Notes on a scandal? The 5 things marketers really need to know about this week's ISBA programmatic report...

One of my favourite quotes about the financial crash of 2008 is from Charlie Munger.

"When you make a system vastly complicated so that only a few high priests in each department can pretend to understand it, what you’re going to find all too often is that those high priests don’t really understand it at all. Then all of a sudden the system goes out of control."

He has spoken widely about how the finance system became so complicated that bad actors found it easy to commit massive fraud under the noses of banks and regulators. The film 'The Big Short' details this problem too.

I think the worst parts of modern advertising resemble the worst parts of modern finance circa 2008 - unnecessarily complicated, driven by algorithms that nobody except ‘experts’ understand and fraud ridden.

Finance flew blindly into an enormous crisis after ignoring warnings and failing to adapt.

This hasn't yet happened to advertising.

Until now.

This week a bell-weather report from the ISBA and PWC sent shockwaves across the business world. When a story about marketing technology makes it into the Financial Times with the headline 'Half of online ad spending goes to industry middlemen' you know something big has happened.

But the headlines only tell half of the story. Programmatic has become an incredibly complicated, sometimes alarming topic for marketers who aren't knee deep in it day to day, meaning the report itself is incredibly detailed and nuanced. This means it's time consuming and difficult to figure out what it really means for your brand.

No CMO or senior marketer really has the time to go deep on programmatic advertising. That's a big problem. Because if you look at the headlines alone without understanding the nuance of the report, your knee-jerk reaction might be to switch off all your digital ad spend tomorrow.

But trust me. That shouldn't be your immediate takeaway.

So in this short piece I'll quickly outline the key findings of the report, point you to some of the best analysis and offer my own personal take on what to do next. I'll do my best to help you understand what are the real implications for marketers behind the headlines.

This is categorically not an attempt defend some of the behaviour that the report highlights. We should all be alarmed by the numbers, the opaqueness and murkiness offered up by an area that promised so much transparency, control and efficiency.

Instead, I want to help busy marketers who may not have the inclination to dive into this complex topic that's only becoming more and more difficult to understand as time goes by.

What to read:

Read the full executive summary of the report here. (The whole thing is only available if you're an ISBA member or can 'borrow' it from elsewhere, as I have done).

Read the ISBA and PWC's own analysis here.

Wayne Blodwell offers an excellent fair minded analysis from the pro-programmatic side here.

Campaign's Global Tech Editor Omar Oakes puts forward an excellent analysis here.

Former Ebiquity exec Nick Manning offers a brilliant perspective in Mediatel here.

The 5 key takeaways from the report:

1) Some advertisers’ budgets are being halved through fees

For the first time in the world ISBA and PWC have done a deep dive into the programmatic supply chain from all sides (advertiser and publisher). At a top line level they found:

• a 15% gap, money that no one in the system can account for which they refer to as an 'unknown delta'.
• 51% of total spend ultimately filtering through to publishers.

Previous reports have outlined similar findings, including a Canadian report entitled 'Programmatic: Seeing Through The Financial Fog' which outlined how an average of 72% of spend went to "working" media and 28% of spend was claimed by "non-working" activities, with exact ratios varying from 85:15 to 30:70 depending on the campaign.

The overall take-away is that few advertisers really know what they should actually be paying and what value they are getting.

However, there is a caveat...

2) Some of the supply chain costs incurred are legitimate

Many media outlets are reporting these findings as clear indications of fraud or malpractice. There's certainly a lot of at best opaque and at worst downright dodgy stuff going on. But to label any 'non working media' as 'lost' or a 'tax' isn't right either.

Firstly, in the above image, many of these supply chain intermediaries do add value to the process, just like in any business. Usually this cost is incurred to help brands improve targeting. So blanket labelling all supply chain costs as ‘tax’ or ‘value-eroding’ isn't correct. Some of these costs are value adding.

