Email a colleague    

July 2011

Robots for Hire: Verifying Accuracy In the Age of Complex Mobile Billing/Charging

Robots for Hire: Verifying Accuracy In the Age of Complex Mobile Billing/Charging

Thanks to the iPhone, the flat rate or tier-priced postpaid billing is the current runaway billing method for smartphone service in the U.S.  Yet in the years ahead, odds are that real-time charging will steadily eat into flat rate’s market share, which means operators must now get serious about billing accuracy issues that the large flat-rate bucket plans have masked for some time.

For instance, if you sell an SMS package where a certain number of SMS messages are free, and 10 cents for each SMS above the limit, how do you ensure it’s billed accurately?  What if your rating system is fine but the switch is sending the wrong CDR?  How do you know that the offer your customer sees is the same one reflected in billing?  And if your customer roams, how do you know your roaming partner is billing you at the right rate?

These issues are with us today, of course, but as complex lifestyle calling plans and real-time offers gain traction, you can bet that investigating billing quality issues will rise in importance.

But not to worry, says Xavier Lesage, president of Araxxe, whose firm has been helping operators in Europe and Africa tackle billing quality issues for seven years now.  Araxxe, headquartered in Lyon, France, offers an ingenious managed service for bill testing.  The service requires no purchase of call test generators or other robots, though Araxxe itself is a big buyer of this equipment, deploying across 100 countries and 400 major international routes.

In this interview, Xavier explains his firm’s unique service and its power in detecting billing errors as the customer sees them.  In addition, he shows how Araxxe’s robots verify roaming charges and identify international fraudsters.

Dan Baker: Xavier, it would be great to get a quick understanding of how your service works.  I think most operators are familiar with the test call generators that companies like Roscom and The Board Room sell.  But yours is a new approach.

Xavier Lesage: You’re right, Dan.  In fact, we call our business a billing monitoring service — and operators subscribe to it.  The service leverages our main asset — network robots, computers with phones — that are able to generate voice calls, SMS, MMS, email, Web sessions, and any type of data traffic.

This is a real managed service.  There is no IT interface whatsoever, and our customer is typically either the CFO or head of Customer Service.

The client first sends us, say, 10 SIM cards and gives us on-line access to its PDF invoices.  That’s all we need.  Usually a test run of 5,000 to 10,000 calls is sufficient to check all the rating and switch-to-bill chains in the area the client needs to investigate.

Our clients vary from large clients with 20 million+ subscribers to small operators on remote islands.  Sometimes the scope goes up to something like 100 SIM cards and 100,000 calls.

There are plenty of revenue assurance software vendors out there.  How does your service differ from what they do?

We do not compete with WeDo, Subex and the others.  Assurance software checks for real-time issues, such as alerting that 5 percent of SMS messages are missing.  Most of the techniques they use are statistical checks based on the processing of billions of call data records.

But in our case, we check each of our calls individually, looking for billing-configuration issues and price issues in a well-designed sample of transactions generated by our robots.  These are not checked by revenue assurance software.  For example, if the CDR is not configured to what the billing system is expecting, RA software will usually not detect that.

If you sell a bundle of 60 SMSs for $5 and 10 cents thereafter, we generate 100 SMS messages.  Then at the end of the month, if the 61st SMS is not being billed, we raise an alert.

At first glance, this may appear to be a very tiny issue, but it could be 2 percent of the revenue on this precise bundle.

Before you go to work for a client, do you obtain the rating plans of the operator?

No actually.  Instead, we use public information such as what’s featured on the operator’s website or in a marketing brochure.  And frankly that’s the beauty of our approach: Our billing tests assume nothing.  The rating system could be in error, the switch-to-bill configuration could be wrong, or the advertised market price could be different.  That’s what our analysis finds out.

Which brings me to the other key value of this service:Look at the bill from the end user customer’s eyes, which is the best place to begin.

The assignments operators give us are pretty diverse.  Two months ago, a client asked us to check 200 complex destinations to see if the new rate plans on the prepaid/IN system were properly configured.  Another two clients are switching their billing system from one solution to the other, and so they have asked us to check if the migration has introduced any new errors.

I also understand you use your robots to detect fraud problems.  What type of fraud do you address?

Our analysis usually investigates the fraud around high-priced call termination fees in countries of EMEA who have “ee facto“ monopoly to terminate calls toward their own customers.  Fraudsters in such countries make a lot of money bypassing interconnect gateways.  Here we use various techniques such as purchasing interconnect minutes or calling cards, then testing them.

Here’s how our fraud service works.  Let’s say Telefónica in Spain wants to monitor interconnect revenues and using our robots makes calls from many countries.  So we call some specific targets in the Telefónica network in Spain and check: Did the calls enter the right routes and go through the interconnect gateways?  Or did those calls enter another way, say through a hacked PBX or using a fraudulent trunk or SIM Box?  A SIM box is essentially two phones on different networks, rigged so that a call arriving on one is routed out again on the other.

We report to the operator the percent of calls passed through legitimate ways and the number that went through fraudulent paths.  And we identify the SIM boxes that the call went to.

Let’s say our client is a mobile operator in Morocco who wants to check the legitimacy of calling cards to Morocco.  We would usually test calling cards that have very cheap prices to Morocco and buy those cards off the Internet.

If Verizon wanted to terminate a call in Morocco, Verizon needs to pay the operator 18 cents a minute.  But if you can hack the PBX of a large hospital in Morocco through a hidden door on the Internet, you can bypass the Moroccan operator’s gateway.  So the fraudsters pay nothing for the phone calls and can charge, for example, 9 cents a minute to Verizon enjoying a hefty margin.

It’s very interesting case.  And it goes to show there’s still some mystery and intrigue in Casablanca.  But tell me, Xavier, do you have any North American customers?  And where do you feel U.S. and Canadian operators need your type of service?

We are just entering the North American market actually, so our understanding of the market’s needs is limited.  However, we recently won a contract to check the accuracy of bill roaming when an American operator’s subscribers call from international locations.

Our hunch is this roaming area is a key one and it seems to be an area where North American operators have less experience than their European counterparts.  In Europe, you are constantly crossing territory and country boundaries, and there are plenty of operators with a very limited footprint, which means you require many roaming agreements and need to deal in multiple currencies.

We usually recommend operators do a test in one roaming location for each of the regions of the world.  Plus make tests from countries where your subscribers go abroad.  If Americans mostly travel to Brazil, the U.K. and France, do tests of the roaming in these countries.

This article first appeared in Billing and OSS World.

Copyright 2011 Black Swan Telecom Journal

Xavier Lesage

Xavier Lesage

Xavier Lesage is president of Araxxe, a managed services firm that uses robotic test call systems to detect revenue assurance and fraud issues.

Prior to joining Araxxe in 2005, Xavier managed major revenue assurance and billing programs at Accenture for telecom clients such as Orange, Vodafone, Telecom Italia, AT&T and Liberty Global.   Contact Xavier via

Black Swan Solution Guides & Papers

cSwans of a Feather

Related Articles