community games design massively multiplayer mmo signup processes web 2.0

Customer Relationships and Support for Online Games and MMOs

Here’s a question about increasing the profitability and decreasing the development cost of any MMO, although probably no-one except the web-people will recognise it as such (and even some of them won’t get it):

How do you improve the customer support for an existing MMO?
[where do you start, and what do you target?]

Or, to put it another way, here’s three questions that I bet most games companies cannot answer without waffling:

  1. What is “good” customer support?
  2. Why do we care about customer support?
  3. How good is our own customer support?


Before I go any further with this, I want to point out that there are (at least) three main areas of Customer Support, of which I’ll only be covering one. The others are all covered reasonably well within the industry (hmm…maybe not so well, actually – but certainly better than this one).

Those other two areas are:

  • Controlling what CSRs (Customer Service Reps) say and do to make sure they are “on-message” with what the Marketing and PR departments are trying to do
  • Managing a community via forum-moderation, live events in-game, real-world events, etc

The other kind of Customer Support

…can be looked at in two different ways:

  1. Handling routine questions, complaints, rants, and moans from customers. Helping them fix their PC enough to play your game. Helping them get their credit-card payment to go through successfully
  2. Buying future revenue for unrelated products, one person at a time

This latter view emphasises the idea of CRM (Customer Relationship Management). I’ve worked with plenty of people who felt we “ought” to be nice to customers, and make their experience with us a pleasant one. They generally disliked (or detested) the first view, but they themselves were only half-way between the two views; they didn’t really know why we cared (or should do) about customer support. I was like that myself for a long long time, until I sat down and thought about it properly.

We still don’t know what “it’s a service not a product” actually means

I’m sad to say this, but it’s true. On the whole, MMO and Online Game developers/publishers *still don’t get it*. They think they do, but they don’t.

Various people started chanting the mantra “MMOs are a Service, not a Product” back around the time of Everquest (the first one) and Ultima Online. In the game industry at large it peaked around the time of Gordon Walton’s “10 reasons you don’t want to make an MMOG” talk at GDC 2003. By now (5 years later) the industry has understood a couple of things about this subject, but on the whole it’s failed to think about it strategically, and has pretty thoroughly *missed the point*. Most people see the trees but not the wood – the mantra is so short and simple and easy to understand, people tend not to think it through, and so don’t realise the connotations.

What’s the most important high-level goal of a Product company?

“Shift more boxes”

What’s the most important high-level goal of a Service company?

“Purchase more customers”


Yes. The primary goal with a service-oriented business is to BUY something, not to SELL it. Because a serviced-customer is a cash-cow that can be milked at any point in the future, every day for the rest of their life (in the case of corporate customers “the rest of their life” can be a very long time, maybe even measured in centuries). It’s worth buying them, even at great cost.

For people who are accustomed to the box-shifting view of business, this feels like it flies in the face of everything they know about business. Actually, it doesn’t, but it exposes an unstated assumption they’ve made throughout their lives: with EITHER business, you are NOT really selling (or buying) anything – you’re entering into contracts. A “sale” is, to give it its full title, a “contract to exchange a thing of value (a good or service) for a price”. In the case of box-shifters, the terms of the contract merely state that they are receiving “an amount of cash”. In the case of service-managers the terms of the contract state they are receiving “a batphone connected directly to the mind + wallet of the consumer”.

Or, to simplify: box-shifters “sell” for cash right now, and service-managers “buy” a relationship that they can later rent for cash in the future.

And there we have the root of all that follows: any company that chooses to sell a service instead of a product has – implicitly – chosen to FORGO cash IN LIEU OF taking possession of a RELATIONSHIP. i.e. they’ve actually *paid money* to get this “relationship-thingy”, so they’d better make sure they know what they’ve bought, that they didn’t “over-value” it, and that they know how to extract the “rental money” in the future.

Yes, you can still charge cash AS WELL as buying the relationship – but most people are doing that well enough already, and don’t need help from me to do it better.

