Case studies  | Publicis Groupe

"We've found that the highlights of working with ServiceNow are their ability to answer questions and their ongoing support of our deployment of new functionality." – Bob Callaghan, Change and Configuration Manager at Publicis Groupe

This ServiceNow case study is based on an interview with Bob Callaghan, Change and Configuration Manager at Publicis Groupe.

Publicis Groupe is the world's fourth largest communications group, with nearly 45,000 employees in 750 agencies in 104 countries and 196 cities. It is ranked as the world's second largest media agency, and provides clients worldwide a complete range of advertising services through Leo Burnett, Publicis, Saatchi & Saatchi, Fallon and Bartle Bogle Hegarty. Publicis Groupe offers healthcare communications, corporate and financial communications, sustainability communications, shopper marketing, public relations, CRM and direct marketing, event and sports marketing, and multicultural communications. 

We're global and we typically grow by acquisition. Of course, the companies we acquire present a variety of processes and tools, so we usually go through a number of conversions every year.

The IT landscape at Publicis Groupe

We have about 675 IT employees spread around the world. We have adopted ITIL as our standard framework, and it's as widespread and diverse as the agencies within the company. We have adopted a shared services structure encompassing service entities like finance and IT, so each agency goes to a central entity for these services. The shared services organization breaks down into a series of regional organizations. For example, we deploy people worldwide who support standard tools for e-mail, the wide area network, file and print support, and of course, ServiceNow.

Looking for more flexibility and support for higher volume

We had four regions within the company with about 12 different tools deployed among these regions to support help desk and the routing of calls. We needed a common tool for passing tickets, incidents and events from one group to another so that we could take advantage of our reach across time zones. Our teams are geographically dispersed, from Hong Kong, Paris and Chicago, so with a dozen tools deployed across the globe, it was impossible to share work. 

Across North America, Publicis Groupe had deployed HP OpenView Service Desk 5.1. Europe had two or three different solutions, Latin America had four or five different solutions and our Asia Pacific region had another two or three. HP offers many of the basics to support ITIL and we found it to be consistent in the way that it allowed us to deal with certain processes. 

However, we were consistently frustrated when trying to customize HP Service Desk. It was just impossible to change the tool to support certain ways in which we run our business. We also saw occasional performance issues with HP Service Desk. So if we were going to deploy it globally and increase the load on it by a factor of three or four, we had serious doubts that it could handle the additional volume. 

In mid-2008 we began the task of finding a global tool and to adopt ITIL processes consistently worldwide. First, the company decided to establish several global organizations. I can't emphasize this enough – we took the time to study all of the processes, policies and procedures that support the various ITIL processes around the globe like change, incident, and release management. Then we took three months to rewrite all of those policies and procedures to a common set of documents. For instance, we whittled down the set for North America from 70 pages to six and got agreement on them from all four regions around the globe. 

We were then ready to look for a tool with which we could actually implement those policies and procedures worldwide. Executive management decided we would move off HP Service Desk.

Business requirements for the ITSM tool

We were committed to ITIL.so we encouraged all of our IT staff to go through ITIL training and a number of us have become ITIL-certified at various levels. So our number-one requirement was that the tool must support ITIL. Another requirement was global scalability and always on 24x7 availability because we have staff around the world. 

Finally, we were looking for a tool that would support foreign languages. English is our global standard but many of our employees prefer to enter tickets and issues in their native language.

Stump the ITSM vendor

We decided to bring in ServiceNow and put it through its paces. When evaluating a tool I like to play "stump the vendor." We grilled ServiceNow on the opportunities and functionality that we were looking for in a tool, and asked them, "Tell us how you solved this." Every time we thought we had a "gotcha" moment, ServiceNow were able to show us a solution. 

For example, the HP Service Desk had a fat-client that ran fine on a PC, but it never worked on a Macintosh. With all of the creative teams in our company, roughly 35-40% of our computers are Macs. So we asked ServiceNow, "Does it run on a Mac?" The first time they presented for us, they did the entire demonstration on a Mac. 

As another example, we use a common directory to get away from tracking things like user ID and passwords within the service management tool. We're a Lotus Notes shop, and there is an LDAP directory associated with Notes, so we challenged the ServiceNow to show us LDAP implementation for a single sign-on. Not only did they come in and show us how to do it, they helped us customize it, and it runs very well. 

We also did a proof of concept with group in North America and other regions asking them to stump the vendor. Honestly, it's the only time in my 30-year career a vendor has answered every question satisfactorily. I've been involved with a number of service desk tools like BMC Remedy and Service Desk Express, and Peregrine. But we realized ServiceNow provided everything that we needed and we couldn't find any gotchas. It was very unusual for us, but we decided that an evaluation population of one was good enough for us and we settled on ServiceNow. 

I would encourage other companies evaluating ITSM tools to make sure to ask all your "what if" questions. "What if we do this?" "Can the tool do this?" "Can it tie into this?" Make sure to get satisfactory answers. We've found the highlights of working with ServiceNow are in their ability to answer questions and their ongoing support of our deployment of new functionality.

The implementation experience

Implementation went very well for us. During our global rollout, we started in Europe and made both incident management and ServiceNow Discovery mandatory in order to populate the CMDB. We did that because with ITIL processes it's very important to be able to track what happens to every configuration item, whether hardware, software or a combination. So in every implementation in the 20 or so countries where we're currently live, we started with incident management and configuration management. 

Also, in every region we rolled out a self-service component with the ServiceNow service catalog. The most sophisticated service catalog implementation was for North America because the processes and procedures were so highly developed here. But we rolled out basic self-service capabilities in each of the regions. For the mature regions with available resources, we also rolled out change management. 

