An Adweek story from early last week about bad internet traffic caught my eye, and before I could write up a post they followed it up with a little more detail. Unfortunately the outcome is less on detail and more on “we know something is up, but can’t figure it out.”
I’ve shared this same frustration in the past where I was working on a project and we decided step was to increase traffic by buying traffic through some networks that do this sort of arbitrage work.
These stories first center around some “suspect” content networks that are being credited with millions of page views a month. Good for them? Well the issue is that they really are not very sites and the fact that they are related though a central publishing group makes you question are there really users going to the site and generating clicks.
The problem that crosses the line is when these page views generate millions of page view X multiple ad units per page. When these numbers get tallied the revenue estimates for these sites again reach into the $millions.
Here are the stories:
- Meet the Most Suspect Publishers on the Web – The rise of ghost sites, where traffic is huge but humans are few.
- Alphabird, Digimogul: Tell Us Who’s Behind the Botnet – The CEO of Spider.io speculates on who is behind a massive Web fraud effort
- Spider released data on a new bot network dubbed Chameleon that Spider believes is costing advertisers $6 million a month.
My personal feeling is that traffic buying strategy should be an act of last resort or only used as a primary strategy if you absolutely need to generate traffic today (even that raises concerns of finding a better plan). Our experience is that if you find the wrong vendors they can serve you up the equivalent of traffic-crack where all your traffic loves your sponsored links. From there management loves the spike in revenue, and you’re stuck with no proven strategy that will maintain that new revenue.
I remember the addition of one vendor that caused our query traffic to go through the roof for “Rolex Watches;” suddenly, the whole world wanted a Rolex, and they were clicking on the top ad link for more data. You know it can’t be true, but – Cha-ching! — management loved it and the biggest issue became swapping in a plan to ween ourselves from the purchased traffic and that provided us sustained organic traffic. We were able to pull this off, but it took 4-6 months, which is not surprising for good content and SEO to get recognized, and many “trust me, this is better notes” to the people upstairs.