Time for me to answer some questions…

Why do we care about Customer Support?

(NB: CS == every time a customer needs or wants something and get its from something you’ve done or said, whether they contact you directly, visit your website, or merely go back and read past emails you’ve sent them)

ANSWER: Second only to the in-game experience itself, CS is the richest, most direct part of the Relationship that you’ve purchased. For a service, it is more important than all the rest of your Marketing and Sales.

Marketers fight constantly to get their voice heard loud and clear – and without distraction – by consumers. In practice, thanks to free speech, anywhere that YOU can talk to the consumer, so can all your competitors. And so you find yourself desperately trying to “stand out from the crowd”, and get your message across. Even then, you cannot personally visit each customer, you have to rely on communications channels, different media (print, TV, news reporting, etc) – and each one of those channels introduces Chinese Whispers, corrupting (or deliberately censoring e.g. your mighty claims) your message.

A direct, unfiltered, uncensored, uncontested channel to every consumer’s mind is the best thing a marketer can hope for.

And you have one. It’s

  • …sitting down in your CS department swigging a bottle of meths and wondering why no-one cares about it.
  • …lieing in a filty heap of smelly clothes out the back of your website, wearing a tattered hat marked “Account Management”.
  • …parading itself in a smokey bar full of leering shadows, doing lap-dances in a bra covered in sequins that spell “Abuse” and a thong that says “…My Email Address”.
  • …erected a toll-booth at every corridor in your User-Experience Building, with three forms you have to fill out in triplicate for everything from getting a glass of water through to going to the bathroom. The three forms are titled: “Username”, “Password”, and “Best friend’s neighbour’s mother’s maiden name” – and each corridoor has a different layout of forms, and a different set of valid answers. Some of them swap about randomly every morning “…to confuse the Enemy!”.

Fine. Enough poking games-companies in the eye with a blunt implement. Where do we go from here?

What is good Customer Support?

That makes it quite easy to answer this question now. It has to:

  1. Monetize the relationship we paid so much money for
  2. Prop-up the relationship when it starts to falter
  3. Cement the relationship and make it stronger
  4. Remind the customer how much nicer you are than their last Girlfriend/Boyfriend, and that if they leave you they’ll never find true love again
  5. KEEP THAT RELATIONSHIP AT ALL COSTS! (up to the difference in how much it’s worth and how much it cost to buy in the first place)

I’m sorry to all the people who diligently work in CS with no thought of monetization and think they’re just genuinely helping people. Yes, you are helping people. But you’re paid to do it because someone else in your company (your boss? your boss’s boss?) is using that as part of how they monetize it, or as part of something that helps to make sure the customer is still around in the future solely in order to BECOME monetized.

I put that last item in caps for a reason other than dramatic effect. Since the first item is “to make money” the last item is “…(profitably)”. If you calculate the total FUTURE revenue from this customer, and then spend up to that amount in order to keep them, you are guaranteed to always be profitable. Since you cannot guarantee they will remain a customer, you have to put a percentage discount on the expected future revenue that is proportional to how many of them you think you will lose unavoidably. Obvious stuff, and obvious difficulties abound there … all makes for a busy time for CFO’s and CMO’s to extract the most profit possible.

How good is our own Customer Support?

Most companies cannot answer this. In desperation, they collate graphs such as “number of support queries per month” and “percentage of support queries marked as Resolved by the customer, and with a customer-rating of 4 stars or above”. So what? That tells you some stuff about how good your CSRs are at being nice to people (not a lot, but some); it’s largely irrelevant from a wider CS point of view.

