Google has filed a patent application (it's another matter whether they'll be granted it) to listen to your phone conversations. But it doesn't care what you're talking about or who you're talking to. It looks for the background noise, to extract information from the environment you are in and thus offer you related advertisingPure science fiction or is it a horror novel?
The name of the patent already gives the first clues: “Advertising based on environmental conditions”. And he explains something else: “The information about an environmental condition of a remote device is received, this environmental condition is determined based on a signal from a sensor of the remote device or a sensor attached to it. An advertisement is identified based on that environmental condition and the advertisement is sent to the remote device ”.
The idea is similar to that of related ads in our searches or Gmail emails. Google robots read the keywords that we search for or the content of the emails and display advertising related to them. Only in this case the process is a little more complejo.
The patent explains that the system will be able to analyze the sounds surrounding the conversation to extract information about where we are. For example, if you hear bad weather (rain or wind) it can return advertising related to the meteorology. Or if we are in an enclosed space, such as a shopping mall, send product advertisements. If this is combined with the GPS location, it can be fine-tuned much further.
However the patent also includes analysis of photographs and videos. When taking an image or recording something that we upload to the network, the system will analyze its content, offering advertising alternatives.
It is only a patent application and there is still a long process for its implementation if it is approved, but the idea can provoke division of opinionsSome will say that this way they will receive ads that they really need at that moment and others will assure that it is just another intrusion into their privacy.
Via The Next Web.
How this environment-driven advertising would work technically
The approach described combines device sensors (microphone, camera, barometer, ambient light, proximity) and paired sources such as w or smart speakers. The signal would be processed to identify environmental conditions (rain noise, wind noise, traffic noise, in-store music, indoor/outdoor echo) and convert them into advertising signals.
To mitigate risks, it is key to on-device processing where possible, reducing the sending of raw data and limiting the persistence of information to ephemeral attributes not linked to identity. The combination with approximate coordinates and local catalogs would allow hyper-relevant ads without the need to identify people.
The use of images and video supports scene recognition (beach, snow, shopping mall) and non-sensitive object detection (umbrella, bicycle, backpacks) to adjust dynamic creatives. With local inventory catalogs, the platform could vary offers depending on stock and climate.
From contextual relevance to advertising sustainability
Hypersegmentation by environment opens up improvements in performance (CTR and conversion), but it also forces us to consider the environmental impact of the infrastructure that makes it possible. Google has introduced reports of Carbon Footprint for Google Ads, a tool that allows you to estimate the emissions associated with the use of their advertising platforms.
The objective is to provide the market with transparency and a methodology consistent with standards such as Greenhouse Gas Protocol (GHGP) and the Global Media Sustainability Framework from Ad Net Zero, so that brands and agencies incorporate criteria ESG in your media buying.
Google Ads Carbon Footprint: Key Concepts and Methodology
• Google advertising products operate on a shared computing platform, composed of multiple internal services (software) that runs in data centers and content distribution networks. Breaking down their consumption by customer is a technical challenge for the sharing of resources.
• The calculation is based on the energy hour by hour: the expense is separated dynamic (actual load) of the inactive (idle machine), assigning CPU/RAM/storage to each service. In the CDN, for less complete data, a conservative estimate by type of equipment.
• To location-based electricity emissions, the hourly consumption is multiplied by the carbon intensity from the local network. The factors come from Electricity Maps and if they are not available, factors are applied national media of energy agencies.
• In focus market-based, emissions consider the clean energy purchases Google in each region. Annual factors and a scale factor which adjusts the results by region to create monthly reports by client.
• The tool breaks down emissions of scope 1, 2 and 3: direct, electricity, and other indirect aspects of the value chain. The data can be integrated into ESG reports and guide investment decisions in media with a smaller footprint.
Allocation to advertising services, clients and non-electrical sources
First, identify what internal services support Google Ads, Display & Video 360, Search Ads 360 and Campaign Manager 360. Daily emissions per center are consolidate and then assigned to accounts specific using activity data (records), so that only what is processed for that client and day is computed.
In addition to electricity, sources are incorporated non-electric: implicit emissions of equipment (amortized over useful life), facilities (adjusted to capacity and amortized over the long term), consumption of fuels in centers and travel and displacement of the associated staff. These items are updated on a monthly or annual basis.
Important: emissions from final user (access networks, public devices) are not included, although the data in the report may help to estimate those downstream impacts.
Availability, adoption and practical utility
Carbon Footprint reports for Google Ads are offered to large advertisers upon request and will be extended. Brands such as L'Oréal, Carwow, giffgaff and LVMH have tested the reports, improving accuracy compared to estimates based only on spending. Organizations such as BFI Europe y Name Net Zero support sectoral standardization.
For advanced equipment, it is possible to complement with Google Cloud Carbon Footprint if cloud services are also used. Think with Google can be consulted. use cases and good practices to reduce emissions without sacrificing performance.
Privacy, ethics and transparency in environmental metrics
The ambition to measure better brings debates about methodology: hourly vs. annual factors, differences between location and market, or the treatment of clean energy purchases. In AI, too, controversies arise over excessively high numbers. optimistic when indirect consumption is excluded or measures that hide peaks are used. The industry is moving towards independent audits and comparable metrics, key to avoiding greenwashing and strengthen confidence.
If advertising learns from the environment to be more useful, it must also learn to measure its simulator: Contextual relevance and sustainability must go hand in hand, with controls privacy, robust methodologies and informed decisions that improve the experience without unnecessarily increasing the environmental footprint.
