Panelist: GDC Online/Austin 2009: Leveraging Customer Service Quality Through Tools


I participated as an expert panelist at the 2009 GDC Online convention in Austin, Texas.

The panel was moderated by Gordon Walton (BioWare), and I was joined by John Erskine (NCSoft), and Phil Dean (Cryptic Studios).

The subject was:

Leveraging Customer Service Quality Through Tools

Overview: All customer service personnel want to help our customers, and tools are their enabler. This panel will review the state-of-art learning on what tools enable higher productivity along with generating higher customer satisfaction. Each of the panelists will bring real world examples and lessons learned in building and extending customer service toolsets for online games.

From Article:

Gordon Walton (BioWare) – Moderator
John Erskine (NCsoft) – handles all support related functions for US and Europe
Phil Dean (Cryptic Studios) – in charge of all billing, in-game support, closed and open beta for Chamions Online
Jason (BioWare) – C/S director for BioWare Austin

What’s important about tools?
John: Tools make difference between being “nice” to people who ask questions, apologizing etc… versus being able to say “we can see what happened here, we can fix this, you can keep playing like you were before this happened”. Difference between being polite and fixing problem in game, forever.

Phil: Cryptic’s focus has been on internal tools for C/S and tech… no external apps… our in game ticketing system took 18 months to develop. Main focus is helping players solve their issues themselves first, versus having them send in an issue.

Jason: C/S is a people business, but what those people are has shifted over time. Anyone can run into battle but the better your weapons the greater your success. You can hire great C/S people, but if you don’t have good tools you’re damaged. Before people always wanted to speak to someone, now people expect that there won’t be someone there and they should be able to solve their issues themselves. Tools are the class higher sword you are looking for.

What’s the #1 vendor you’d recommend for C/S?
John: We use vendors for all our ticketing systems and we’d recommend them – we are happy with their product. It is relatively abstracted from our game data – which is why it works for us.

Phil: Used Primus – they build an internal knowledge base for our C/S agents to use. By using this internal knowledge base, it helps the C/S team provide the right answer every single time. It is a living breathing doc updated by the C/S agents themselves, every time they have an interaction with a customer.

Jason: I am from the old AOL world. I am hard-pressed to recommend a specific vendor. AOL’s tools grew beyond their expectation and they weren’t able to scale fast enough. At AOL, the average agent took 15 minutes to log in to all the different tools they needed to use that day. So I believe it’s good to find the right tools.

What’s the biggest productivity gain you’ve gotten out of a tool in the last 5 years?
Jason: When we gave people the option to chat immediately with our AOL service, we deflected a lot of C/S calls.

Phil: Two examples. From gaming side, we had a huge gain in deflection of C/S issues by building in self supportability. You’re a player, you have a quest, you complete it but item doesn’t drop. Is it because I am new? I don’t know. By using in game ticketing system, you can view all tix and see if anyone has the same issue as you. Very frustrating not to know if it is just you or other people are having the same issue. In game support people can then put note in there if it is a known issue. Outside of game industry, re: phishing with PayPal, we built a robust internal tool that would allow us to get a phishing site down within 8 hours as opposed to a week. App we built was very automated and would alert us and track offending sites. Decreases productivity costs for C/S agents and increases customer satisfaction.

John: We syndicated our self serve content so it was available in Google, etc. The first place you usually go for help is Google, not necessarily our NCsoft site. So we make sure all our knowledge base issues are searchable via Google. The goal is for our content to be there when someone looks for it. In the last year, we’re using Twitter to push messages out to our playerbase about game updates and outages, we’re monitoring Twitter and Facebook updates re: NCsoft and then we can proactively contact those customers if we choose to, then hopefully make amends for the issue. We find that is a very efficient way for us to handle those issues… we’re contacting people early and resolving these issues before they become larger.

How many people do we devote to social network monitoring?
John: We don’t devote any more resources than we already have. Our existing C/S and community team handles it. On the GM side, we’re utilizing some tools that monitor Twitter feeds, etc for key words. It’s all part of our team’s day.

Which tools are you using to monitor Twitter?
John: The names escape me, but there are a lot of tools that do the trick.

Phil: Re Twitter, it’s a good thing to let players get the word out and tweet (he said “twitter”, but I changed it) from within the game. Free PR.

What’s the tool you are trying to build now or wish you had?

Phil: The ability to automate the reports coming in from people re: griefing, etc. One thing we are looking to do now is that if a player gets reported for griefing too many times, then a bouncer comes out and beats up the offender. But that could be exploited. Realistically, if you had a tool that would do more analysis on the reports coming in to weight them according to where they are coming from, their frequency, etc.

John: I have been fortunate to be at NCsoft for many years, so there is not one frontier we haven’t yet tackled. But I now have a dedicated tools prog who reports to me in terms of schedule time – allowing me to manage what he works on. We can tackle some of these little things piece by piece. Surprising how big a difference these little things can make.

Jason: A universal interface for your C/S tools system is pretty critical. #1 reason people stayed with AOL was ease of use – they made it very simple. #2 reason ppl stayed was for the service they received. Also, fraud is a critical issue – the more robust our anti-fraud tools, the easier C/S’s life is.

What are the most difficult things to handle from a C/S perspective?
Phil: Currently our biggest issue outside of griefing is key issues – people bought the copy somewhere and have trouble activating. One positive thing that was put in place that takes up a lot of C/S time is the number of key commands they have to remember. All our C/S agents work through the backend, not in-game, so we can get people items, move them, etc from our CRM system without having to log into the game.