What you need to evaluate (again, self-evident from all the above) is more like this:

  1. How much money are we spending on each customer? (min, max, average, median)
    • this is a simple headline figure, it solves no problems, but it can hilight that there IS a problem … somewhere
    • should be “total cost of the Relationship” not “how much do we pay our CSRs”
  2. Segmenting customers by type, what’s the profitability for each Relationship?
    • Choosing those types is what you pay your Marketing Director for, it’s not trivial (inventing them is tricky, but working out how to actually MEASURE each type can be really difficult)
    • examples include:
      1. “people who bought our product at retail”
      2. “people who bought the digital distribution version via steam”
      3. “Spike TV viewers who saw our review in January 2005”
      4. “Parents who liked our game so much that they bought a copy of our game for their children”
      5. “Parents whose children liked our game so much that they bought a copy for themselves”
      6. “People who created an account on our website”
  3. Ditto what’s the loss-of-relationship rate?
    • i.e. ONE of the inputs for calculating that “discount percentage chance-of-losing-a-given-customer-over-time” figure mentioned earlier
  4. How much money are we making from each customer?
    • YES, it’s “what are they paying in monthly subscription / virtual goods purchase volume”, but NO that isn’t all it is
    • How much cash have we made by selling them some unrelated product or service (careful: that one will need to be monetized too)?
    • How many unique products have we sold them?
  5. What are the trends in all the above for our userbase, zero-aligned?
    • i.e. if you measure all those figures and graph them over time for a user, you get one graph for each that shows e.g. “after 12 months, they bought their first secondary-product”
    • …if you average that for “all users in a given segment” (see above) then you get a graph that is both observational (based on fact) and also predictive for any future consumers of the same or similar type
    • You can then use this to spot trends in your relationship-management and relationship-capitalization
  6. Then get fancy: instead of graphing the above by “time” on the x-axis, graph it by “milestone”. This way you can see if e.g. “having to visit the website to file a bug” is damaging your Relationships (people buy less other stuff once they’ve done that), or is failing to capitalize as much as intended (people don’t buy ANY MORE THAN BEFORE after they’ve visited your website to file a bug)
    • Read that example carefully. Think about it. Most MMO/online games companies don’t think about it.
    • HINT: Remember what I said earlier, about how the Relationship is a direct channel to the customer. Think about what that SHOULD have been going down that channel while the user was filing the bug
  7. …and so on…

All the above list is, to a marketing person, teaching a granny to suck eggs. Good ones should know this stuff inside out. On a daily basis they ought to be working with more detailed, cleverer, more difficult-to-measure-but-we-measure-it-anyway-because-we’re-hard-workers demographics and actions. I’m presenting it more to illustrate the point than as an actual guide (I wouldn’t advise any real company to blindly do the above verbatim).


The Relationship is *everything*, and it must be:

  • Guarded
  • Monitored
  • Strengthened
  • Monetized profitably

at all times. It may seem that the last of those conflicts with the first three. In fact, all four of them are mutually conflicting, and you have to continually comprise, and re-compromise, finding the dynamic balance that best fits your company’s overall strategic aims.

The mistake many game companies make is to obssess about just one of the above (usually the “guarded” part if you “care about the company’s reputation”, or the “strengthened” part from a partially-enlightened marketing person). Many just ignore all four of them, and instead only look at the “spending” half of the word “profitably”, and ask continuously “how can we reduce CS costs?”.

Many game companies consider that the roles of the Sales and Marketing departments are to do this kind of analysis and activity on “future customers”, and fail to recognize the inherent waste of potential profitabilty that comes from ignoring the most valuable asset the company has: the hundreds of thousands of Relationships that it has bought, and paid for, but is only partially monetizing.

UPDATE: I just spotted this short post by Furqan over at Altgate that’s pretty relevant to this topic – about measuring the “value” of a free (i.e. non-paying) customer.

17 replies on “Customer Relationships and Support for Online Games and MMOs”

Good piece Adam. I just wanted to throw in something else and get your thoughts.

I would add to your list the customers view of a customer supports value system.

I’ll clarify!

I’ve been involved with mmo development on and off since 2001 but as a player from the mid ninties, I would have to agree that bad CS can colour the players entire experience and be a significant factor in customers withdrawing thier subscription. At the same time players will endure significant failures of service if they feel they have a significant and valuable relationship with CS and the operator.

The example that springs to mind is Eve Online, which I played for almost 4 years. In general the perception of the CS within the player community was very favourable.

