Though on the outside the boxes may look different, inside all washing powders are the same. It makes customers think they have a lot more choice than they really do. Increasingly it is the same with telecommunications. A few special additives may lead some customers to pay a premium, but the average customer just wants the basic no-frills product, at the best price. So it comes as no surprise that Vodafone UK, has turned to two of the most straightforward marketing methods long used to sell washing powders and the like. Sales flagging? Cut the price to encourage some new customers. The brand fails to attract a certain share of the market? Put the same product in someone else's box, with someone else's brand on it. Neither tactic is new, but it is fascinating that even Vodafone UK is now being forced to use them, and is eroding margins in the process. Read the story here. So the only way to boost those margins will be to squeeze costs and to grab every penny available. Which should make for interesting times for Vodafone UK's revenue assurance team.

Posted by Eric on March 31, 2007 12:05 AM | Permalink | Comments (0) | TrackBacks (0)

The results of opinion polls should always be treated with caution. They usually tell you more about the bias of the people paying for the poll than anything else. So it is no surprise that the new UK government quango responsible for promoting online security, "Get Safe Online" has produced some pretty terrifying stats on the level of internet crime in their first survey. See the press release for the GSO Internet Security survey here. The figure that most worried me is the high proportion of internet users who did not hold themselves responsible for keeping their information secure. Whilst 48% agreed that they were primarily responsible for the safety of their online information, a staggering 13% thought the ISPs should be responsible. A further 16% thought it was the responsibility of their bank (do they also blame their banks if they spend their money foolishly or loose it down the back of the sofa?) Given that the poll also claimed that the average internet fraud cost the victim £875, that is a lot of people willing to risk a lot of their money in the hope that somebody else will cover their losses. See the BBC's write-up of the GSO survey for these and more stats.

So ISPs should beware. Whatever the legal rights and wrongs, ISPs need to have cast iron contract terms and conditions, and support that with clear communications to customers that explain internet fraud is not the fault of the ISP. At £875 a time, they want to take a lot less risks than their customers....

Posted by Eric on March 26, 2007 02:25 PM | Permalink | Comments (0) | TrackBacks (0)

Imagine a future where everybody drives electric cars. Or a future where people both buy electricity from national suppliers and produce their own for domestic use, perhaps using solar panels. One of the biggest problems for energy conservation will be loading, trying to balance the use of electricity with its production, in an effort to minimise the difference between the peak load and the average load. This is crucial for energy conservation because of two reasons. First, power stations are most efficient if their output is continuously being delivered at the designed optimal level of output. This is better than increasing and decreasing output to reflect changes in demand. Second, transmission over distance involves wastage. The better the match between local demand and local supply, the less energy is used just to move electricity around the grid. Of course, the problem of loading is already with us, but increased demand through using electricity for cars, or increasing the variability of demand through erratic home micropower generation, might accentuate that problem further.

One partial solution would involve a change in mindset. Solutions historically have focused on supply-side management techniques, like building lots of small local Combined Heat and Power stations to service local communities, building international connections to transmit between different national grids, or by storing energy through techniques like pumped storage of hydro electricity. In contrast, there is less effort placed on demand-side management, and techniques used to manage demand are relatively crude. These focus on the promotion of energy conservation and applying different charges for consumption depending on the time of day. This in turn promotes relatively crude mechanisms for storing energy locally. For example, heat energy can be stored in order to utilize off-peak electricity supply. Put simply, electricity is used to heat up a large pile of bricks during the middle of night, and the bricks radiate this heat during the day. A timer in each house will click on the supply of electricity at the start of the off-peak time, and click off at the end. This may help balance the load overall, but it is pretty unsophisticated, and arguably not very efficient. Imagine if, instead, a remote computer could chose when each house's pile of bricks is heated up, with the goal of trying to keep a constant demand by switching on and off the use in each house in order to match demand with supply. Think then of the battery in the electric car, or of a battery connected to the micropower generator in the person's home. These would simply be more sophisticated stores of energy than the piles of bricks. They could also be used to help balance the load, storing electrical energy during off-peak times, but also contributing electricity during peak times. The car's battery would be fully charged during the night, ready for the commute to work in the morning. Any excess left in the car battery after the return home at the end of the working day could be run down during the evening to help meet household needs, especially if there is a sudden surge of demand in the house. Balancing the load in such situations would require a lot more co-ordination, meaning a lot more communication between the electricity company's computers and the ones in the car and in the home.