In some of these areas of expenditure marketers also have tighter controls – e.g., buy-side services, tech and costs such as agencies, trading desks, DSPs, data providers, ad serving and verification services.

The reality is the digital supply-chain simply incurs more charges. That's a cost of doing business at the moment. The problem is because there are so many extra mouths to feed, these fees takes up nearly half the initial budget.

Secondly, and a separate point, even the 15% unattributed isn't being clearly labelled as illegitimate (yet):

In fact the report clearly states that "the unknown delta could reflect a combination of: limitations in data sets, necessitating occasional estimations; DSP or SSP fees that aren’t visible in the study data; post-auction bid shading; post-auction financing arrangements or other trading deals; foreign exchange translations; inventory reselling between tech vendors; or other unknown factors."

Basically there is likely some bad stuff, and likely some opaque good stuff causing these supply chain costs. This is a point that someone reading only the headlines would tend to miss.

The overall issue is that from a high level it's hard to figure out who's really adding value and who's not.

Which brings us to our next point...

3) The study itself actually suffered from the complexity of the ecosystem

The data collection ran from 1 January 2020 to 20 March 2020 this year. But the study owners could analyse only 12% of impressions served from study advertisers to study publishers, or 31 million out of 267 million. Why? The rest could not be mapped due to low data quality.

Basically the data collection process for the study itself actually reinforced its own critical conclusions. The supply chain is a complex, unwieldy mess that's almost impossible to parse.

Problems included the fact the 15 advertisers had nearly 300 distinct supply chains to reach 12 publishers, there was a lack of uniformity around data formatting and some supply chain intermediaries were not clear about what was required for permissions to access data.

Programmatic clearly suffers from a profound lack of transparency and business practices making it hard for brands to know where their money goes.

But not all formats are created equally...

4) Video performed better than any other format

The level of working media was found to be much higher for video formats.

According to the report, a greater proportion of advertiser video spend (65%) reaches publishers than display (54%) and the same is true of private marketplace spend (PMP, 54%) vs open marketplace (OMP, 49%).

This tells us that while no format is devoid of problems, certain formats and buying approaches are far more prone to issues than others.

So video tends to be a more premium product than display, which brings us to our final point...

5) This study looked only at the premium end of the market

This isn't a study that examined obscure long tail sites, weird apps created for ad fraud purposes or porn. In fact this study looked at the premium end of the programmatic market, involving some of the UK’s leading publishers, media agencies and best-known ad-tech companies.

So even when big brands are working with reputable supply publishers and vendors, 15% of budget is 'unattributable' or lost. That's incredibly damning.

What would the % of 'lost' budget could be when you get into the even less transparent, murkier sections of the internet?

This indicates the need for marketers and agencies to put better stopgaps in place (including white & blacklisting, quality context/viewability parameters and ad fraud tracking) to ensure they know where their ads are being served.

What you should think/do now:

• Don't come away from this with a blanket 'programmatic is bad' mindset.

For direct/activation activity (the 40% part of the 60:40 rule), programmatic buying can be extremely useful, particularly if you're using your own first party data. This report highlights some of the worst parts of programmatic (and indeed there's plenty worse than what's in the report too.) But with the right parameters in place it can still be a hugely powerful tool in your arsenal. The ability to trigger advertising based on location or weather, to personalise creative based on where a consumer is in the path to purchase or to generally use data to be more relevant to people is very powerful. In its ideal form, programmatic offers targeting precision, scalability, cost efficiency and real-time optimisation.

• Speak openly to your procurement department, agencies and other partners about transparency and incentives.

This is absolutely critical. I believe that part of the problem in programmatic is the unhealthy procurement led relationship that has developed between brands and agencies. This isn't just a programmatic issue, but a wider media and indeed creative agency issue.

In the above piece Wayne Blodwell outlines this perfectly.

"Poor procurement practices incentivises poor behaviour. These poor behaviours are often realised through untraceable off-grid commercial arrangements."

