Archive for November, 2010

It’s all well and good coming up with fantastic ideas to set you ahead of the competition, assuming that you’ve thought the proposition through properly. Take Samsung for example – in these times of paranoia and heightened security consciousness, they released a function onto their handsets which allows you to enter numbers into it so that, should your phone be stolen and a new SIM card installed, the handset automatically sends a default SMS from the new SIM to the designated numbers without the new user ever knowing. Genius! This means that you can contact the authorities with the details and they can duly track down the offenders and pursue prosecution. If you’re really lucky, you may even get your phone back!

However, what if your phone is stolen and leaves the country? This happened to me. When I first saw this remarkable little application I thought ‘Great! A nice added extra that’ll ensure that I can nail someone for stealing my stuff’. In fairness this is probably very true when it comes to opportunists who swipe your handset off an unguarded table in a bar, but falls far short of the mark when dealing with more organised crime.

It wasn’t long after the event of theft that my partner received a message from an unknown number. A very odd message indeed, as it wasn’t English and came from a phone in Morocco.

It seems as though handsets are a shared asset out there or passed on quickly, as my partner continued to receive messages every week. Suffice to say that she and I have been unimpressed by unsociable reminders as to my loss, with many rude awakenings in the middle of the night. Still – I thought that I could contact Samsung for a speedy resolution to this irritation.

Wrong. After around four phone calls and much back and forth between departments, I finally spoke to someone who had heard of this rather unique function. Sadly, as the recipient numbers are stored only in the handset, Samsung have no means by which to change them or reset the service. Their only advice was to speak to my network provider about it. Further confusion ensues here when I can’t speak to them regarding stopping the messages because the recipient account has nothing to do with me. Then the only advice they can offer when speaking to the offended account holder is to try and have the numbers added to a blocking list. There are two major flaws to this solution:

  1. to accomplish this you must have a crime reference number from the police and, as the handset stolen was mine and the associated crime number bears no relation to my partner, this wasn’t possible
  2. (the more inhibiting problem) as the service is only activated when a new SIM is installed and this is only likely to happen once the phone is resold/passed on, we can’t know the offending number to add to any such list until it has already started to act as an extra alarm call!

It seems that, due to lack of foresight in the concept of this service, there is no hope of controlling the transmission of these messages, unless the latest user sees it themselves an duly deletes the stored CLIs. Well there hasn’t been any activity for a month or so, so we live in hope now – although I have a drafted message explaining the problem to send back to any future number. I can but hope that’ll do it!

This is a small application and has done little damage, all things considered, but it serves as yet another example of poor implementation from a manufacturer or provider, when the how this should have been done is so easy and obvious (white list repository, network operator agreement, etc, etc, rather than something integral to the handset). Normally these initiatives are flawed through a lack of funding and/or time to get something to market as quickly and as cheaply as possible, with testing being what seems the natural sacrifice of any project*. The overwhelming approach now been operated by most companies, to stick to a deadline rather than ensure delivery of something with quality is a worrying and poses such challenges when tasked with assurance and risk.

The value add of RA & Risk functions tends to be directly proportional to what sway it has with senior management rather than the message carrying an independent level of gravitas. As RA and ERM groups tend to be the only areas of a business that understand the potential ramifications across the board and can give some form of measure or scale to these challenges, it seems clear that reputation has a big part to play and our functions need to step out hard into areas of development within our companies to try to offer prevention and not cure. Indeed, RA and Risk functions should be involved from the outset for all change activity in a business, including the definition of strategies.

Anyway, I digress! In short, if you have this particular function active on anything, set the recipient to be someone you don’t like.

*I am not forgetting the damage here also done by a ‘yes’ man approach from Sales to prospective clients, or indeed the innovation of a marketing function who don’t understand or appreciate the complexity of support or billing of a solution when desperately looking for new USP’s.

Bookmark and Share

I sometimes write about the size of the revenue assurance world. Estimates can be made based on revenues from RA sales, numbers of employees of vendors, attendance figures at conferences… but it is not easy. The first rule is never to trust a stat if it comes from GRAPA! They boast a few thousand ‘members’ (read: email addresses which they spam). In contrast, the TM Forum’s RA community has several hundred members, but people can only join that if their company is in the TMF. There is another source of data about the size of the RA world that I look at on a regular basis, though I have sometimes been reluctant to share it – the popularity of talkRA itself. Endless caveats should be applied to any statistics about the popularity of websites, because the methods used for calculation vary so greatly. For anyone interested, here are the most reliable numbers I have about talkRA’s audience.

