Why is activating and debugging a Deal ID still a nightmare in 2026?

The other day, I was chatting with an agency's Head of Innovation, and we shared the same thought: โ€œWe've been doing programmatic for over 10 years, and debugging deals is STILL a nightmare in 2026!โ€

According to Alliance Digitale's Open Web Programmatic Barometer, only 20% of Display investment goes through Deals ๐Ÿ˜ฎ (Figure cited in our original article; the publicly accessible barometer summary does not give the precise table or edition needed to independently check this percentage. It should not be interpreted as the share of all programmatic formats.)

Why?

  • Activating a deal is more laborious than activating a line in the Open Auction.
  • And, frankly, avoiding the ordeal of debugging deals is a motivation in itself...

And nothing has changed in 10 years... ๐ŸŒ

A Deal ID identifies a private or curated programmatic supply arrangement that a buyer can activate in their DSP. Compared with Open Auction buying, it introduces extra configuration and synchronisation points between the seller, SSP (or Curator) and buyer. (Not every deal type or platform has exactly the same workflow.)

What can a sales house see on the SSP side of a Deal ID?

On the SSP side, the sales house can generally see:

  • The volume of bid requests (BR) sent by the deal.
  • The volume of bid responses (to calculate the participation rate).
  • The volume of impressions won (to calculate the win rate).
Sell-side metricWhat it tells youWhat it cannot tell you on its own
Bid requestsOpportunities the SSP sends towards demand for this dealWhether the buyer's DSP actually considered them eligible
Bid responsesHow often a bid is returnedWhich DSP-side rule prevented bidding on the rest
Impressions wonHow often returned bids result in winsWhich auction or eligibility reason explains a low win rate

Indicative formulae: participation rate โ‰ˆ bid responses รท bid requests; win rate โ‰ˆ impressions won รท bid responses. Denominators and counting conventions may differ between platforms, so compare like with like.

If the participation rate or win rate is too low, the sales house often has no way of knowing what's blocking delivery on the DSP side from its own screens alone.

๐Ÿ˜€ They have the motivation to solve the problem... but not the information ๐Ÿ˜ฎโ€๐Ÿ’จ

What can the buyer see inside the DSP?

On the DSP side, the buyer may have visibility into:

  • The total volume of bid requests received through the deal.
  • Detailed reasons why the DSP filters or rejects bid requests โ€” for example targeting, budget, frequency, creative eligibility or bid conditions, where the relevant diagnostics are exposed.

In short, the buyer can see far more of the reasons behind the non-delivery. (Our original observation was โ€œthe buyer can see everythingโ€; in practice, not all DSPs expose every rejection cause or event. Google's DV360 Troubleshooter, for example, documents many filters but also notes some losses have no reported reason.)

The problem is that they have 25+ lines to manage per campaign. They don't necessarily have time to investigate every single case...

And why would they, when their Open Auction lines are already delivering without any trouble?

๐Ÿ˜€ They have the information... but limited bandwidth/motivation to debug the deal ๐Ÿ˜ฎโ€๐Ÿ’จ

What are the main reasons a programmatic Deal ID does not deliver?

Here are a few practical starting points when diagnosing the gap between SSP requests, DSP bids and final wins. They are possible causes to investigate, not automatic diagnoses (based on DV360's troubleshooting documentation, Microsoft Monetize's guide and our separate French-language hands-on tutorial).

What you observeWhat to check first
Deal ID not visible in the DSPBuyer seat, deal activation, deal dates and SSPโ€“DSP synchronisation
Few bid requests on the SSP sideSupply eligibility, deal rules, inventory volume and sell-side configuration
Many SSP requests, few DSP bidsDSP targeting, creative approval, advertiser exclusions, budgets, frequency and line-item state
Bids returned, few or no winsFloor prices, bid amount, auction losses, creative restrictions and intermediary fees

Illustrative example: if a deal sends 1,000,000 bid requests and receives 50,000 bid responses, its indicative participation rate is 5%. If 10,000 of those bids win, the indicative win rate is 20%. The metrics show where the funnel narrows; they do not tell the SSP why 950,000 opportunities received no bid. (Fictional numbers, not a campaign benchmark.)

How could sharing DSP troubleshooting data with SSPs fix the problem?

The solution?

๐Ÿ’ก What if we automatically gave the most motivated person all the information needed to get the deal working?

๐Ÿ‘‰ DSPs would simply need to give publishers' SSPs direct access to their Deal Troubleshooting module.

The technical complexity of such a project doesn't seem outrageous (potentially automated via an API, while respecting data security). This is our proposal, not a claim that every DSP already provides such access or that implementation would be trivial across platforms.

In practice, a useful shared diagnostic feed could include:

  1. Deal-level rejection categories โ€” targeting, creative, budget, floor and technical errors.
  2. Counts or proportions by reason โ€” enough to prioritise what to fix first.
  3. Clear ownership โ€” what needs action from the publisher, SSP, Curator, buyer or DSP.
  4. Controlled access โ€” authenticated, authorised sharing, preferably aggregated, without disclosing sensitive advertiser settings or individual-level data.

Imagine the difference: instead of asking the trader for yet another screenshot of a waterfall, the seller would see the main rejection category and suggest a concrete fix. (Implementation, contractual permissions and confidentiality would still need to be designed.)

That's not magic AI. It's simply putting actionable information in the hands of the people who can act on it. ๐Ÿ”ง

Would better Deal ID troubleshooting help publishers, DSP clients and the Open Web?

It could benefit the entire ecosystem:

  • ๐Ÿ“ˆ It would allow the sales house to do its job better.
  • ๐Ÿ“ˆ It would save DSP clients considerable time and improve efficiency.

And perhaps it could encourage more buyers to test high-quality private and curated deals, rather than defaulting to the Open Auction because it is easier to get impressions flowing. This is a possible outcome, not a measured forecast.

You could even wonder whether this might become a deciding factor for Traders when choosing their DSP?

We have long argued for simpler activation on the French Open Web. You can see how programmatic curation and Curator seats work, and why publishers are demanding more transparency from the curation ecosystem. The information gap in Deal ID troubleshooting is another side of that same problem.

Sources and further reading