If you squeeze agencies so much that they're making no margin, the default response might be to find other, more nefarious ways to make money.

If you incentivise agencies based upon inherently short term, easily massaged metrics, then they won't invest in media that will build your brand profitably over time.

So if you're an advertiser you should engage openly with agencies to understand how they're currently remunerated, what they're optimising towards and most importantly the steps they are taking on your behalf to maximise the proportion of their spend that reaches publishers.

You also need to ask the direct question 'what % of my media is being siphoned off on fees and other non-working elements?'. A good agency will see that as a fair ask and have a clear, immediate response. If, for many advertisers only 51% of your programmatic media investment is actually reaching the publisher, you need to be sure of the value that the other 49% of your budget that goes on various agency, data & tech fees is giving you.

• Keep it as simple as possible, but no simpler.

Here's a good rule of thumb from my own experience with programmatic:

the more complex and difficult to understand it becomes, the more ineffective and fraudulent it truly is.

If you're working with an agency or a vendor that's bamboozling you, the likelihood is they don't understand themselves.

If you're reading this you're probably a smart strategic marketer so take this as an opportunity to strip back the complexity of your programmatic spend and start from scratch.

Work with someone to analyse whether you need that huge ad tech stack, whether you're really making use of your first party data and if a much simpler approach would actually yield better results for less spend and less time invested.

• Take advantage of the current situation to test and learn.

One big benefit of the timing of this report is we're currently in the biggest ‘zero based budgeting’ experiment in the history of marketing. In many categories media spend has dropped off a cliff. But that’s a massive opportunity.

Because it create a clean controlled environment for testing what happens when we start to switch programmatic media back on. It this creates favourable conditions for tracking the real impact of your spend. Use this time to hone in on non-working spend and try to understand what’s good ‘wastage’ and what’s merely waste.

In summary

The promise of programmatic was that it would deliver greater transparency, efficiency and measurability.

Instead what it has mostly delivered is unnecessary complexity, within which bad actors can hide.

There is huge value to be had if you approach programmatic with the right mindset and incentivise your partners correctly. Plus as outlined in the Campaign analysis above, "a healthy programmatic media system is good for brands, media owners and consumers, provided their privacy can be safeguarded".

But as it stands, the whole area is very messy.

Programmatic buying as a broader approach isn't going away. It's the infrastructure that the future of media buying will be built on. In fact over the next five years almost every media, from OOH to TV to radio, will be bought programatically.

The onus now is on the entire sector (including agencies, platforms, clients, middlemen and publishers) to clean itself up, become more legitimate/mature and start to apply more stringent regulations across all actors.

As Jerry Daykin said when the report came out, it's now time for change. We need standardisation, more transparency and better cross industry collaboration to solve these problems and re-inject this whole space with trust. This isn't about pointing of fingers, real action is needed.

This report should be welcomed by everyone involved in programmatic advertising (except bad actors who will now be flushed out into the open).

It's an incredible starting point to generate a wider conversation around the transparency issues and supply chain complexities inherent in the ecosystem.

As the saying goes, never waste a good crisis.

Because if the advertising industry doesn't fix this now, it will have enormous repercussions.

Just ask anyone involved in high finance circa 2007.

--------------

I hope this post helped you make sense of the biggest story in marketing this week.

If it did, please share it with others.

Note 1: This post is designed to be non-exhaustive and to offer cliff notes on a vitally important topic marketers without the time or ability to dive deeper. There's been plenty of commentary this week so if you've anything to add please do so in the comments below.

Note 2: This whole area is rife with biased commentary that doesn't help matters. So to declare my own bias, I work for a company that has invested heavily in (transparent) programmatic services and is recognised as one of the global leaders in the space. I think my above commentary is fair, measured and balanced. But you can make up your own mind on that!

Shane O'Leary

@shaneoleary1

Join 2000+ other marketers by subscribing to my email list...