There was the perception that the CS was fast, fair and proactive. Players felt thier problems were taken seriously and resolved to the best of the CS ability. There was also the sense of a significant dialogue with CS, a ticket could be opened and if needed a lengthy ‘conversation’ had to resolve it.

Significantly a sense of ‘justice’ is vital to a healthy CS / Customer relationship. Even if a decision went against you as an Eve Online Customer, the willingness of CS to justify and explain it, generated the perception of ‘fairness’.

This was vital when major failures of service occured. A particular incident springs to mind to illustrate this.

A couple of years ago it became public knowledge that a small number of individuals within the CS team had allegedly been aiding a guild (corp) within the game by spawning significant resources for them using thier GM tools.

Now this is of course a nightmare scenario for an MMO operator, with players questioning the fundamental existance of a fair playing field within your product. The operator acted quickly and publicly to resolve the issue and reassure thier audience. This in itself would never have been enough alone, the significant faith the customers had in the operators and the history of an open constant dialogue (via devs and CS) helped enormously.

As difficult as it can be to quantify, the customers perception of the ‘character’ of your CS is a significant factor in the relationship. In particular with customers who have never used the CS. They hear about the experiences of others and base thier assesment of whether to bother using the CS at all upon this shared perception. Certainly we have all at one time or another played an MMO in which it was widely believed that the CS was worthless to use and in fact an active hinderance to customers game experience.

Wow sorry didnt intend that to be so wordy but you got me thinking!



I see that as just another tool in the CRM (is marketing) toolbox. It has a cost; is it worth it to your game given your audience profile and the game brand? How will you leverage it a d can you afford to capitalize on it if you do decide to do it?

Like almost everything elaein CS that tool is “trivially” profitable *if* you can afford the capital expenditure required to get it up and running. Most smaller companies cannot, at least initially. Weak companies therefor write it off permanently – and you even sometimes hear them speaking at the leading industry confereces,saying so in front of hundreds of people. Smart companies know to revisit and periodically reevaluate it until they can afford it.

One of my employers liked to spend all the cap ex on it and get it flowing nicely then instead of reaping the rewards they panicked at how much theyd spent and insisted on shutting it down. No it doesn’t make much sense when you view it like that, but my point is they almost certianly weren’t viewing it that way, when they probably should have been.

Clive, I’m sorry, I was a resonably early player of Eve and your description of them is entirely unrecogniseable to me.

Yes, the common questions could be answered by the help database. However, getting a petition answered? Usually took a minimum of a week. And was usually a useless answer, in text. Note that the in-game and out-of-game systems never properly synched. Also, I know a fair few people who reported abuse (and I was there at the time) while not doing anything remotely abusive themselves were handed bans.

And, there was allways, allways issues with GM’s having access to data. On more than one occasion, an enemy fleet zero’d in on us after a GM’s visit to fix something. Some GM’s were downright notorious for their bias.

The foum moderation? Was both uneven and heavy handed, with some people’s downright abusive behavior tolerated, especially if they were for example writing chronicles or articles for the Eve magazine (including an outright permaban reversal when a certain author said they’d pull out if it wasn’t reversed) while other people were banned for extensive periods for RPing moderators didn’t like, “oversized” sigs which were not oversized at all (happened to me, *twice* – two of my three “offences”, the other one being critical of a dev descision which lasted 72 on the test server), etc.

Also, the bug reporting website (STILL, and note it’s not in-game and it’s an entirely seperate system…) randomly drops issues, and since theres no confirmation emails or the like…this is how things like the economics exploit get missed.

The entire thing with BoB and CCP was overblown, it was a drop in the water compared to known past issues with anti-BOB GM’s, frankly (including incidents of 200+ non-BoB battleships restored, with no BoB ships restored during server lag).

Don’t take me as pro-BoB (indeed, there’s emnity from them towards me which on at least one occasion has spilled over into RL behavior), but…

fits right in:

Carl Shapiro and Hal R. Varian: “Information Rules – A Strategic Guide to
the Network Economy”. Harvard Busines School Press, 1999.

