I bought a new red car a couple of years ago. “I can’t believe how many red cars are on the road now,” I remember thinking. Red cars currently make up about eight per cent of new car purchases which, if anything, is slightly down on what it has been over the last couple of decades. My mind had tricked me with a classic case of confirmation bias. I was driving a red car so I was more inclined to notice other red cars.
Many people have contacted me recently to tell me an anecdote about their phone listening in on their conversations and then seeing related ads. The plural of anecdote is anecdotes. It is not data. I have largely dismissed these stories as an example of confirmation bias. Besides, Google, Facebook and Instagram have all categorically denied that they use audio to decide what ads to serve to you, and multinational tech organisations can be trusted, right?
Well…in a world-first enforcement action brought by the Australian Competition and Consumer Commission (ACCC) against Google LLC and Google Australia Pty Ltd (together, Google), the Federal Court has found that Google misled consumers about personal location data collected through Android mobile devices. The court found that when consumers accessed the ‘Location History’ setting on their Android device and turned the setting off, Google continued to collect, store and use a consumer’s personally identifiable location data.
Maybe my readers were on to something.
Why do we see certain ads pop up in our searches and our feeds? For an advertiser, the greatest aspect of using online ads is the ability for an ad to be targeted. When I am using an electronic device, the apps have an incredible amount of information about me so ads are targeted specifically at me and my interests. They know my age. If I am seventy years of age, I am more likely to see ads for aged care facilities. They know the time of day when I am using my device. I am more likely to see ads for coffee when browsing in the morning compared to late at night. They know my parental status; my interests; the type of device I am using and probably my income level.
And here is the biggie. They know my recent search history.
You can see how important this last one is. I recently needed to buy a new studio microphone and the local supplier was out of stock with a three-month delay. I went online and ordered the unit I required. For the next few weeks I kept seeing ads for the same model microphone that I had just ordered. This is the power of targeting. I was obviously in the market for a new mic so seeing ads for microphones made a lot of sense. The AI wasn’t quite clever enough to know I had already made a purchase.
So back to another microphone – the one on your phone. Try an experiment. Pick a random topic that is obscure and something that has not recently interested you. Talk about this topic near your phone but never type the topic on your PC or phone. Do this for several days and sit back and watch. I typically think that the ads are very cleverly targeted based on all the information that is already known about us so I don’t think you will see any ads related to this obscure topic but, after the ACCC case, I am not quite as convinced as I once was.
Let me know the microphone experiment you conducted (and the results) at ask@techtalk.digital.
Mathew Dickerson