There is already an elegant solution for communications for demand-side energy management: transmit data between the relevant devices using power line communications. Despite having lots of promise, power line communications have so far failed to evolve into an effective competitor for delivering broadband, and even less so for voice. Problems with power line communications include poor quality because of the inherent noise in using power lines, and interference with radio. But for electricity demand management, power line communications provide a straightforward solution once the infrastructure is installed. All the relevant devices will already be connected by electric power lines - the battery in the car, the home meter, and a computer to manage the regional electricity load - so this mechanism avoids the need to connect a separate communications line to each device. Also, the bandwidth needed for such communication will be small, decreasing the issues with quality and interference. In future, power line communications could enable far more intelligent load management, balancing demand and supply by controlling exactly which household devices receive electricity and when. Better still, it can be used to deliver cost savings in the here and now. Power line communications could be used to eliminate the need for a person to visit the home in order to read the electricity meter. It would also improve billing accuracy by doing away with the need for estimated bills.

Like everything else in life, there is some bad news and some good news for demand-side management. The bad news is that the many politicians who have jumped on the green bandwagon mostly think in terms of old-fashioned solutions to problems. Prefering solutions that focus on the state and central control, most politicians talk in terms of punitive taxes for the perceived "undeserving" use of energy, such as personal travel, especially when this supposedly has to do with carbon emissions. Note that according to the International Energy Agency's 2006 report, more energy is used globally for electricity than for transport, and that coal, a fuel with only a negligible level of use in modern transport, is a greater source of carbon dioxide emissions than oil. Other favourite solutions for our unimaginative leaders include the building of more nuclear energy power plants and increasing storage capacity to manage load. At the same time, we also see some tinkering with daylight savings time and increasing promotion of the benefits of pumping up tyres on cars, but these initiatives are better at getting cheap publicity for politicians than they are at reducing emissions. So what is the good news? You may have thought what I wrote about demand-side management of electricity using integrated communication is science fiction. Thankfully, it is not. Not only is it science fact, but it is becoming commercial fact too. Forward-looking businesses are already planning for demand-side management of electricity. Take this press release from Manitoba Hydro, explaining their plans for a pilot installation of new automated meters. They will be using Cannon Technologies solutions that incorporate power line communications. Though Manitoba Hydro is starting small, they identify that the benefits will be long-term. At first, the power line communication will enable Manitoba Hydro to remotely read meters and identify outages, but it is also an enabler for management of demand. Intelligent real-time balancing of electricity supply and demand will likely be a key technology for increasing energy efficiency and reducing carbon emissions. The sooner it is implemented in practice, the better.

Posted by Eric on March 25, 2007 07:50 PM | Permalink | Comments (0) | TrackBacks (0)

Here is a brief catch-up on recent revenue assurance sales announcements:

Lavastorm sells a business assurance solution to Com Hem;

ECtel makes a first-time sale to Digitel GSM and gets a follow-on order from CAT Telecom;

ATS secures a new contract with a Tier 1 Canadian carrier and renews a contract with a Tier 1 US carrier;

SubexAzure wins a Fraud and RA contract from CellC; and

CANTV selects Radcom's network monitoring solution.

Posted by Eric on March 23, 2007 12:21 PM | Permalink | Comments (0) | TrackBacks (0)

Communications is getting better and better of course. But sloppiness abounds. What kind of signal is given by the failure to deliver messages on time? Vodafone UK's failure to deliver 11,500 SMS texts in time for a TV show vote is revealing on two levels. First it says that Vodafone UK is happy to lose nearly £6,000 in revenues. Of course, over 80% of that revenue was going to go to other businesses. So will Vodafone be compensating their business partners as well as refunding their customers? It is only because of the unusually high levels of scrutiny currently being given to TV-related premium services that this is a news story. We have little public data on how often this kind of thing happens; annual losses may run high. Second, it says the service provider does not care about whether a text message arrives in 2 minutes or 2 days. Saying the 11,500 texts delayed is a "small proportion" of all the texts sent that night is not very reassuring. These service levels may be good enough for quick, cheap and trivial text messages sent between friends. But standards will have to be a lot better to justify the hype - and cost - of the advanced data-driven services that Vodafone and others sell as the future of mobile. After all, if you look at the financial reports you will see revenues from those advanced services get added to the same total as revenues from SMS.

