Publisher considerations for partner consolidations

This article was originally published in March of 2023.

The cold air cloaking Newport Beach during last week’s Base.Camp felt fitting given the chilly state of the programmatic market. YoY CPM growth is declining, addressable audiences continue to shrink, once prominent SSPs are shutting down, and the only certainty is continued uncertainty around third-party cookie alternatives and timelines.

So when AdExchanger asked me to comment on the Jounce Media report that stated, “The average size of an Ads.txt file among the top-grossing 10,000 publishers has tripled between 2020 and 2022,” I made it clear that the trend didn’t surprise me. While it’s unfair to single out publishers as the sole reason for this increase, the path between increasing the value of a website's inventory vs. increasing the revenue you earn per session has never been murkier.

The +40% CPM increase that publishers benefitted from during the advent of header bidding resulted from improved auction mechanics that priced publisher impressions more fairly. That is fundamentally different from the boost in revenue per session that comes from exploiting price fragmentation through things like bid duplication and out-of-view refresh. While this distinction may not matter in affecting short-term business outcomes, header bidding was a sustainable leap forward because it reflected a marketer’s willingness to pay more for an impression while the latter is the utilization of a technology loophole.

The challenge with continued reliance on bid density and duplication is that the loophole is inevitably going to close with the advancement of measurement and DSP-driven QPS (queries per second) restrictions. The Trade Desk limited publisher supply paths in 2020 and has publicly discussed the importance of limiting QPS on earnings calls. Additionally as companies like Scope3 have highlighted, bloated publisher stacks create an outsized carbon footprint that is bad for the environment.

Some have argued the industry should first change its incentive structure before publishers are asked to adjust their monetization strategy. Or, this sentiment should be DSP-mandated and SSP driven instead of publisher-focused. While I understand the framework of that argument, ultimately, this isn’t a measure of blame. It’s an exploration around the sustainable opportunities and outcomes that a cleaner partner stack presents when combined with a publisher’s focus on ad quality and organic audience.

Partner Evaluation, Assessment and Paths to Consolidation

March Jounce Media data shows the average publisher is connected to 27 SSPs while authorizing 16 of those partners to resell inventory. Seller’s Guide data shows the average publisher has authorized 169 lines in their ads.txt file.

When this article was published, the largest website we run had 14 SSP connections and authorized 3 of those partners to resell inventory for a total of 34 lines in our ads.txt file. While that’s hardly sparse coverage, it places us 50% below the Jounce SSP number, 80% below the Jounce reseller figure and 80% below Seller’s Guide data on a site with an average session length over 21 minutes long.

While I’m pleased we under-index, it’s an outcome of our larger goal which is building deeper and more effective partnerships with enabled SSPs. I don’t contest the point that adding more partners can lead to short-term gains due to an increase in bid density. The logic of that strategy is widely adopted and in many cases, at least in the near term, will lead to a positive outcome. However, my position is that the benefit from that strategy, assuming it still exists, is decreasing and no longer big enough to matter strategically vs. the preferred position we gain as a high-growth publisher through our consolidated approach. Why?

Fewer connections means you can be more strategic with enabled partners and programs. You’ll also matter more to them since your revenue won’t be sliced so thin. Success can be driven by a deeper familiarity between publisher and partner focused on audience and inventory that leads to more inclusion in curated media and SSP-driven PMPs. It might be an improved revenue share negotiated from your increased ad spend, preferred DSP pathing through a specific exchange, “First Look” for a portion of your inventory or tighter controls around 1st party data you’ve elected to put in the open exchange. Beyond impact to yield, fewer partners also means a lighter carbon footprint, an easier path for ad quality controls and the opportunity for faster payment terms. Regardless of the goal, success in this automated marketplace remains relationship-driven and maximizing that benefit outweighs maximizing enabled partners.

Additionally, working with fewer partners doesn’t mean the door is closed for testing. While I hope to continue consolidating our pathing, if a partner can drive value to our existing stack, we’re eager to test. The only requirement is a meaningful revenue share of voice or offering differentiated demand - when it works it’s usually both. It’s not about a large or small number, it’s about impact, relationship, and strategic value add.

If that sounds appealing or at least worth considering, here are some things I’ve used to evaluate and consolidate my partner stack.

Clean up your ads.txt

Run your domain through Sellers Guide. It’s a free and easy way to measure the status of your ads.txt file, eliminate non-existent or incorrect paths and better understand where you sit vs. the industry. The tool also gives your ads.txt a score you can benchmark against your cohort.

Understanding partner performance

Start by selecting a time period you think provides a good data set for the partner stack you currently run - usually the longer the duration the better.

Order your partner stack by total revenue share of voice and do the same thing across your different ad format product lines. For example, you want to understand your stack in aggregate but also between products like display and video which can have different results. Timing matters as some partners are particularly seasonal or only focus on a specific format like video.

These numbers are going to vary between publishers, but broadly when I think of a tier 1 SSP partner it’s driving more than 5% revenue share of voice.

3-5% is still quite valuable and I’m doing everything I can to better understand how I can help those partners increase their share. It typically means they are strong in one but not all ad formats.

The partners below 3% can be removed. Will your performance look better the next day? Maybe not, but you’re not going to lose anywhere close to the perceived value of the combined revenue share of the partners you’re cutting. Further, you’re significantly reducing bid requests and putting the partners that are already driving better business outcomes in an improved position to succeed.

If 3% is too high for a first pass, start with 2% but why not test? The beauty of this approach is it’s a two-way door.

Reducing supply path duplication

If you’re opposed to the concept of removing partners you can also increase efficiency by reducing duplicate paths to your inventory via TAM or Open Bidding.

We removed Open Bidding from our stack after recognizing that the cumulative revenue from enabled partners was near last overall combined with the second lowest CPMs. While that signals an ineffective supply path, I was pleased to see our partners preferred spending via Prebid.

We turned off Open Bidding entirely without impact to our bottom line while reducing our total requests.

Working with resellers

While there’s no question resellers can drive positive business outcomes, publishers need to be increasingly intentional about their approach. Resellers should be significant contributors in your stack, not marginal contributors. Using the methodology I defined above, enabled resellers should be driving 5%+ revenue share of voice monthly to justify inclusion.

You can define this by agreeing on a monthly/quarterly spend minimum or average spend rank amongst enabled partners.

Remember the massive uplift they promised during their pitch? Hold them to it. They will probably insist on some level of testing before making a commitment but it is critical for you to define what success looks like for you once at scale.

For already-enabled resellers, ask them to break down the efficiency of their endpoints in your ads.txt. Typically the top 5 entries drive over 80% of revenue and the rest can be eliminated.

Conclusion

We’ve all heard the quip that the only constant in our industry is change. The reason we keep politely nodding and smiling when someone makes that joke is that its truth is striking.

I believe we can add the advancement of technology to the list of industry constants and as a result, partner and supply path consolidation is a choice today but will be a requirement tomorrow. Bid duplication may yield but that yield doesn’t drive improved outcomes for marketers and advertisers and technology advancement will eventually eliminate it. With a few thoughtful moves to manage your inventory, you can improve your position and get ahead of these changes.

Emry DowningHall heads up Programmatic Revenue & Strategy at Unwind Media, Emry has 15+ years experience leading programmatic strategy across startups and public companies. Emry is an active member of the Beeler Tech community and in 2018 received the AdMonsters Digital Media Leadership Award.