The onslaught of programmatic in recent years has raised issues and benefits, however, marketing guru Gil Snir says many agencies are fluffing out the media buying technique, making it confusing and overwhelming.

“Since 2015 we threw over five billion dollars into digital advertising – that’s a 20 per cent leap from the year before,” the chief marketing officer at programmatic digital marketing agency Benchmarketing told the company’s annual Game Changers afternoon in Sydney on Thursday.

“Clearly we’re putting more effort into this area, but I see some big hurdles happening and some of the culprits tend to be the large agencies.”

At the talk, Snir outlined seven areas – his Seven Deadly Programmatic Sins – that were told too often and overcomplicating the matter.

Viewability

Viewability is one of the top priorities for media agencies this year and yet Snir called out the absurdity of the current way we measure viewability. At the moment, an ad is considered viewable if half the pixels are on screen for at least one second, two seconds if it’s a video ad.

At the Game Changers event, Snir flashed a Facebook webpage onto the screen for one second then asked the audience who noticed/remembered the ads. No one put their hand up.

He did the same with a native programmatic ad – an ad that is supposed to blend into the site’s content – with similar audience results. He said viewability was like the proverbial tree in the forest. “I ask the question: If an ad loads in a browser and no one is there to notice it, does it really matter?”

And added: “The point here is that we’re really starting to digress into metrics that don’t matter.”

Snir isn’t alone in his viewability viewpoint. Marketing director of travel website Expedia, Vic Walia, said he also fell into the trap of thinking viewability matters.

Still, many others believe it’s a critical metric for digital advertising. IPG Mediabrands and GroupM have launched their own viewability initiatives for clients, and industry body IAB (Interactive Advertising Bureau) says it’s important, just the technology isn’t there yet to measure it effectively.

Ad fraud/brand safety

Another two hot topics in adland many agencies and advertisers are concerned about is ad fraud and brand safety. Snir said that while the bots in ad fraud eat up impressions, giving inaccurate results, if the objective is pushed back to something beyond an impression, ad fraud doesn’t really matter much.

“If you set a goal that occurs beyond a click – like a point of engagement or lead or sale – what happens to these little online gremlins out there trying to eat up all your ad dollars? If you ignore them, they simply go away.

“If you optimise to a point after an impression or click, ad fraud matters a lot less.”

Brand safety is low on the priority list for marketers and agencies alike, according to a report from the Audited Media Association of Australia (AMAA). However, Snir said a lot of the rhetoric about brand safety is coming from agencies, and when he and the Benchmarketing team investigated it, Snir said they found many of the big advertisers in the UK were using un-safe brand networks. The analysis was conducted using online service Similar Web – a platform allowing the comparison and analysis of various websites.

“It highlights the point that you need diversity and you need a mix of channels to be able to test these things,” he said.

Time-based programmatic

We used to think about marketing in campaign timelines, said Snir. However, now much of the chatter is about ‘always on’.

“If you believe you can sell products always on, then you should be running an ‘always on’ strategy,” he remarked. The idea, he said, is to still focus on timely events – Christmas and mother’s day for example – but also to establish a baseline for strategy.

“Always on means always on accountability,” he added. “You can always be asking what happened and what didn’t.”

Not disclosing attribution models

Snir believed the biggest type of ad fraud is attribution ad fraud. An attribution model refers to rules, or a set of rules, that determine which touchpoints lead to someone buying the product. Snir said many marketers don’t know what the attribution model is when it comes to their programmatic offering.

The issue for Snir however is that there are some agencies that use a combination of attribution models – post click (where a consumer clicks on the ad) and post view (where a consumer has just viewed an ad) – to determine the industry standard.

“I’m not saying that’s necessarily the wrong metric,” he stressed, “because it depends on the product and price on the sale side, but it’s really important for you to know exactly how it’s being measured.”

Reporting magic tricks

When advertisers and marketers are handed reports from agencies, Snir said he often sees little tricks to ensure a positive report, such as missing columns, graphs without trends and selective time-frames, only showing the good moments and “hiding negatives”.

These are some of the red flags he wants marketers to be aware of when receiving programmatic reports.

Advertisers using their own technology

The rise in programmatic has led to many agencies and brands developing their own trading desks. However, if an agency is using its own capabilities for a client, and says there’s a mandate to only use certain vendors, there’s a “misalignment of independence”, meaning advertisers need to trust the agencies have their best interests at heart.

It’s a similar argument the founders of programmatic company Real Programmatic have been arguing. Justin Boersmna and Jonathan Despinidic penned their thoughts in the December/January issue of B&T a few months ago, claiming there’s many agencies ripping off clients when it comes to programmatic.

Yes, it’s a complicated landscape, but Boersmna and Despinidic said CMOs should at least try to understand the landscape.

Setting the bar low

The biggest issue for Snir however is agencies setting low expectations for programmatic. “It’s a really big problem because programmatic advertising can be a very, very powerful tool.”

“We can’t test, learn or evolve if we don’t know all the ingredients.”