Netflix has boasted of its improved ad targeting and measurement capabilities as it competes for legacy TV network ad dollars.
But behind the scenes, ad buyers and buy-side ad tech execs exhort Netflix for more data in order to justify their spend, according to two sources with direct knowledge of the matter.
Both sources, who exchanged candor for anonymity, told AdExchanger that they receive insufficient IP address data from Netflix for geotargeting campaigns or cross-platform measurement.
One source, a CTV media buyer, said Netflix provides abridged IP addresses that aren’t as useful as the more complete IP address data they get from other platforms. Another source, an exec at an ad tech platform that works with Netflix, said the platform does not provide any IP address data.
Both sources also noted that Netflix lags other video ad publishers in terms of how much data it offers for programmatic campaign targeting.
Buyers have demanded more sophisticated targeting and measurement levers since the streaming behemoth birthed its ad-supported platform in 2022. But the data they receive remains basic and brands are practically begging Netflix for more, said the ad tech exec.
What’s the IP?
To Netflix’s credit, the company has inked deals in recent years with more data partners for targeting and attribution. It also made its inventory available in a handful of demand-side platforms.
But without IP address data, both sources say the platform’s geotargeting loses its edge.
According to the CTV media buyer, Netflix passes incomplete IP addresses to its sell-side platform, Magnite, which uses that incomplete data to create signals for geotargeting. When Magnite passes those geotargeting signals to buyers in the bid stream, they appear as ZIP codes that represent where Netflix subscribers live, the same source added.
But both sources who spoke to AdExchanger confirmed that Netflix does not pass IP addresses to DSPs or buyers accessing the streamer’s inventory programmatically. Magnite declined to comment.
When asked what data Netflix makes available for targeting, Netflix told AdExchanger that it does not need to provide IP addresses because it instead offers what it calls deterministic identity signals of its member base.
And to be fair to Netflix, there are industry-wide concerns about the veracity of IP addresses as a targeting mechanism in the first place. IP addresses rotate often enough to be considered a semi-persistent identifier.
Even so, IP addresses are considered deterministic enough to be a user privacy concern for publishers and platforms. But many publishers have come up with ways to share IP address data without identifying an individual, such as decoupling the data from other possibly identifiable information.
Both sources told AdExchanger that IP addresses are still an important ingredient of the streaming ad targeting recipe. That’s especially true for geo-based campaigns that blanket a concentrated region, such a designated market area, and that aren’t necessarily trying to target individual viewers.
Generally speaking, IP addresses hover between 50% and 75% accuracy at the city level, according to the CTV buyer. But when an IP address is abridged, as they are seeing with Netflix data, the accuracy rate drops well below 50%, they said. Which means geotargeted campaigns on Netflix have a higher risk of running in the wrong city or state, for one thing.
Fighting fragmentation
Both sources who spoke with AdExchanger about Netflix campaigns acknowledge that advertisers aren’t worried about making egregious mistakes and errantly wasting dollars on Netflix, so much as they have fear of missing out on promising opportunities.
In its campaign reporting, for example, Netflix offers basic reach and frequency counts, completed views and unique reach, according to the ad tech exec. By contrast, other streaming and CTV platforms often report content type and time stamps in addition to IP-level data that allows buyers to attribute campaigns across platforms.
The CTV buyer said that it’s highly unusual for a programmatic buyer not to find IP-level information in a DSP bid. Except when it comes to Netflix.
Without that information, it’s difficult for media buyers to control reach and frequency across the seemingly endless streaming channel options available to viewers. This advertising obstacle, in turn, is largely why wearisome ad repetition remains a major consumer pet peeve and a burden on the entire streaming media industry.
But Netflix only knows its subscriber base in the context of how they interact with Netflix, said a third source, an exec from another ad tech company that works with Netflix.
And if Netflix does have information that would help buyers make sense of how those subscribers behave across platforms, the company is not making that data available, they said.
Netflix told AdExchanger it is taking steps to improve its attribution capabilities by investing in first-party measurement specifically for the Netflix environment, including brand and conversion lift metrics. The streamer also launched a conversion API in March.
Plus, Netflix has recently hired ad sales and measurement experts. Earlier this month, the company added Ted Vlachos to its North American ad sales team and Lisa Kim to its ad measurement team in the same region. Both hail from Amazon Ads, from which Netflix also poached Maggie Zhang as director of measurement for its North American biz earlier this year.
Coming clean
Frustrations about Netflix aside, programmatic buyers have gotten used to operating with incomplete data. Walled gardens aren’t exactly known for their transparency.
One way that black box-style platforms try to meet demands for measurement accountability is by using data clean rooms.
Netflix, for one, works with clean room providers InfoSum (owned by WPP), Snowflake, LiveRamp (acquired by Publicis) and AWS. For what it’s worth, Netflix was the only major streaming publisher to mention data clean rooms during its upfront presentation last month.
Still, buyers need some degree of IP-level data for their own verification and cross-platform measurement, according to both sources. Which is one reason why other major streaming providers make IP address data available programmatically, the CTV buyer said, citing Paramount, Disney and NBCUniversal as examples.
In the current streaming advertising landscape, performance marketers looking for local targeting and/or improved attribution are shifting spend to streaming platforms that deliver superior data quality, said the ad tech exec. Many of these local advertisers are the smaller or medium-sized advertisers that must justify greater budgets for TV ad campaigns. Both sources said they’ve seen some marketers reallocate programmatic spend elsewhere in search of data accountability.
However, for bigger marketers looking for mass reach and other upper-funnel objectives, the ad tech exec noted, advertising on Netflix might very well suffice.
In April this year, Netflix reported that its advertising base grew 70% year over year to more than 4,000 global brands, with 50% of non-live ad sales transacted programmatically. Netflix can earn a tidy sum from its current advertising base, which includes many larger Fortune 500 brands. But if Netflix wants to attract the growing batch of digital marketers that expect more transparency and performance from their streaming investments, it may need to hand over more data in exchange.
