role of scapegoat, to own the responsibility for the impending disaster.

**malvolio** · 1 March 2017, 17:09

Originally posted by Abraham View Post

It is amazing that the Senior Management "always" listen to those who are determined to fail the project. I have seen this happening in few projects, where the "politicians" bring in their favouites, who in turn work towards the goal of failing the project.
In this situation, the contractor is there to perform the role of scapegoat, to own the responsibility for the impending disaster.

Yep, that was my subtext. You defeat them at source by (a) providing them with a plan that shows why their "approach" isn't going to work and (b) not being there when they refuse to implement it.

Or you deliver what's needed and make the senior heroes look stupid - not that you care because you will have gone by then.

**MrMarkyMark** · 1 March 2017, 18:12

The current senior management are allowing the business to lead the migration and due to the inexperience they have allowed the business to set the time-frame, budget and changing the scope on a daily basis.

Another angle would be to reach out to the business.
Explain a few home truths, surely you have been involved with such projects before?

Using your past experiences and scenarios can often help sharpen their minds.
The inevitable fall out that could come, by trying to do too much too fast, sometimes even alluding to jobs being lost at past clients etc.

Do I leave a failing project?