  • Since the site launched in August 2008, talkRA has generated 65,000 page views from 32,000 visits.
  • On average, the site attracts over 600 unique visitors per month. There are seasonal variations in numbers, with February tending to be the most popular month, despite being the shortest month of the year. talkRA had over 1,000 unique visitors in February 2010.
  • About 60% of all visits come from long-standing recurring visitors. The visit and visitor numbers corroborate that talkRA has a very loyal audience of repeat readers.
  • Reader loyalty is also emphasized in the traffic sources. 42% of readers visit by typing in the URL or because they have bookmarked the site. 36% of visits come via web search, with the dominant search string being “talkra” and variants thereof. 21% of traffic comes through inbound links.
  • By far the most significant inbound link is from, because of the migration of readers when my blog first moved from there to Links from the TMF generate roughly ten times the traffic received from specific vendor sites. There are inbound visitors coming from a wide variety of vendor sites, including WeDo, Subex, TEOCO, Martin Dawes Analytics and cVidya’s Gadi Solotorevsky, but the number of visits from these sites is small.
  • talkRA has been read in 165 countries. Analysed by continent, talkRA has 43% of its readers in Asia, 28% in Europe, 18% in the Americas, 6% in Africa and 5% in Oceania. The site has most readers in: India, USA, UK, South Africa, Australia, Israel, Portugal, Malaysia, Canada and Indonesia.
  • Comparing readership numbers to national population, it becomes clear that talkRA does especially well in English-speaking countries which are also home to RA industry players – which is not that surprising when you think about it. Penetration is lowest on the continent of Africa, though the reason for this is unclear, especially as talkRA does well in South Africa. Low African visitor numbers may be due to poor internet connectivity or because of GRAPA’s popularity in African telcos.
  • After the home page, the most visited pages are: my bio (!), Downloads & Links, and Who Are We?.

The readership of talkRA has steadily and consistently grown, after allowing for seasonal variances. The audience is currently about 150% of the numbers when the site was first launched. Everything suggests that readership retention levels are very high, but new readers are added only slowly. This probably reflects two things. First, the total numbers working in revenue assurance is growing, but only slowly. Second, talkRA has been successful in its core mission, and appeals to those RA practitioners with the greatest dedication, but is less successful at picking up new readers and transient readers. As trade-offs go, that suits me. If forced to choose, I would rather talkRA appeals to RA decision-makers and pioneers than to those with only a passing interest.

Bookmark and Share

You have to love any website with the word ‘talk’ in the title. The good people at cost and revenue management vendor TEOCO have launched their own internet forum at You have to register to post, but it has been worth swinging by just to read their blog. The content has been very focused on the the US market so far, but for anyone interested in the complicated particulars of assuring communications services in that part of the world, it is well worth a read.

Bookmark and Share

The thing about collaboration is that you can never be sure where it will lead. Güera asked about the differences between crowdsourcing and collaboration. I responded by questioning the ease and possibility of either crowdsourcing or collaboration, at least in the field of revenue assurance. Lionel Griache of ProActiveRA came back and commented about barriers to collaborating on the development of an open source RA tool. I replied that people could use talkRA to set up a collaborative workspace for open source RA, if they wanted. Joe thought that was a good idea.

Is it a good idea? The quality of an idea is measured not just in the abstract – it all depends on the possibility of turning the idea into reality. Over a decade I have been looking, and finding, and sometimes failing to find, ways to collaborate with my industry peers. Sometimes the results are amazing. The publication of the talkRA book, which will be out next year, encourages me to believe that collaboration can work. Just as often, the results of collaboration are frustrating. You never know what the final result will be. But one thing is for certain: you will never accomplish anything if you do not try. And talkRA was set up to let people come together and try to make RA better. Our mission statement says talkRA:

provides a platform for thought leaders, allowing them to communicate and exchange ideas

I am no code hacker – a man has to know his own limits. But if people want to step up and use talkRA to develop open source RA tools, they are welcome. So this is a call for the people who would be the leaders of an open source RA collaboration. I have set up a new page to be an initial workspace for you. There is nothing much there… yet. If you want something good there in future, then leave a comment here, identify yourselves, and start bouncing ideas around. talkRA is happy to accommodate and support you, for as long as you find that helpful. Ask for what you need. I might not be able to give you all you want, but like they say: “if you don’t ask, you don’t get”. RA needs leaders, and for open source RA to become a reality, the world of RA needs to find the leaders who will get it started. Those leaders also need to find each other. If that time is now, and the place is here, then let me know, and make yourselves at home.

Bookmark and Share

Güera asked “how does crowdsourcing differ from collaboration?”. I dare not offer an answer as such – I would need to ask around and find some consensus first – but I will offer some personal observations…

‘Crowdsourcing’ is jargon. It is one of those buzzwords that make it sound like people know something about the information era. We can infer crowdsourcing must be different to collaboration because people have long been able to collaborate, but have only recently started to crowdsource. Wherein lies the difference? If crowdsourcing is new, then the difference must lie in using new technology to combine the efforts of many people who would not otherwise be able to work together.