I’ve posted a few articles on my blog about CS:
Customer Service doesn’t matter
Customer Service still doesn’t matter

To put some of my thoughts in the terms Adam describes above:

1) Relationships with companies are generally not important to people. I’ve bought a lot of items through, but I have nowhere near as much loyalty to that company as I have to even the most casual acquaintance I have. So, as a company, you’re trying to foster a one-sided relationship.

2) Other relationships matter more than the one your customers have with your company. I talked about a poor CS experience I had in WoW in the second article above. The harm to my relationship with Blizzard did not matter compared to my relationships with my friends that still played the game. I only left the game when they left.

On the other side, I had a free account on a game because I knew some developers at the company. When my friends were playing, I was right there with them playing the game. As soon as they left, I left the game as well even though I had a very good relationship with the company. (Admittedly, the company wasn’t making much money off of me directly, though, so it was still a one-sided relationship.)

3) Other issues are more important. Game players are novelty seekers, in general. So, a newer game with crap CS is often going to be more appealing than an established game with the best CS. Or, people want to be in the “in crowd”, so they’ll flock to a popular game despite CS issues.

3) Costs are almost impossible to measure. I listed three types of people in your game: needing no CS, needing a little CS, needing constant CS. That last group drives up your costs significantly, and it is hard to really measure. If you have a phone number, someone is going to use your CS hotline as a cheap therapy session. How much is that really costing you if they call one hour a day every day for a month? It’s a question of how many other people are not getting the CS they feel is necessary to keep playing the game, and that’s hard to really measure.

On the flip side, how much does a customer that you lose really cost? Were they getting bored with the game and about to leave, anyway? Are they already so bored with the game that they grief the CSRs for fun now? If they leave, will they take a bunch of their friends with them? All these examples have different cost.


All that said, relationships are important. In my estimation, one of the reasons why WoW did so well was because of the strength of the Blizzard and Warcraft brands. Blizzard had built up relationships with people over many years by developing high-quality single-player games. I believe this relationship is much, MUCH more important than any relationship a current MMO company can hope to build with their audience through traditional CS means. But, not every company can do what Blizzard did to build the relationships necessary for the game. So, they have to rely on all the other factors I mentioned in my blog posts.

In the end, CS still does not matter… as much as all those other factors.

Thanks, Brian!

I think your CS posts hit the nail in the head wrt plain CS, and I think the last three sentences of the first post sum it up very well. I particularly liked your counters where commenteds brought up some of the classic “thought experiment” justifications people give for CS being worth it, which are what people like to tell themselves are true but which just don’t hold up in reality as soon as you get access to the stats of real live services.

In that context, what I’m trying to point out is that if you view CS from a different context, with different metrics both of success and of value and of cost, and hence also change the meta activities you engage in, you can justify doing a lot more with it.

I read your posts a long time ago and they influenced my thinking: they were in the back of my head when I saw examples of organizations somehow contradicting your statements by spending a lot on CS and yet also somehow getting actual business value out of doing so. This post is partly trying to explain that unexpected difference that I saw.


You seemed to have addressed a lot of my issues and I remember your posts on my blog, so I figured you had read them. I was restating a lot of my issues in your terms. For example, are you really building a relationship when only one side considers it such (the company)?

I read an article related to this concept a few days ago: The upshot is that a company wanted to build a relationship with customers by allowing them to automatically create an account, but most customers didn’t want to get into a relationship, they just wanted to buy stuff off the site. (The article is an interesting read.)

But, how many people are just playing a game for reasons that are truly beyond the control of the company? As I’ve said, most games I play for any length of time are because my friends are on it. There’s not much a game company can do to keep me there if my friends are not playing, and they often leave because of outside influences instead of game-related issues.

So, in the long run, I’m still not seeing a huge benefit to CS beyond the minimum necessary to keep most people playing; what people refer to as “poor CS”. You can try to measure it in a different way, but I’m not sure that stacks up, either.

