Last time we covered why it’s OK to be afraid of programmatic. Now, let’s dive into the root of that fear – bad traffic – and how you can avoid it. The short answer is: tracking direct-response metrics is the only way to ensure traffic quality. On the flip side, tracking only engagement metrics is an open invitation for your staff and vendors to send you bad traffic.
Some Hard Truths
Truth #1: Bad traffic exists in the digital advertising ecosystem. It can be as nefarious as bots programmed to “engage” like humans, or as ambiguous as a misled human user who came to your offer, video, or landing page unintentionally.
Truth #2: Programmatic buying eliminates the need for media buyers and sellers to have a personal relationship. In the “old days,” the story goes, professional reputations ensured that sellers were accountable for the quality of the traffic they represented. The flawed perception is that since programmatic removes personal accountability, the evil traffic vendors have run amok. The reality is that anonymity has allowed buyers to misbehave as well.
Truth #3: Bad traffic exists in the digital advertising ecosystem because there is a market for it. This is not discussed enough. If you are buying traffic based on an “engagement” metric like impressions, click-through rate (CTR), video views, or bounce rate – and are pressuring your staff to pay your vendors as little as possible for that traffic – you’re likely buying fraud. And when you look in the mirror in the morning, you should chant the mantra “plausible deniability.”
Welcome to Canal Street
With programmatic traffic quality, accountability is a two-way street.
At a recent conference, one of the panelists discussing programmatic fraud said, “No one would accept paying for a fake Coach bag, why should we accept that we have to pay for fake clicks?”
— Adam J Epstein (@adamjepstein) September 17, 2014
This media-buyer-as-victim analogy does a great job of illustrating the very point that it is missing.
If you have been to Canal Street in Manhattan, you know that there is a thriving market for fake Coach bags. Buyers will pay a low price for a product that others think is real. A designer bag reflects status – and a fake bag can as well — so long as you don’t get caught. Likewise, a large volume of cheap traffic with excellent engagement metrics (look at all those video views!) is awesome – unless you get caught.
Media buyers buy fake clicks for the same reasons frugal fashionistas buy fake designer bags – to attain status (by hitting metrics) while saving money.
Fortunately, it’s not difficult to determine if you are buying bad programmatic traffic – provided you actually want to.
You Get What You Pay For
In programmatic, if you are putting in low bids and winning traffic, and you’re not taking steps to discover if your traffic is coming from actual people, you can pretty much guarantee that it is not. Essentially, you’re buying Coach bags on Canal Street.
Click-through rate, bounce rate, and video-view metrics are the most susceptible to fraud, especially if those are the only metrics you track. There have been some high-profile examples of programmatic traffic inflating video viewership because the high CPMs for “view engagement” combined with the low cost of bot traffic, provides an easy arbitrage opportunity.
If you want to avoid fake clicks, you must use direct-response, performance-based metrics that cannot be easily gamed by bots or confused users. If you are using the wrong metrics to evaluate traffic quality and campaign performance, everyone – from bad actor publishers to your own employees – will hit your metrics at the price point you demand.
Figure Out What You Want to Pay For
Direct-response metrics will vary by campaign, but credit-card purchases, subscriptions, and registrations are typical. Purchases and subscriptions are the best metrics to track – because a real person had to actually pull out a valid credit card to complete those transactions – but any activity that can be tied to human behavior can be effective.
Some advertisers in our system track form fills, but optimizing against these metrics is only useful if the advertiser also tracks continued interactions with these users. Ideally, your direct-response metrics should be analyzed side by side with engagement metrics. For instance, if you notice that your programmatic ad buy is generating excellent time-on site-metrics, but none of your direct-response metrics are performing well, your clicks are fake.
Bottom line: Every programmatic campaign needs to capture performance metrics that reflect desired human activity. Without these “hard” metrics, advertisers cannot truly know if they are buying Coach bags on Madison Avenue or Canal Street.
Image via Shutterstock.
Video consumption keeps increasing and Facebook is serious about a video-first world, encouraging us all to explore its full potential. Ian Crocombe, ... read more
Mike Andrews Ph.D is Chief Scientist (Forensiq) at Impact Radius, and is carrying out some fascinating work around digital marketing and ad ... read more
A new organization, The Coalition for Better Ads, has been launched to “leverage consumer insights and cross-industry expertise to develop and implement ... read more