Really easy examples of crowdsourcing involve asking people about numbers, for example to make economic predictions, and then treating the average result as a new and useful piece of information in its own right. Of course, the same result can be achieved with a traditional survey. At the low end, crowdsourcing is no more than taking a survey, though it sounds sexier. Indeed, a lot of ‘crowdsourcing’ is nothing other than taking suggestions over the internet and perhaps letting people vote on whether they like suggestions that were taken previously. So there is little genuinely new here, except that the internet makes it a lot easier. At the low end, crowdsourcing is just a virtual suggestion box or an online opinion poll.

At the high end… well, that is one problem with the concept of crowdsourcing. The high end cannot be too high, because the idea is that a large number of people contribute to the collective task. People could collaborate before, without technology. Technology just changes the scale and spread, meaning more can participate in a collective activity. Having a large number of unpaid people volunteer to contribute to a task assumes a high degree of common motivation and that the task is straightforward and consistent enough that the people responding can be relied upon to do a consistent job. Tasks that involve subjectivity, specialized knowledge and judgement do not lend themselves to crowdsourcing. For example, think of how you might get people to draw a picture together. One children’s game involves drawing a person in stages, with the paper folded so each kid can only see their part of the drawing. One child draws a head and folds the paper so the next child can only see where the neck should join to the torso. That next child draws the body and the arms, and leaves lines to join the legs, before they fold the paper and pass to the last child. The last child draws the legs, ignorant of how the rest of the drawing looks. When finished, they unfold the paper, to see the funny result of three independent drawings being joined together to make one odd-looking portrait.

With crowdsourcing, everyone works independently, with nobody directing or coordinating what is done by each individual in the crowd. This is unlike collaboration, which involves some common agreement and push-and-pull about who does what and how the individual pieces are meant to join together. The best examples of crowdsourcing I can think of are where newspapers ask readers to help them analyse large volumes of public documents. Just such a campaign occurred in the UK when a Freedom of Information request resulted in the publication of lots of paperwork about politicians’ expenses. The general public could be relied upon to analyse the expense forms because all they needed to do was read something and point out if they found anything interesting. A high degree of conformity could be assumed in what people found interesting – people care if politicians make big and/or dodgy expense claims. But looking at the current progress of that particular crowdsourcing project is a reminder of the importance of maintaining the interest of the crowd. The documents were made available last year, but at date of writing, the crowd is still less than half way through reviewing them!

Collaboration does occasionally exist in revenue assurance, but it is not the norm, and I increasingly doubt it should be. The obstacles to collaboration are difficult to overcome; they make crowdsourcing of RA impossible. Just take a look at the opinion polls about how much leakage the industry suffers. Poll twenty people new to RA and they will tell you leakage in their business is terribly high. Assuming they stay in the same job for the next five years, imagine asking them the same question after that time. After that length of time, some might say their leakage is high; others will be less keen to point out that five years of their life seemingly made no difference.

In its early days, people shared their RA knowledge relatively freely. They did so just because they did RA and they wanted to talk to others about doing RA. Those were more innocent times. Things have changed, though some people continue to be innocent to the point of naivety. These days, RA is normalizing on a different model to develop and propagate information – the buying and selling of information. If collaboration can be rewarding because of the give-and-take with peers, imagine a scenario where most want to take, or where most want to receive knowledge so it can be validated against some pre-determined scale. Those are not fertile conditions for collaboration, and certainly not for crowdsourcing.

Collaboration and crowdsourcing imply the group works to create something new and valuable. The free market can also promote the creation of something new, but in the free market, people create not to share but to sell. Creating in order to sell discourages collaboration, unless collaboration also involves a sharing of the rewards. That is why you will never see a GRAPA publication marked as ‘copyright GRAPA’. They are all the copyright of Papa Rob Mattison. But we should not be surprised if people try to spice up a business model with some old-fashioned appeals to a community. The clue is in the word ‘crowdsourcing’. It is derived from the word ‘outsourcing’. In many cases, crowdsourcing should be called ‘freesourcing’. If you outsource, you expect to pay your supplier. Crowdsourcing assumes that if you ask enough people, some will volunteer to work for free, out of altruism or a sense of common interest. Hence it is inevitable that some – whether newspapers or self-proclaimed gurus – will appeal to the notion of community in order to supplement their essentially profit-driven business model. There is no question about why they do it: they get something for free, and they can use it to make things that can be sold at a profit. There is only one question that needs to be asked: what benefit does the crowd receive? That is a question where I would like to see the crowdsourced answer…

Bookmark and Share