My thoughts.

Thanks Adam,

This is a really tricky subject, and one that is close to my heart. Brian, you make good points too. Here’ s a rambling stab at some of my thoughts…

I believe that MMO CS is a tricky beast compared to traditional CRM logic because people do not experience the product in isolated cases, but rather social networks. In fact, I can’t think of a tighter social network than an MMO.

If I buy shoes from Zappos and have an awesome experience, I feel happy, and I guess I have a ‘positive relationship’ with them. But I don’t spend my evenings hanging around in chat rooms with other Zappos customers comparing notes and having any sort of a shared recreational experience. In fact, if the shoes fall apart, I don’t blame Zappos at all, I blame the company that made the shoes. Anything that Zappos can do to help me get them replaced at that point is seen as ‘good CS’ even if the product was total crap to begin with. Several points in this paragraph could be entire essays or dissertations on their own.

I think the real crux of CS in the MMO space is that there are many ways that you can totally blow it that will lose customers who are ‘on the edge’ of quitting, who are new to the game, or have low investment in general. Especially customers that might buy a ‘future product’. And, the biggest risk is one to the reputation of the company overall. Here in this thread we have one positive description of EVE’s service, and one negative one. Which one is right? Which one is more influential? I never played EVE more than a few hours, but these descriptions might change my opinion about whether I would play a game from them in the future, or whether I would play EVE again. I have plenty of friends who play EVE now, so if I decide to play I have to be willing to risk my free time on something that might be a bad experience. My perception will be a big factor of this, even if my friends tell me ‘things are great’.

The real job of CS is to be genuine and to make sure we don’t drop the ball on the straightforward stuff. Did we actually read your question? Did we make sure not to send you some troubleshooting steps which were provided earlier in the thread? Was our response polite and genuine? Was it actually helpful and honest? None of these things cost money, but they do have an impact on how people feel about the company. The genuine and honest things are the really big ones.

I’ve never worked for a company that said ‘we are happy with bad service.’ But, I have worked for companies that said ‘we want good service for less money.’ Every year, less money. I’ve never worked for a company that has said ‘we want the best service in the world no matter what it costs.’ If we are talking about MMO’s, I’m not even sure the last one is the best, but somewhere between the second and third option surely lies the most value.

It is very tricky to measure whether someone ‘would have quit’ no matter what you did. But it is relatively easy to measure how many people quit playing after contacting support. If that number is higher than the number of people who quit playing in general, then you’ve probably got some work to do.

Good post, and not incorrect. Just incomplete and a little off target.

While you’re on the right track, I think you are a victim of the problem and fail to see the forest because you’re standing in all those trees.

Here’s where you’re missing the mark: Only a very small portion of customer service in an MMO is actually done by the CS department. In a service business, the product *is* the service. Everyone who actually provides the *service* on a daily basis is what counts most.

Take cable television service as an example.

-How often do you use the service?
-How often do you actually call them to get a problem fixed?
-When asked to rate the service, do you think more about those few telephone calls with CS representatives, or do you think more about the quality of your television viewing experience each day?

Imagine if your cable company was run like most MMO game companies. They would sell you a package that included about 3 months worth of television programs. Some of the shows wouldn’t be much fun, some wouldn’t let you watch them all the way to the end, and some wouldn’t start at all. If too many other customers were watching the same show, then you’d see the show skip and stutter, which may result in you having to start it over from time to time. When you call customer support…. huh? ….call customer support? ….OH THERE’S NO TELEPHONE SUPPORT, so you have to use e-mail or the web site. They may or may not update the shows and get the broken shows working, but there’s no promises. The next time they plan to release another three months worth of shows could be YEARS from now, or never, since there’s no publicly available schedule or contract agreement saying that they promise to provide any more than the original three months of shows.

Sure, it’s apples to oranges in *some* ways, but not so much as the game companies think. If they can’t provide mainstream quality service they’ll never break into mainstream entertainment revenue numbers.

Comments are closed.