This article is the fourth in a weekly series that focuses on issues that (should) matter to fraud and assurance practitioners. Each post is scheduled for a consecutive Wednesday, and by comparing the response to all ten posts I intend to rank each topic by seriousness, and so perform a ‘reverse’ opinion poll. Today’s article is not one that telco employees tend to think about much, because they cannot influence it. In contrast, vendors think about the topic a lot, though they can only fight their own corner. In any market there should be an optimal number of suppliers. If too few compete for business then they may lack the motive to cut costs or innovate, and so improve the quality of their offering or their price. If too many compete then most will suffer from having inadequate resources, and their priority will be short term survival instead of long term investment. So what is the right number of vendors for RAFM, and do we currently have too many or too few?
The current trend
This analysis could begin in many different places, but perhaps the most useful starting point is to observe market trends. We have fewer vendors than before, and it is likely that the number will continue to decline. Some vendors have been bought by others. Some have gone bust. Some have gradually declined and either subtly changed their focus over time or were absorbed into a much bigger business that could lose them without anyone noticing. The decline in the number of vendors will be underestimated by many; a vendor who makes few sales can exit the market without attracting attention.
The best way to understand the context is to look at the wave of liberalization of telcos across the world, and then to look for spin-offs and start-ups that followed that liberalization. Over time telcos tried to solve fewer of their problems with in-house solutions. Because there was growth in the number of customers, vendors were more likely to develop and market solutions on a commercial basis, and that was true for the supply of RAFM solutions too. But not every firm that entered the RAFM market (or at least dabbled in it) has chosen to remain.
To pick one example of a vendor discretely exiting the market, when Basset was acquired in 2014 probably very few still thought of them as a player in the RAFM space. Years before they landed a big deal with the Millicom group, and so could reasonably hope that more deals would follow. Anyone with sufficient experience would know they tried to sell their product more widely. But they never did. So ultimately they stopped trying, and worked on other products that were more lucrative for them.
Another factor that reduces the number of competitors is the coming together of RA and FM technology. I only recently adopted the combined acronym RAFM because there used to many vendors who did one but not the other. The split between RA and FM continues to exist in the org charts of many telcos, but is now rare when it comes to vendors. A few offer highly specialized services that are only relevant to one half of the market, usually the anti-fraud portion. For example, there is a long list of niche players that offer test call services designed to detect bypass. Test calls could also be used to do revenue assurance but only a minority also adopt that approach. In general, the same underlying technology can be used to tackle both fraud management and revenue assurance. So whilst start-ups may have begun life by focusing on a subset of the RAFM space, they have tended to expand their portfolios until they can all claim to deliver similarly comprehensive functionality that will interest managers on both sides of the RAFM divide. And obviously this creates opportunities for cross-selling and for offering keener prices to telcos who buy both the RA and FM toolsets. This bridging between RA and FM vendors inevitably put pressure on the start-ups that grew least, forcing them to exit a market where they no longer had any realistic prospect of growth.
There are some new vendors entering the space, and the wider you define this market the higher the probability of finding a recent addition. The competition with fraudsters lends itself to innovation, which creates new opportunities, and predictive analytics could arguably be included within this domain. But we should avoid the mistake of paying all our attention to start-ups that loudly announce their arrival whilst we ignore the former start-ups that have quietly slipped out of the back door.
The trend overall is downward, and that is no surprise when you analyze profits. Nobody can sensibly measure the profitability of the smallest vendors, but trends for the bigger firms can tell us plenty. There have been good years, but few have made sustainable returns for their investors. To their credit, Subex keeps grinding out profits, but it does feel like a grind. WeDo’s owners probably place less emphasis on profits as they maintain a posture of pursuing a strategic growth plan. And firms like Cvidya which had to generate profits were unable to do so, leading to a takeover by Amdocs which resulted in job losses. Probably Amdocs will milk their investment as much as they can, but their goal is only to retain the rump of Cvidya’s business whilst abandoning any former hopes of significant growth.
Less is more?
Given the trends identified above, there is a good argument that everyone in RAFM would benefit from further reduction of the number of vendors competing in this space. Lots of promises were made to investors, not least by VC-backed firms like Cvidya. Those investors are not going to keep throwing money at enterprises which cannot pay them back. So future investment in RAFM innovation has to be sustainable. In short, if we want continued development of RAFM technology it will only come by the reinvestment of profits generated by the vendors. To be profitable, those vendors will have to be able to command high enough prices, not be engaged in a fight to the death with many over-promising price-cutting sharks.
Scale also generates a significant benefit. For software, the relationship between revenues and costs is not linear. The cost of developing software is what it is, whether it is sold or not. So the vendors’ profitability rises with revenues. On this basis, it could be argued that reduced competition between fewer vendors would be in the interest of their customers, because more can be recycled into continued development of products. Having a fourth or fifth major alternative may not greatly reduce the prices paid by the telco, but having only two or three choices would lead to significantly higher profitability and so the chance for more competition based on innovation and the quality of the product, rather than the price.
But what about the cost of the technology itself?
