The hidden cost of a ‘set and forget’ programmatic strategy.

When most agencies talk about the importance of automation within their programmatic strategy, often what they are really referring to is a ‘set and forget’ approach which is entirely dependent on the standard off-the-shelf bidding algorithms that most DSPs offer. They will pull their report in-platform and see that ROAS and CTR were both up week-on-week whilst their eCPM remained stable. On the face of it this is a successful campaign ticking all the boxes, however it’s not until you dig deeper that you start to understand what is really going on.
‍

The model is learning from vanity platform metrics that don’t deliver any real-world value.

Metrics like CTR, CPA and ROAS have been the industry’s bread-and-butter for years. They’re the metrics clients are drawn to immediately on a media plan and the stats plastered all over weekly reports and quarterly business reviews.

Good-looking dashboard doesn’t always mean good advertising.

When traders set up a campaign in their chosen DSP and switch on that click algorithm because they have been given a CTR goal, more often than not they have absolutely no regard for the supply that algorithm will be buying.

At best they might have some kind of broad ten thousand site inclusion list applied, at worst they’re reliant on an exclusion list which hasn’t been updated since 2018.

Very rarely will they have hand selected three to five of their preferred exchanges to run the campaign on (no one needs to be targeting fifty exchanges). Very rarely will they have built a carefully vetted inclusion list specifically for that advertiser. Very rarely will they have selected to buy from authorised direct sellers only.

Very rarely will they be using independent analytics and ad verification tools such as FouAnalytics to truly optimise the quality of their supply and eradicate ad fraud from their supply path.

DSPs will quickly learn which bid requests you’re responding to and send more of the same. The algorithm will learn which supply is ‘performing well’ against the goal it has been set and will start to prioritise that, and before you know it you have a self-perpetuating cycle of high CTRs and a poor supply path.

The algorithm only knows the goal you give it.

These are all very simple controls that traders have at their fingertips in any DSP but when they choose not to use them it’s a sure fire way to end up with a site list which, at best, is propped up by low-quality gaming sites and murky supply chain intermediaries and, at worst, ad fraud.

None of which is delivering real-world business outcomes for clients.
‍

The model is overly reliant on user IDs.

Surely when the model is geared towards conversion none of these guardrails matter, right?

This is another common misconception and where many agencies and traders fall foul. This belief that all conversions are made equal is fundamentally not true and misunderstanding this can lead to a lot of wasted media budget.

When traders are given a ROAS or CPA goal and they activate that conversion algorithm in their chosen DSP, the path of least resistance for the DSP is often user IDs.

But what does this mean?

It means users who are part of some kind of audience list. This could be a third-party intent audience or a first-party site visitors audience, either way they are a part of that list because they have already shown clear signs of conversion intent.

This leads to over-exposure to existing customers and non-incremental customers i.e. customers who were going to purchase whether they saw an ad or not.

Whilst re-engaging existing customers has an important role to play in any brand’s marketing strategy, true long-term growth comes from acquiring incremental new customers.

The path of least resistance isn’t always the path to growth.

The other impact of ID-obsession is over-exposure in more addressable browsers such as Chrome and under-exposure in more privacy-constrained browsers such as Safari.

Many programmatic teams will actively exclude Safari from their campaigns because this environment will not satisfy their vanity platform metrics. Whilst this might be true, in doing so you are actively excluding a significant portion of potentially high-value customers.

With a lack of user IDs in Safari comes less competition. Combine this with the fact that many programmatic teams choose to avoid it altogether and what you often get is the same quality programmatic supply at a lower price.

At ByAtlas it is not uncommon for us to see CPMs between 50-70% cheaper on Safari vs Chrome for the same quality supply.

What we also find is that whilst attributed conversions might be lower in-platform when targeting browsers such as Safari, MMM and incrementality measurement partners often credit these strategies with ROAS numbers sometimes 75-100% higher than their Chrome-targeted counterparts.

Quality first. Performance follows.

Programmatic campaigns should be optimised for quality and performance will follow.

This is such a simple concept but one that I have seen overlooked so many times over the years.

When you’re so focused on legacy performance metrics it is easy to forget about what is actually important.

Clients will quickly begin to see programmatic delivering real value for their brand when their campaigns are optimised for quality before anything else.

‍