Potential advertisers sometimes ask me for stats about the popularity of Commsrisk. I always refuse to give them an answer. There are two key reasons why I have adopted this policy.
- I have literally no economic reason to tell the truth. Whilst I am moral, and do not wish to lie, there is no financial incentive to give away accurate information that is also commercially sensitive. If the real information would influence the advertiser’s decision, then misinformation is more likely to give me the result I want, with none of the downsides that come with sharing real data. And if that is true for me, it is also true for my competitors, making it likely they exaggerate, and encouraging an upwardly spiraling cycle of bogus claims about the popularity of websites. Rather than get sucked into a game where the most deceitful are also the likeliest to win, I prefer not to take part. And by refusing to answer, I can honestly tell potential clients that they are daft to assume any answer they cannot independently verify would be honest.
- Even if I did not invent the stats I wanted, it is easy to inflate the popularity of a website or an advert through meaningless traffic. All it takes to boost traffic is to set one computer to repeatedly ping a website. Though the action would be meaningless, this technique could be used to massively increase stats like page views, clicks or average duration of visits. It is easy to find criminals offering to sell popularity-boosting services like these, some of them relying on bots running on hacked computers. And it is not hard to engineer your own scam. I actually know one deadbeat who tried to sell a website by drawing attention to its very high Alexa ranking. Anyone with half a brain would have asked why a website about finding love and success was receiving all its traffic from inside London. In case you have not guessed, the deadbeat happened to live in London.
Jamie Bartlett recently pointed out in The Spectator that…
…as much as half of the underlying economic model of the free-services-in-exhange-for-ads system that keeps social media afloat could be fraudulent.
I recommend you read his article for a succinct, no-nonsense evaluation of the madness of the online advertising business, as backed by plenty of numbers about the extent of fraud. Bartlett has an explanation for why such crazy levels of fraud are tolerated, and it rings very true:
There are lots of reasons this isn’t getting sorted. No one wants to rock the boat too much, and plenty of people still get paid even if it is fraudulent. (Although advertisers always lose). Everyone in the chain seems to think it’s someone else’s responsibility. And for all its faults online advertising still allows far greater nuance and accuracy in terms of targeting a market of people you want to reach.
To appease potential advertisers, I usually tell them to look at the number of social media recommendations for Commsrisk articles, and then to compare the results to articles on other sites. I do this because I believe in following genuinely good practice, and it is good practice to rely on independently verifiable measures of performance before making a purchasing decision. (Though it does not help that social networks sometimes ‘break’ or deliberately turn off the APIs that share their stats.) However, whilst inflating the popularity of content on social media is significantly more difficult, I still expect it happens a lot. Just look at the video below, showing a Chinese ‘like factory’, where banks of phones are being used to automatically and repeatedly boost content on social media.