Muki′s AdTech Diary - March

Muki′s AdTech Diary - March

Thinking outside the Sandbox

Everybody is writing about Google’s privacy sandbox, how complex and expensive the implementation is and lowering expectations. I appreciate Google’s ambitious effort to build replacements for all the third party cookie use cases, while preserving privacy. But this solves the issue only for 30-70% of my traffic, depending on the website. I want something that scales across my entire traffic.

Personalized advertising & attribution

We have been so accustomed to measure everything and make data driven decisions that there is little room for the thought that we can’t measure something precisely. The narrative I hear the most is that personalized advertising is more effective. “The new dot com bubble is here: it’s called online advertising” begs to differ. I highly recommend reading this article, even if it’s 5 years old. One key takeaway is this

Economists refer to this as a "selection effect." It is crucial for advertisers to distinguish such a selection effect (people see your ad, but were already going to click, buy, register, or download) from the advertising effect (people see your ad, and that’s why they start clicking, buying, registering, downloading).

There’s a high probability that you target users that would have bought your product anyways, because it improves your KPI. What an advertiser really wants is to attract people that would not have bought the product.?

I would argue that personalized targeting is overrated and will become so expensive to implement and so rarely allowed to be used, that sooner than later it will be gone. Google gives privacy sandbox grants to SSPs, because the investment is so high. Privacy regulations are created all around the world - it started with the GDPR, CCPA and COPPA, and there are a lot more. So either the technical possibilities or legal basis are missing to deliver personalized ads. I hope this will benefit the entire ecosystem as this is a major selling point of the walled gardens, which is a thing of the past.

And attribution? I think this is the bigger issue as there are enough targeting options available, but how to measure the effectiveness of a campaign? I like the idea of coupons to track conversions down to an ad without requiring any user tracking through the entire internet. I hope Apple will not extend the removal of tracking parameters as long as they are for highly aggregated reports.

What about identifiers?

In order to generate deterministic identifiers you need something stable. Forcing users to leave their email address works for larger sites with unique content better than it does for smaller and unknown publishers. Apple already provides alternatives to avoid this via random generated apple email addresses. And don’t forget you require consent to do this. While I think it’s a piece of the puzzle, it is similar to the privacy sandbox. It may work for some publishers. It may work on a portion of the traffic.

Context is key

Advertising has been historically contextual. You book an ad in a certain newspaper or magazine to reach the audience that you think is consuming this kind of content. Attribution has been done by coupons or not at all.?

My niece waits every morning for her school bus. On the other side of the street is a billboard that shows candy, chocolate or toys. It always takes a few days until she asks her parents if she can have the thing she stares at 5 days a week for a couple of minutes.

The most important thing is to understand your audience and where you can reach it. This gives publishers the opportunity to gain more control over the advertising experience on their page. The hardest part will be to earn trust and build scalable solutions to provide the necessary first party signals.

Cut the noise - focus on the core value chain

All this change will bring more focus on the essential value chain in the online advertising system. Publishers produce content that people want to read. Advertisers can buy space in or around the content from the publisher, to sell consumers their goods. This is the ideal, clean world without any fraud detection, SSP reselling or dressing up your bid requests, brand safety vendors or CMPs. I know that this is a utopia, but we should strive for it. Publishers have accumulated too many connections in fear of missing revenue. Advertisers buying cheap inventory on MFAs.

We should think outside the sandbox and focus on what really drives our business. Users that enjoy our content and advertisers that can reach their audience at the right place.

Alexander Savelyev

VP Product at Verve

11 个月

Thank you for putting time to write and share it! I am of slightly different opinion when it comes to this section: "There’s a high probability that you target users that would have bought your product anyways, because it improves your KPI. What an advertiser really wants is to attract people that would not have bought the product." There is marketing to the new users and there is a re-marketing to the existing or historical customers. Both are very needed strategies to grow businesses, but they are different use-cases. Re-marketing is extremely important when it comes to subscription businesses. Think of streaming services. You always need to remind your current customers which new shows are you releasing, else they would switch to one of a dozen of competitors tomorrow. Advertising was always developing to be more data driven, because this is how one can achieve efficiency in pricing. The Real Time Bidding and auction was introduced to allow buyers stop buying ads in "bulk" for everyone. I agree that Sandbox is a good development to solve for privacy, but in my mind it shouldn't be presented to the rest of the economy as if their current approach that they are dependent on to sustain their businesses is wrong.

要查看或添加评论,请登录

highfivve的更多文章

社区洞察

其他会员也浏览了