Posted by Eric on March 21, 2007 10:00 PM | Permalink | Comments (0) | TrackBacks (0)

If you hang around revenue assurance departments long enough, you notice a curious thing. Most of the time they report that losses are going down, because they keep finding things wrong and fixing them. But every so often something remarkable happens. And it usually happens at just the point in time when it seems that everything has been fixed and there are no more leakages to plug. The description may vary, but essentially a "new" kind of loss is suddenly discovered, or the scope of RA suddenly grows, and the value of losses reported shoots up. Of course, the revenue assurance team gets back to work and starts to diligently reduce the losses again until they nearly reach zero when... they suddenly shoot up again. So a graph of revenue losses against time ends up looking like the jagged edge of a sawblade. Of course, somebody cynical would say that the losses fall over the year in order to show that targets are being met. And somebody cynical would say that they shoot up at exactly the time it comes to set budgets and targets for next year. That cynical person would say that the numbers reported are more to do with securing ongoing headcount and investment in technology than they are to do with the actual level of benefits delivered. But I would not say those things ;) I would say it is often to do with people changing jobs. If the guy in charge of revenue assurance changes, the old guy wants to leave boasting that everything is fixed, whilst the new guy wants a report that justifies his budget - not one that justifies cutting his budget. By the same logic, if the exec who reads the reports (rare, but it does happen) changes, you can safely assume the business case for revenue assurance will be "bolstered" by finding a few extra causes of loss not reported to the previous exec. Of course, the saw can cut both ways. If nobody changes job for a while, depending on simplistic reports of benefits will in the end undermine the value and purpose of revenue assurance. And, incredible but true, but some execs are smart enough to see through the implications of a sawtooth report of leakage. So if you see a sawtooth trend in leakage reports, you learn more about the rate of personnel turnover than you do about the losses suffered or benefits added. Companies that want to consistently have the best results are better advised to work steadily through all issues, not just handle a few at a time only to be repeatedly surprised by "new" leakages. Because, of course, the leakages are very rarely new, and if new they are usually small, because there blockbuster revenues from totally new products are rare. Which means any very large "new" leakages will be large old leakages that have been ignored for a long time - and costing the business money all that time.

Anyhow, this kind of thing goes on all the time, and it is not for me to criticise if the revenue assurance team finds sawtooth reporting to be the best way to get business buy-in. However, I would be very cautious about letting the figures go public. But my old chums at Cable & Wireless UK do not seem to be as cautious. They have seemingly endorsed a press release that states the value of benefits Cable & Wireless UK have realised thanks to using cVidya technology. In fact, it states two values for the benefits. In the headline the benefits are "over £2.5m" but in the first paragraph the revenue benefits have fallen to "over £1m", which is not a great advert for the accuracy of reporting by either company (don't people proofread before they issue press releases any more? they also do not seem to know which year it is). Presumably the distinction between the two numbers is that £1.5m of the total benefit is from releasing network capacity, which is itself an interesting extension of the scope of revenue assurance. Of course, we do not know how these figures are calculated. For example, how do they put a value to the liberated network assets? You could argue that the assets would have never been liberated without work like this - so the value of liberating them is equal to the combined capex and opex costs of replacing the assets from now until the end of time. Of course, that basis for calculation would be silly. But I think everyone must be aware that numbers get manipulated for commercial reasons (Enron, Worldcom, SOX... surely people get the idea by now!) so these numbers are useless as presented. We can safely assume one thing: there will never be a press release stating that the benefits of a tool were less than its cost. But if you ask me, it will only be when that happens, and the sawtooth turns into a flat line, that revenue assurance can be said to have really come of age.

Posted by Eric on March 20, 2007 01:21 PM | Permalink | Comments (0) | TrackBacks (0)

New offerings like Voice over WiMax often prompt a lot of concern from revenue assurance people. The trick is to remember that there will be a few different risks whilst many of the fundamentals will be unchanged. So monitoring for indicators of fraud or checking completeness of EDRs are basically the same for Voice over WiMax, but some of the technology used may differ. This article about the risks of leakage for Voice over WiMax scores 9 out of 10 for being well-written and balanced. It would have been 10 out of 10 apart from the fact that the author is from ECtel, ECtel sell IP probes, and the major piece of advice is that people should buy IP probes...

Posted by Eric on March 13, 2007 02:04 PM | Permalink | Comments (1) | TrackBacks (0)

The problem with email spam is that businesses think of it as something-for-nothing. They get advertising, and the marginal cost of each email is zero. So they hardly care if it is annoying or fails to work most of the time. But some victims of spam have decided to fight back - by successfully taking the spammers to court. Taking advantage of the EU's love of arcane and ill-publicised regulations, they have identified a repetable formula for winning damages when sent unsolicited marketing; see here. The references in this story are to UK statute, but presumably most other EU countries have their own version of this law. In another great example of internet consumer activism, there is now the Scotch Spam website, dedicated to explaining how to take the spammers to court. It gives useful resources and explains the most recent court case where damages of £750, plus interest and costs, were awarded. So next time you find your inbox overflowing, make the spammers pay!

