Every programmatic impression with a thin bid request pays an invisible tax — a discount applied by DSPs to compensate for uncertainty. Most publishers have never measured this tax. The ones who have are universally shocked by its size.
When a DSP evaluates a bid request, it's making a probabilistic estimate: how much is this impression worth? That estimate has a confidence interval. The wider the confidence interval — driven by missing or insufficient signals — the more the DSP discounts its bid to protect against overpaying.
This isn't nefarious. It's rational behavior. A buyer at auction who can't accurately assess value bids conservatively. But when every buyer in the auction is applying the same conservative discount to the same thin request, the clearing price can be 20–35% below what the impression would clear with full signal context.
The uncertainty tax shows up in your data as bid shading. Most SSPs report bid shading percentages in their reporting — the average gap between the DSP's maximum bid and their actual submitted bid. On thin bid requests, bid shading runs 25–40%. On signal-rich requests, it runs 10–15%.
The uncertainty tax isn't applied uniformly. It's concentrated in specific scenarios:
These taxes don't add — they compound. An impression missing identity signals pays the identity tax. If it also lacks contextual signals, it pays both taxes plus loses access to contextual campaigns entirely. Missing quality signals adds a third layer of discounting. An impression missing all three categories may be clearing at 40–50% of its true market value.
Signal enrichment eliminates the uncertainty tax by giving buyers the information they need to bid confidently at true value. Metrux publishers consistently see bid shading rates drop from 25–35% to 10–15% within the first billing cycle after enrichment activation.
The uncertainty tax isn't just a revenue loss — it's also a signal to the DSP ecosystem about the quality of your supply. Impressions that consistently clear with high bid shading get assigned lower quality scores in DSP algorithms, which can reduce their future auction eligibility. The tax compounds over time. The sooner it's addressed, the more of the cumulative cost is recoverable.
Metrux delivers 20–40% yield improvement through signal enrichment — no dev work, no tag tax, no risk.
Request Early Access →