Shining a light on the darker side of bid shading

Danny+Spears.jpg

In this piece originally written for AdExchanger, Danny Spears, our COO, talks about what this means for publishers’ businesses.

Two months into 2020 and it feels like we’re at a real inflexion point for programmatic advertising - and in a very good way. Thanks to the commitment from advertisers and publishers, the conversation surrounding the need for greater trust and transparency in digital advertising has never been louder. 

With better practices rapidly replacing the persistent problems of the past, we must call out and challenge those things that stand in the way of a brighter future. As a starting point, we need to shed some light on bid shading - and in particular how the technique used by some SSPs is negatively impacting publishers’ business.

Understanding the principles of why SSP bid shading is wrong doesn’t require a technical mind; the SSP by definition of its role in the supply chain should always be acting in the publishers’ best interests. However, their use of bid shading allows the SSP to arbitrarily reduce the price of a publisher’s inventory in a way that is not only self-serving, but also uses the information privilege a publisher has made possible under contract, against that very same publisher.

It’s time to get really under the skin of bid shading and understand what can be done to mitigate the damage SSP bid shading is doing to publisher monetisation. 

Bid shading is a service to buyers

So what is the problem with bid shading? Nothing - if you are on the demand side of programmatic ad trading.

Bid shading was invented by DSPs as a way to protect their customers from price inflation as the market moved from second-price to first-price auctions. Today it is just one aspect of their wider bidding logic. 

To inform bid-shading, the DSP uses historic trading data to predict the minimum successful bid price for subsequent buying opportunities. The DSPs bid values are then progressively lowered to reach optimal price whilst maintaining win-rate. 

This makes a lot of sense as means for the DSP to deliver value to its customers.

SSPs become conflicted (again) 

Where the move to first-price was hailed as the solution to corrupt ‘second-price’ auctions that never were, SSP bid shading lands publishers in an even worse position; with an even murkier ‘first-price’ auction that never is.

The trouble began when SSPs – contracted to work for publishers – began offering bid shading as a service to buyers. Consider bid-shading as a repackaging of buy-side fee logic; a means through which the SSP adjust buyers’ bid values without the publisher’s knowledge. The main difference with bid shading is that the value extracted is passed to the buyer, rather than it being pocketed by the SSP. Either way the value is unwittingly taken from the publisher for the benefit of others.

There is an obvious conflict in SSPs attempting to serve customers with opposing interests on either side of the trading relationship. But the real concern for publishers must lie in the specifics of how the buyer service is delivered.

When an SSP applies bid shading, the SSP is exploiting its privileged position with visibility of the publisher’s own economic data; clearing prices and the minimum ‘floor’ price they are willing to accept. Note: Google does similar to this through its ‘minimum bid to win’ feature ostensibly to improve auction transparency.

The inevitable response from the bid-shading SSPs will be “this benefits the publisher“ as was their response with bid-caching. The truth is of course very different. Whilst publishers won’t see any increase in their overall monetisation as a result of bid shading, what they will observe is share of business move between SSP partners at an ever decreasing yield. As this occurs, the publisher’s yield management strategy is undermined and monetisation hit. 

Individually rational. Collectively insane.

SSPs introduced bid shading in an attempt to buy market share. In reality, the impact of bid shading is deflationary since it compels SSPs to undercut one-another on media price, precipitating a race to the bottom. 

Over time, were SSP bid-shading to become accepted practice, it stands to reason that vendors’ share of spend will quickly return to its original state. Except that the original volume of business will be transacted at much lower yield. As SSP take tumbles as a result, the logical conclusion is that SSP bid shading will be just as damaging to the future health of the entire SSP category as it is to their publisher’s health today. 

As publishers watch on, the frustration I share with many is that it’s the same vendors that they desperately want to see forge a position as a trusted alternative to Google’s monopoly who are driving this death spiral. The only beneficiary is Google. 

Publishers, size your exposure

If you’re a publisher you need to take steps to understand the cost of SSP bid shading on your business today. 

Here are some questions you might want to ask the SSPs you work with:

  • Do you ‘bid shade’ or otherwise adjust the buyer’s bid price?

  • What of my data are you using to inform bid adjustment logic? Consider clearing prices, ‘live’ bids, floor prices, domain/page information and/or user data

  • How does this occur? Is the reduction applied before or after the auction has closed? i.e understand in time-series. Watch out for application after the auction has closed. 

  • How does the value get passed back to buyers? At bid level in real-time, or as an accrued ‘value’ payment/rebate?

  • What % of my transactions through your platform over {period} have been treated with bid shading?

  • What is the average CPM reduction that you have applied to affected transactions?

This approach will help you calculate the cost of bid shading on your revenue.

It’s time to clean up, really.

SSP bid-shading needs to end; if not motivated by the SSPs duty to its own customer, then for the sake of the SSPs own health.

It will take SSPs agreeing to outlaw the practice in much the same way the UN led its members around the world to withdraw their use of napalm. This is a weapon that might deliver fighting advantage but whose perceived benefit is eclipsed by the unacceptable destruction and suffering it is known to cause. Granted this is an extreme analogy, but it will require similar collaboration and conviction between competitors to bring about change.

In the meantime, publishers must act to understand the extent of bid-shading in their supply-chain and to quantify the cost to their business. With this insight, the publisher can take a position - whether that’s blacklisting certain vendors, handicapping bid-shading SSPs (to reward clean partners) or by addressing the issue through contract terms. 

At The Ozone Project we’ve decided to take the latter path; we only contract directly with SSPs who commit to abstain from bid shading (and indeed all other forms of auction manipulation). This is just one of the ‘best practice’ commitments we require of Ozone Ad Partners and which we make available for all to see here.

Meanwhile, the realisation has dawned that there is something far more important than whether the auction is second or first-price. Without confidence in the auction being clean, this argument is academic - putting an end to SSP bid shading is a necessary step towards this.