Thanks to Neil Hay for telling me about this story, through an unsolicited email ;)

Posted by Eric on March 12, 2007 01:18 PM | Permalink | Comments (0) | TrackBacks (0)

In the revenue assurance market, why does nobody sell Software as a Service (SaaS)?

As soon as I write the question, lots of reasonable answers spring to mind. But for every hurdle I think of, there is also a way to overcome that hurdle... apart from one.

Software needs to be tailored for each customer and each revenue assurance problem

Perhaps. But if that is so, why are software vendors already offering what they describe as COTS solutions with pricing models based on the volume of transactions assured? Why not instead charge for the real challenge, which is the tailoring and integration? Of course, the fact that software vendors use a SaaS-compatible pricing model does not mean their software is SaaS-ready. Probably a transaction-based pricing model is just better for getting the initial sale and earning higher fees in the long term. But tailoring software to integrate with the needs of each client just means that the software needs to be further developed to make it more readily configurable. The problem is not insurmountable; the same data analysis engine is ultimately being used for each telco and each kind of error. Expensive, slow, manually-intensive writing of new code, or the use of specific programs to extract and parse data ready for analysis could just be made more efficient. A powerful solution would accept any dump of a telco's data, so long as adequate mappings had been constructed to permit the relevant fields to be checked appropriately. Creating the mappings is just a combination of brain-work and knowledge: someone has to do that anyhow, whether it is the vendor or the customer. Making the mappings part of a repeatable configuration process, as relevant to SaaS, should also have benefits in making it more efficient and easier to update. So, the challenge is really just that the software vendors find ways to move their solutions up the SaaS maturity model, as described in this paper from Microsoft. And that will be driven by a combination of how forward-thinking the vendor is, and how demanding the customer is.

No capability to bill for revenue assurance SaaS

How pathetic an excuse would this be? A software company analyses a telco's revenues, but cannot find a mechanism to bill it in line with the level of work performed? Probably it is not something that the average vendor has had to consider, but in terms of technical know-how you would have to question the competence of a business that claimed to be able to analyse billing defects but could not calculate a bill based on SaaS usage. If nothing else, it would be possible for any software vendor to partner with another business that already has a SaaS metering and billing solution, like that offered by LeCayla Technologies.

No market for revenue assurance SaaS

You could argue that there are no telcos interested in a SaaS-style approach to getting revenue assurance software. The "long tail" theory tells us that this should not be true; there should be smaller players where buying the software outright is not economical but who would pay to use the software as and when needed. In fact, the attraction should be very clear: paying based on use would allow the telco to buy as much as they needed. It would take much of the risk out of traditional procurement. The telco would simply try the product, see if it suits their needs, and keep on consuming it if it meets their needs. If anything, the downside for the vendor is that they might stop using it if it does not meet the telco's needs. But it would make it much quicker and easier for a telco to extend its capabilities on either an ad hoc or ongoing basis, or to switch suppliers if not happy with their current revenue assurance software. And whilst some smaller players are being gobbled up in telco consolidation, the market should actually be growing as other businesses, like content providers, enter the market and look for solutions more suitable for their current scale.

Network reliability

Unreliable networks are a major obstacle that has slowed the roll-out of SaaS. The risk of Interruptions in network service may well be a good reason not to implement a software solution that depends on network quality. But hang on, what kinds of businesses are we talking about buying the SaaS? Telcos! There has to be something wrong with a world where telcos do not buy SaaS because they do not trust the networks to be reliable...

Security and privacy of data

Regular readers will know I get worked up about the security and privacy of data. So sending data backwards and forwards would be a justifiable concern. However, this obstacle gets surmounted (or ignored) easily enough most of the time. And there would have to be a degree of hypocrisy involved (and not for the first time) if a telco is happy enough to allow customer data to be sent to other businesses for all sorts of commercial and cost-related reasons, but not happy to do so when the data is being used to check if they have made errors...

Buying human brains, not software

The one hurdle that is not easily surmounted takes us right back to the first point I made: that getting the software to do a useful job is as much about human brains as it is about software. Software businesses often find themselves selling human brains rather than the software. Sometimes they sell the brains as an expensive addition to the software project. Margins on telling people how to use their software are often better than the margins on selling the software itself. On the other hand, software is somehow more tangible and easier to get approval for. I have seen more than one software vendor give a presentation explaining how their software project generated huge amounts of value, but where most of that value obviously came from one or two issues that were identified by a person, not by the software, early on in the project. So the source of the value is human brains, not software, although it is easier to sell the software than the brains. Which crucially brings us to the last hurdle - the people in the telco who have a job depending on the software. Some jobs depend on interfacing with the software each day, some on tweaking and developing it, some on managing and purchasing it. SaaS could be a threat to any and all of these jobs, depending on how the business currently manages revenue assurance. In particular, good, quick and effective use of SaaS solutions may serve to highlight previous inadequacies in the thought processes behind revenue assurance.

