7183121469_765ce8ce4f

Client turnaround – at the edge of the Grand Canyon

We were in a crunch. The marketing platform implementation project had been vastly under-staffed. The client was yellow and in danger of going red. The client was very concerned and our management was in a panic. Our Executives had been summoned by the client. It was now a certified turnaround situation.
Prior to this project we had properly managed hundreds of engagements without a hitch.
How do we fix this mess?
Turn on a dime
The key elements of the turnaround were put in place in a one week period. Taking a quick look at the synopsis for that week is an excellent way to learn from the experience.
We had all been working remote to this point and we setup a 5 day onsite to get us all together and put the project on the right course.
Monday
Go-live was in 14 days. There were 10 of us in a room trying to move many activities forward in a distributed and disorganized way (note it was 8 client team members and 2 consultants in the room). The client was responsible for project management at this point and it was not happening. I was one of the consultants and it was not clear to me what needed to be done and by when and when I asked several people the answers were not the same.
We tried to break the room into a few teams to progress in parallel but this was not working because there were constant interruptions from one person to another. The breaking point was being interrupted and asked a question that was not even critical to go-live. So my thought was how can we get to go-live if we are not even clear what that means? And at the same time how can we work more efficiently?
Tuesday
On Tuesday we gathered the troops, projected a list on the wall, and confirmed that those were the exact items that were required for go-live. From that moment on we were tracking progress on the key priorities that were critical for initial go-live and we had a goal for each of those items for the end of the week. So the 2 roles brought to bear at this moment were getting the room under control and clarifying priorities for the week. This role was taken on by one of the senior consultants.
After hours the 2 consultants got together to review the complete laundry list of what needed to be done overall and it amounted to hundreds of hours over the next few weeks. The staffing mathematics did not work out and this needed to be resolved.
Wednesday
We started the day with a review of the key go-live items and where they stood. Then the next breakthrough was splitting into 2-3 teams and making all but one of those teams leave the room. Monday we were basically all in one room and it was not possible to work efficiently and by Wednesday we were able to.
To solve the staffing mathematics we stripped anything that made sense off the core team and gave it to freelance teams. In a period of 48 hours we stripped and handed over 150 hours of work to three different freelance teams. The roles I want to highlight here is that of gathering the resource and packaging and stripping the tasks.
Even with all the additional help, there were still bottlenecks in the core team – specifically build management. One of the senior consultants was handling build management and this person was one of the builders as well. Even all the stuff the contractors were building had to be reviewed by build management. So we had a bottleneck in build management. I would like to highlight the roles of build management, build and build review here.
On this week we dedicated one of the senior consultants to creating a formalized test program. We were responsible for everything performing correctly at go-live and the only way to make this possible was to run multiple rounds of testing on everything – this required a formalized test program. See my blog post on testing to better understand this:
http://www.marketo.com/cheat-sheets/when-boring-means-amazing-how-testing-makes-go-live-day-a-snooze/!
The role I want to highlight here is managing a formal test program to draw out all the issues prior to go-live so they could be fixed prior to go-live.
By the end of Wednesday we had 2 challenges:
1. We had stripped everything we could from the core team and given that to the external resources. Nothing more could be stripped to resources that were not in the room. We needed more in-room resource.
2. We were still not communicating well with our executives and the client executives
By the end of Wednesday I felt there was more turnaround to be done but we were heading to likely success.
Thursday
Thursday we got that in-room resource. The engagement manager joined onsite and brought 2 additional consultants that he had marshalled.
From this day onward the build management role was completely covered. We now had 2 consultants focused on build management. And build was also in great shape. We had 5 consultants building now (including the freelancers).
With the engagement manager onsite communication with the client at the senior levels improved. There were a lot of issues flying that day and the engagement manager also helped us prioritize which were really critical and ensured those got communicated and pushed forward. The role I want to highlight here is client management, and a piece of project management dealing with prioritizing issues and driving them forward. Also note the acquisition of resources role (grabbing the 2 consultants) again.
And the engagement manager started sending daily communications to our executives so they knew what was happening.
By the end of Thursday I felt like the project could not fail.
Friday
Friday was dominated by our first round of formal testing. As we ran through the test scripts we found many issues which needed to be resolved over the coming days. Note this is expected with testing – you will need to perform multiple rounds of tests to get each program and asset to the point where it is functioning exactly as expected.
We had a constant need for technical help primarily because the client had heavily customized javascript and any small change to a page or a form caused something to break. This complexity also created technical hurdles in getting everything to work together on the same page for the first time. Basically each round of testing drove the need for additional technical fixes. Our technical resource had to handle each issue quickly including off hours and I am sure other work was impacted due to the rapid-turnaround requirement – we could not wait long for the fixes because if each fix cycle slowed us down for a day we would not make it. The role I want to highlight is rapid-turnaround technical resource. The senior consultants managing the build had to supervise this as well.
By Friday the project was running like a regular project. The portfolio of key roles were all accounted for. We had enough staff. From the client perspective we were green at this point. From our internal perspective we kept the project yellow to keep us on our toes.
Go-live was in 10 days.
Looking back
In my opinion the issue was not simply about more bodies. It was about getting the proper structure and roles in place. It doesn’t really matter who takes the roles and with a team you will have a natural distribution of roles that people feel comfortable with.
In the end we did successfully go live as planned. When asked if the project was successful I would not hold up this project as a model of success. But as turnarounds go it was a successful turnaround!

adminClient turnaround – at the edge of the Grand Canyon

Leave a Comment