Jason: From my past life with AOL, I was driven crazy by their reactive approach to retention. Re: data mining, the more you get the more you can strategize. If you can be more proactive, more people will stay in your game as opposed to “You’re ready to leave now? Let me try to keep you” – need to get to the customer before they are thinking of leaving.

John: There are a lot of pain points that can’t be worked around with tools, but the biggest category for pain points for us are third party transactions – people selling items outside of the game. Ultimately problems with those transactions come to roost with NCsoft.

With CoH’s mission architect system, what lessons have we learned from that system?
John: Paragon Studios would be better to answer in-game issues. In terms of C/S issues, it puts a bright light on the fact that any system that is player driven and player moderated becomes part of the game as well. Players will game that system as they would any other system. If there is any benefit that could come from gaming the system, they will exploit it to do so. Lineage: right before a castle siege, the players will gang up and report a guild to get them in a penalty box for a while.

Would problems with player-to-player transactions, would they be mitigated by using a Playspan-like service?
John: No. Where there is a dollar to be made – i.e. by undercutting a sanctioned player to player service or by offering goods not sanctioned for sale – people will move in to exploit that niche.

Bugs versus C/S Issues Question
Phil: Workflow… for each shift we have a roving C/S person who looks at newest tickets coming in to see patterns coming in. The rest of the team is looking at oldest tickets first. So the person looking at new tickets can tell right away if something has just broken. On Champions, as it is very new, we have a lot of items being reported as bugs because people don’t know how the game works – i.e. many of those bugs are as designed. Whereas #1 issue on in-game support side for City of Heroes is IP violations – people creating Incredible Hulk.

How do you get around tools dev being put aside when game dev takes precedence?
Jason: You start early. And you try to think ahead and allocate the resources necessary to dedicate to those tools full time. The rest of the org needs to know how important it is that those tools exist when you go live.

Phil: Closed Beta and Open Beta are very important on the C/S side because it helps us rehearse and get our tools and workflow down. Put a cost associated with the top 10 issues coming in… if you can put a cost against what it takes to resolve these issues, you can argue for them to be resolved by the dev team. Also, really engage the dev team to get involved with C/S… we have a Telesupport Listing Session.. designers come in to sit with C/S team to see what issues we are having on a day to day basis.

John: Important component to having tools at launch for new game is to carefully design the tools you are asking for. Prioritize those tools in a way that makes good sense to everybody on the team. I have seen very frequently someone come back with 30 pages of tools they’d like. If you turn that over to dev team, they will ignore everything. If you can come back to dev team and tell them you need these 3 things before Beta, then these 4 during Beta, then these 3 after launch, etc that will get done. We have to be more astute with what we ask for from the dev team.

Gordon: From the perspective of the people funding these games, tools are harder to justify as the best C/S in the world won’t drag a bad game up from being a good one. So the financiers are often mostly concerned with what the prime mover of what will make a great game. Rigor around prioritization and communication of cost benefit analysis is crucial. It’s up to you to show you those CBAs before launch – when it’s much easier for them to see the issue as by then they’re paying for the mistake of not making the tool.

In a free-to-play world, are there some issues that are too expensive to resolve? Could you charge a premium to resolve these?
John: In general there are some things that get too messy for us to figure out. They are usually related to person to person issues – she took my account, etc. Actually more common than you would believe that two people can provide all the identifying info for 1 account. There are no tools we can develop to fix those. Account hacks, etc we can fix pretty easily though. In terms of the premium support, there may be a market for that sort of thing, but only if you are offering support that is truly premium. Fixing something that is very broken in a game won’t win you any fans if you charge to do it. But a concierge service or something premium might.

John: We look for people that are way off the curve in terms of they call way too much, we call them and ask them what’s going on. Often when you break that anonymity barrier, they stop calling as much.

Has there been any studies on what bad C/S costs you in terms of actual metrics?
Gordon: Problem is translating soft into hard – can’t say “if this agent talks to this person, then the person will play 1 week longer”

Jason: We worked hard at AOL to do that sort of thing. Europe did it very well. My old title was Director of Customer Experience… experience is proactive, support is reactive. We were able to say that if we did X, player will stay for Y. We had those numbers and those were very powerful.

Phil: Satisfaction scores play a key part in player base retention. We want to do transaction surveys. Players get a survey occasionally… focused on the actual issue a person had. We don’t ignore them. If players gave us a 3 or below on a 1-10 survey, we reach out to that person personally… call or email to figure out why. When we reach out to these people, 9 times out of 10 we find out that people who are leaving the game are not leaving due to a C/S issue. Often you do surveys and never hear back – if you reach out personally, perhaps you will continue playing the game because you believe they care about you.

Gordon: Often if people leave over a C/S issue, it’s a last straw issue – so it was an accumulation of issues up to that point – not that particular issue. C/S gets a lot of misdirected blame. If you look at cancellation surveys, people give a reason for leaving that is usually not the real reason. Takes talking to the person to figure it out.

One Last Point:

Phil: Re lifetime subscriptions. Many people who hold these accounts believe they are entitled to a much higher degree of service. If you’re thinking of offering lifetime subscriptions, perhaps think twice about what it will take to support them.



Leave a Reply