With the telemedia industry facing its own ‘global’ crisis driven by poor operator reaction to the growing problem of ad fraud, Declan Pettit, Commercial Director of Monitoring Compliance Partners (MCP), outlines how the issue can be tackled
Some of you may be aware of the excellent book by James Rickards, ‘The Road to Ruin’ in which he postulates the outcome of the next global financial crisis – if approached by the ‘Global Elites’ in a certain way. The way, he suggests, to deal with any virulent pandemic is to either create the antidote or, in its absence, the virus needs to be quarantined.
Through the economic jigs and reels of Neo Keynesian ‘interventionism’ versus Friedman’s monetarist, ‘laissez faire’ view, Rickards uses the basic thesis of a novel called ‘Cat’s Cradle’ to make his point. The author of that book creates a substance call ‘Ice Nine’ – a solution that rearranges the molecules in water, rendering anything it touches to ice, irrespective of the prevailing air temperature. It was no coincidence that this dystopian novel was written in 1963, just after the Cuban Missile crisis – the ‘Nuclear Winter’ scenario.
The point is; the smart move is to isolate the cause of the virus – rather than its effect – until an antidote is discovered.
Our industry is going through turbulent times currently – our own global crisis. With mobile customers having to run the gauntlet of ever-increasing types of fraud, the Network Operators are naturally concerned about consumer harm – and it shows…
Rather than approaching the problem in a unified way that targets the problem at source – such as the rogue Affiliates perpetrating the fraud – the Mobile Operators are mandating sweeping changes with abandon. These changes not only profoundly damage revenue lines of the majority of well-intentioned actors in the value chain, but also stymie progress in the on-going battle for market share against non-mobile payment tools.
Whether it’s banning traffic sources, increasing hurdles to payment flows (PIN, Captcha…) or excluding certain types of services from being billed on their networks, these knee-jerk actions are in danger of throwing the proverbial baby out with the bath water.
Granted, a small percentage of consumers are being billed without their consent or knowledge – and yes, the majority of advertisers are also unaware and helpless in the face of it, but is a ‘gun-shot’ reaction appropriate when a more specific ‘rifle’ solution would deal with the issue more effectively – and with far less co-lateral damage?
The problem is the Mobile Networks are desperate to protect their brand names – which is understandable considering that the source of most of their bad press – mobile billing – represents a tiny percentage of their overall revenue.
However, actions are being taken by the VAS/wholesale middle management, who, under threat from senior management, have ‘consumer satisfaction scores’ and complaint levels as their KPIs – rather than revenue/innovation.
Fraud, like any virulent pandemic, is a universal issue: once a particular variant of an APK file leads to a successful auto-subscription works in one territory, this is quickly exported worldwide via surreptitious trading exchanges to local bad actors who know how to monetise in their local geo.
And the reaction from Networks is also a universal problem – the only difference being the voracity of their reaction varies from territory to territory. Indeed, there seems to be a regional paradigm to the reaction, which makes sense as Mobile Operator group ownership tends to be regionally based.
MENA, for instance, is a good case in point. The likes of Viva, Etisalat, Zain and Ordedoo are all very strong in the region – hence group HQ dictats prevail and bad news travels a lot faster.
So back to Rickards’ thesis: how do we isolate the source of the problem, while minimising the collateral damage to the value chain? Clearly, there is no panacea but the following blueprint will help:
- Discover and capture as many fraud instances as possible in order to track trends through data analytics.
- Identify the various flow participants and in particular the responsible affiliates (easier said than done as they constantly change their IDs). To do this effectively requires close value chain collaboration.
- Quarantine them by exclusion from white list of ‘accredited sources’
- Reward advertisers who use only these sources by allowing them access to a more frictionless click payment flow rather than OTP/Captcha barriers.
- Advertisers can clean their own affiliates lists via the official accredited white-list (without accessing it as all will want to protect their successful, ‘clean’ sources).
- Implement a real-time anomaly detection tool which identifies fraud ‘live’ and gives advertiser option to reject source. Whether rejected or not, monitoring system will provide proof of maleficence (or otherwise) and take appropriate action with whitelist.
- Continually monitor these accredited sources in the wild – and alter their accreditation status up/down dependent upon their record.
Whilst I would follow Friedman’s economic ‘laissez faire’ view more easily than Keynes’ interventionist view, there are times when a proactive, harmonised industry step benefits the whole value chain in the long run.
Author
Declan Pettit is Commercial Director of MCP (Monitoring Compliance Partners). They specialise in monitoring and anti-fraud solutions in Mobile billing industry.
Buy or build – trick is knowing which one to choose in any given circumstance
In October 2017, MCP announced the release of its Federated Kids Database (FKDB) – in support of a cross Industry initiative to help Advertisers block inappropriate advertising in children’s digital channels; with the key aims being the lowering of related Industry complaints to mobile operators and the regulator PSA.
As of early February 2018, the implementation results to-date have rendered a 20% drop off of inappropriate ads in kid’s apps (month on month November to January). MCP’s database now includes 20,000+ kids’ apps and2500+ YouTube kids’ channels have also now been added. The company captures and loads new kids’ apps and YouTube videos and channels daily so clients can remain current and up to date with the latest block list in real-time and all PRS advertisers in the UK now have access.
Toby Padgham, COO of MCP explains: “This industry collaboration has and will continue to grow FKDB as the default Industry block list, building a bigger and increasingly intuitive database to help advertisers block kids’ apps and sites. We thank all merchants for their input, and particularly the significant data contributions to set up the ‘federated’ list – coming from PM Connect, Buongiorno, Alchimie, Digital Virgo and Creative Clicks.”
Proposals to build ancillary databases through the purchase of apps lists and combine these with the core FKDB are welcomed because this can only support the common aims cited above. However, it is worth managing expectations, because there is no magic big kids’ database available for purchase now.
Proposed methodologies to ‘fingerprint’ a kids’ app and apply it to a list of the top 500,000 apps are likely to take time and the outcomes of generating x% (e.g. 7%) of accurately defined kids’ apps are unknown.
Brian Pettit, CTO of MCP adds: “The process of categorising apps or sites as kids is mainly a manual one – requiring constant input from a team of publishers. The idea of simply buying a database (whatever the size), then ‘fingerprinting’ a kid’s apps list – and applying those fingerprinted characteristics across the whole database in an automated fashion – could cause various problems, not least automated addition will lead to a % on non-kids’ apps being incorrectly categorised and added to the blacklist – leading to subsequent revenue loss. Missed kids’ apps not being added to blacklist – leading to on-going consumer harm,” he continues.
The assessment process is, necessarily, based on categorisation criteria that requires a subjective/human view. Also by its very nature, the app store environment is prone to new apps trending and becoming popular very quickly, and so emphasis has to be placed on a method that pulls kids apps and sites daily and provides this to Industry for immediate blocking.”
Building a verified kids database is one thing, effective implementation to block is another. The coming weeks will determine whether effective block processes are attached to the database and whether they are applied by some committed providers or all providers.
Padgham, COO MCP summarises: “Over the past few months, understanding of the issue and building a solution has happened at pace. A successful outcome will be determined by the propensity for providers to adopt best practice and take action. It may require Code enforcement action by relevant authorities on any ongoing infringements to ensure all take their regulatory responsibilities seriously.”