Footnote

I should really do my research before I begin writing a blog entry. My footnote is that my original premise, that nobody sells revenue assurance SaaS, is seemingly wrong. A quick search of the internet highlighted one vendor that does claim to provide revenue assurance SaaS: Primal Solutions and their "Voice Profitability" offering. I cannot vouch for it, because I have never used it. However, it does raise a new question: if they can offer revenue assurance SaaS, why are their rivals not doing the same?

Posted by Eric on March 10, 2007 04:00 PM | Permalink | Comments (1) | TrackBacks (0)

Interactive television is usually about asking questions. Some of the questions are silly, like "which celebrity should leave the jungle?". Some of the questions are sensible enough, like "what might a woman keep in her handbag?" though might generate silly answers, like "rawlplug" or "balaclava". But only in the last few days have the people paid a lot of money to ask questions of their own businesses - in other words executives - been asking exactly how they make so much money from interactive television. And the answers have not been good. See this news bulletin from the BBC and also this summary of what is going wrong.

In response, ICSTIS, the UK regulator for premium-rate services, will set up a scheme to license paid interactive television. Good to see they did their job diligently - waiting for a scandal to occur and then deciding to do something about it ;) But they can hardly be blamed. Despite all the recent media attention, ICSTIS claims to have received only 10 complaints about interactive television from the public. In the meantime, people had better resort to more traditional methods to protect their own interests, like just checking their bills.....

Posted by Eric on March 9, 2007 05:40 PM | Permalink | Comments (0) | TrackBacks (0)

A few days ago I blogged about how a data entry slip-up had apparently caused ITV customers to be overcharged £200,000. Whenever they used their remotes to vote on the popular X-Factor show, they were charged 50p instead of the correct charge of 35p. Now it seems that the quality of controls over all premium rate services linked to ITV shows are under scrutiny; see this story. This means that some programmes with formats heavily dependent on viewers voting or calling in will be taken off air until the relevant call charges have been audited. See here for a video clip from the rival BBC commenting on the story.

The amazing thing here is that a national broadcaster can put such a cheap price on its own reputation. The cost of sensible checks to catch errors immediately is trivial compared to the £200,000 overcharge caused as the problem goes unnoticed for months. Engaging Deloittes to check their rates are correct will not come cheaply (I used to work there and so have a good idea of the hefty margins they will be charging on this kind of work). And £200,000 is trivial compared to the potential loss of revenue. ITV makes around £20m every year from its late night phone-in quiz show. So you can assume that each night the show is not on air will cost ITV roughly £100,000 in lost revenues. In addition, viewers will tune in tonight, find that their favourite shows are not on air, and realise that their confidence in being charged correctly may have been misplaced. In fact, a certain proportion of customers will only now fully realise how much they have been spending to interact with their favourite television shows.

The only winners from this are, oddly, Virgin Media (formerly ntl-Telewest). If they had succeeded in buying ITV last year, it presumably would be Richard Branson making the excuses for failure and seeing the Virgin brand get tarnished almost as soon as the rebranding exercise had begun. Their revenue assurance team will be breathing a sigh of relief that this mess did not get dumped on them. So Branson has (a small) reason to thank Murdoch for blocking the ITV bid after all. But tomorrow they will all be double-checking that they have no similar problems within their own businesses. A lot of RA people are going to find themselves very popular in the next few days.....

Posted by Eric on March 5, 2007 10:53 PM | Permalink | Comments (0) | TrackBacks (0)

"Data inputting error". What a great euphemism. So whoever set up the rates for the X-Factor's voting line input a rate of 50p per call instead of 35p per call. Easy mistake to make? Perhaps. Luckily a combination of "robust audits" and "close monitoring" ensured the error was identified at "the first available opportunity", a mere three to five months after the calls were made. What I want to know is this: if the audits had not been "robust" and the monitoring had not been so "close", how long would it have taken to spot the error? Would anyone have spotted it? And does this mean not a single customer noticed the mistake?

Of course, UK customers can rest assured that their interests are protected by the regulator. The regulators' exacting standards "guarantee" that no more than £1 in every £50,000 is overcharged. So if somebody made a single error costing customers £200K, it follows that the next £10bn of UK phone bills will be spot on... doesn't it?

Posted by Eric on March 1, 2007 06:33 PM | Permalink | Comments (0) | TrackBacks (0)