A counter-argument begins with the observation that the cost of technology has fallen dramatically, and is likely to keep falling. The cloud, Big Data, software as a Service – there are many forces leading to cheaper innovation of technology. Vendors like ATS have been doing revenue assurance for a long time, but the way they do it has changed dramatically. They now harness technologies that did not exist when they began. The market place might be a lot poorer for the loss of agile smaller firms who have good reasons to change their approach to business and technology and may find it easier to cope with change than a bigger firm would.
The players on the periphery
Another challenge to this analysis brings us back to the question of who should be properly included in our list of RAFM vendors. In the case of Amdocs purchasing Cvidya it is clear that this has not changed the number of suppliers – Amdocs never pretended to be an RAFM vendor – even if it means the downsizing of one of those vendors. But Cvidya is a drop in Amdocs’ ocean. Nobody in their right mind is going to start calling Amdocs an RAFM vendor, just because RAFM will be such a negligible proportion of their overall revenues. So that leads us to rightly ask whether other massive firms should also be counted amongst the number of vendors. Consider Huawei, who employ 170,000 people. They could have two employees doing RAFM, or 200 doing RAFM, and it would be very hard to tell the difference, especially if most of those RAFM sales are bundled with bigger projects. We tend to think of RAFM solutions as being separate to other systems, but code is code. They could have code that satisfies many RAFM goals without needing to transparently separate that code from other systems whose primary purpose is not RAFM. So when we consider the number of vendors, there will also be players on the periphery who may or may not be satisfying a significant portion of the RAFM market, but where it will be hard to judge if they are players or not.
Huawei is a big telecoms firm. The other kind of major periphery is with technology firms who can offer RAFM solutions though most of their customers are in other sectors, like banking. The phrase ‘revenue assurance’ is not widely used outside of telecoms, but clearly there are many anti-fraud firms focused on other sectors, and who are open to adapting their products and services to suit telcos, depending on the profitability of the deals they might be able to land in practice. An example that comes to mind is FICO, who have a history of selling their products to telcos but who generate most of their revenues by supplying to the financial services sector. Whilst a firm like Huawei might try to satisfy the RAFM needs of some customers without making any conscious decision to invest in development of RAFM tools, a firm like FICO will likely sustain investment in the development of their solutions. Even if FICO fails to generate sustainable profits from its sales to telcos, that does not mean it will not sustainably invest in the development of RAFM tools that could potentially be used by telcos. If we couple this to the chances of technology becoming cheaper and more flexible over time, it means that telcos may be able to choose from a wider range of suppliers and see the benefits of sustained investment in technology, even though some of that technology was paid for, and aimed at, businesses in other sectors.
A lack of a conclusion
On some measures, telco RAFM suppliers should be moving towards the stability expressed by Bruce Henderson’s ‘rule of three and four’. Henderson postulated that:
A stable competitive market never has more than three significant competitors, the largest of which has no more than four times the market share of the smallest.
WeDo and Subex are the obvious first two competitors, but there would have to be a lot argument about who comes third, with all of the following firms being in contention:
- TEOCO, a telecoms-oriented firm which generates more revenue than WeDo or Subex but which hardly promotes its RA offerings because that would distract from its strategic focus. However, it may be doing enough assurance-type work to be ranked the third largest vendor, if only they cared to analyze their revenues that way.
- Neural Technologies, a firm with a solid pedigree in fraud management that has broadened its products to cover RA as well, and which has obvious aspirations to grow. However, their reported revenues are lower than estimates from some rivals, and it is not clear how much of their revenue comes from the financial services sector.
- The business formerly known as Cvidya, which probably occupied third place in the market between 2010 and 2012, but whose revenues declined sharply, and is likely to continue to decline now they are under new management with different objectives.
- The business formerly known as Razorsight, an analytics firm that started life as a competitor and offshoot of TEOCO. They certainly sold RA, though their concentration on the North American market and the blurring of the lines between their analytical offerings makes their market positioning unlike most ordinary RAFM firms, impacting the reliability of a like-for-like comparison. Their revenues were unknown because the firm was privately owned, but their acquisition by Synchronoss Technologies in August 2015 gave them a similar base valuation to Cvidya and included performance-related payouts which if satisfied would raise the implied value.
My speculation is that these four firms are below the boundary for the second part of Henderson’s rule, each generating RAFM revenues that are less than 25 percent of WeDo’s. But that suggests the current situation cannot be stable because the market share of the largest firm is much smaller than Henderson anticipated, and more firms need to exit the market in order to deliver stability. In this context, perhaps it is telling that two of the businesses listed have been purchased and absorbed into a much larger company. Or maybe the mistake is to think there is an RAFM market in the first place. Maybe we should define the market more by the technology being sold, and less by the use made of it, though that would make it even harder to fathom what market we should be analyzing.
On the other hand, perhaps the analysis about room for small agile businesses and players on the periphery is more important because Henderson’s rule of three and four does not apply to the telecoms RAFM market. Boston Consulting Group revisited the rule of three and four to see how well it applied to present day markets, and they concluded:
The rule of three and four does not seem to apply to the growing number of more dynamic, unstable industries, such as consumer electronics, investment banking, life insurance, and IT software and services.
That means there may never be a correct number of RAFM vendors. Instead we may be faced with what BCG calls ‘rising turbulence’, which sounds like an apt description for suppliers faced with addressing the risk and assurance challenges of telcos who are themselves confronted by a wave of disruption.