We had to be careful in many ways with how much customization we committed to, because ServiceNow allows a lot of flexibility on its out-of-the-box change management tool. In getting the tool to suit our global needs, it was so easy to customize things that we had to draw a line in the sand and say, "No more after this." 

That was the essence of our implementation roadmap. We started in March 2009 and we finished rolling out to 20 countries in eight months. All IT personnel worldwide now have access to ServiceNow, and half to two-thirds of our users have access to at least minimal self-service. So far we've deployed ServiceNow in French, German, Italian and Spanish and our employees in China have chosen English over the Chinese version of ServiceNow.

Resourcing

Three of us formed the core team here in Chicago. We named one project manager in Latin America, one in Europe and one in Asia Pacific. Then we personally trained them in incident management so they could train the trainers for the different countries and organizations in their region. 

For configuration management, I trained one person in each of the 20 countries in which we deployed. This meant a lot of Webex web conferences and phone calls at all hours to accommodate time differences. At last count our CMDB contained more than 210,000 configuration items.

Discovering all the assets

We had evaluated several discovery tools – offerings from HP and other vendors like Altiris – but they were always rather difficult to integrate to our existing tools. But the discovery tool in ServiceNow is very well integrated. You can write scripts and take advantage of routines to de-dupe the CMDB. There are flags in the system to decide what to do if, for instance, a laptop has multiple IP addresses. Do you want to record both? Do you only want to have one? 

The hard work is all up front: setting up Discovery to run the way that you want, deciding on ranges of IP addresses, devices and credentials. After that, it is very much hands-off. 

As we discover things like desktops, for instance, we can find out if they have standard builds. If a user calls for service, the service desk representative can enter the caller's name to the incident form and automatically examine the hardware and software installed on their PC. 

We can also have an entire history of all of the changes made to servers or to infrastructure items like routers and switches. In the future, we hope to take advantage of the baselining functionality within ServiceNow where we create a baseline and then periodically find all infrastructure changes in an interval. Then we can identify any unauthorized changes and the affected devices.

Selecting from the service catalog

To determine our service catalog items we've used ServiceNow Discovery to populate the CMDB and identify relationships that constitute a service. Depending on your role within the organization, we can assign an appropriate role within ServiceNow and make sure that you can see only the choices you need to see. 

For instance, for one of our divisions we have designated two or three individuals to make requests. Using the example of a new employee hire, they see a number of fields for employee name, email address, demographic information, supervisor, type of PC required, software requirements, Blackberry requirement, etc. The request then automatically spins off tasks and assigns them to the appropriate group within IT for fulfillment. 

We've built these for each brand in the Publicis Groupe family – there are about 100 in the U.S. alone – and each brand has its own specific look and feel. One of our goals in the near future is to try to consolidate so we can stay on top of so many service catalog items. Ideally we'd like to have a service catalog offering for each of our four regions.

Service level agreement workflows

For each of our catalog items we define a service level agreement to fulfill requests within "x" number of hours or "x" number of days. We have a simple workflow that tracks whether we've hit 50%, 75% or 100% of our SLA time. The workflow waits for another 25% and of course 25 and 50 are 75, which is why it would then create the SLA 75 warning, fire off another event and then wait for another 25% and if you hit a 100% and you haven't completed things already, then you've breached the SLA. 

We are contracted with each of our agencies for particular SLAs and evaluate our field staff on SLA observance. Depending on the SLA the agency selects and how quickly they want resolution, they pay more or less money. It's a true charge-back model.

Selling it to the users

We conducted live training sessions for incident, change and service catalog. The training differed according to the region and the maturity of the organization. In North America, we focused more on what was different from the legacy tool, whereas in Europe we talked more about the theory of change management and how to do it in ServiceNow. 

Every field in ServiceNow can have a little help icon, a little question mark beside it, and users can pop up context-sensitive help. Training, whether face to face, by video or by reading a manual, we provided all three.

Maintenance and upgrades

We've been through three or four ServiceNow upgrades. In 99.9% of the case, they have no impact upon our existing code. We have always found one little tweak, one little issue here and there with every upgrade. None of them have been totally perfect. But, you know, 99.9% isn't a bad grade, either.

On our ITSM horizon

Our plans are to roll out service catalog to the four regions over the next year and give it a common look and feel in all of them. We're going to deploy change management in regions that don't yet have it, notably Latin America. We're modifying the release management code that comes with ServiceNow and tying that into change management. We're also going to roll out problem management globally next year. I'd like to roll out content management, then put our own look and feel onto ServiceNow. 

Content management allows you to completely rearrange everything, very similar to what you do in myYahoo! or even iGoogle, and present a very familiar and useful user interface for employee self-help. It's like a consumerization of IT. We hope to have a common portal with a particular look and feel for employees to sign onto and get news about the company so that most users will not know whether they're in the portal or whether they're in ServiceNow.

Quantifiable Results

Many people have asked us about quantifiable results, but we don't have a true apples-to-apples comparison because we have so many more capabilities with ServiceNow than before. We've gone from about 12 different tools in support of our IT service management process to one tool: ServiceNow. 

The system is available and stable 24/7 around the globe; we can assign a task to any IT resource worldwide; we can put out a common process to define priority-one and priority-three incidents so that the data centers in Europe and Chicago understand each other. When you don't have a common tool worldwide, as we didn't, it's hard to compare. 

We have a globally deployed change process that we'll expand next year with one repository of all changes that get made to the infrastructure worldwide. Self-service is a huge win for everybody in the organization because users can go in and order things that they want and track that item within an automated process. We quantify these results using the reporting functionality in ServiceNow and people love to report on nuances and data sliced every which-way. Reporting is a huge deal in a large company like this. 

Finally, it's not a quantifiable result, but I can honestly say that the adoption of ServiceNow has made it fun for me to work again. I know it sounds a bit contrived, but I actually come to work and have fun now.