Head of Artificial Intelligence, EMEA H2O.ai
Blogger
Share this content

Cookies are leaving online marketing but data ?

17th Sep 2021
Head of Artificial Intelligence, EMEA H2O.ai
Blogger
Share this content

Since the start of the World Wide Web (and doesn’t that seem like an ancient phrase) as the main way for us to ride the Information Superhighway (I promise I’ll stop with the call-backs to the ancient past now), user data has become the main fuel of online marketing.

Whether it be third-party cookies, web logs, email clicks, shipping cart history, if the user is accessing your site on their mobile or from their desktops, brands, aided by a new scientific and numbers-driven breed of Mad Men, grabbed all that. Advertisers also hoovered up useful contextual and customer behaviour, so his demographics, her expressed interests, their interest (or not) in your customer rewards/loyalty programme… seasonal data—the list goes on.

‘Everyone hates cookies, right?’ Actually--not so much

But then GDPR came along. Which we survived. But a new threat is emerging to us using all this lovely online data as the rocket fuel of our cyber-marketing: soon, we’ll all have to live without third-party cookies, as Chrome and Chromium-based browsers will stop supporting them in 2022. Apple has been anti-tracking for some time, and the latter is also of course, for its own commercial reasons undoubtedly, at war with the rest of the web right now with its very tough stance on privacy. Just to add to the fun, Google has changed its algorithm (yet again) to make sites offer better content and stop robots gaming the system with valueless spam.

Yet people still want personalisation. If you ask your friends, what do you think about online ads and the way they follow you around sites, sure, everybody hates ‘em: cookies are the worst thing ever. But when those same chums want to buy new climbing shoes, they do a search and in 10 minutes they get served with the best ones on the market for their budget because the system knows which shoes they like, then everybody's somehow more than okay with that convenience.

The reality is, 81% of consumers want brands to know when and when not to approach them; 63% of consumers expect personalisation as standard. So, we seem to have a big headache coming up for the marketing sector: how to somehow access data still, but also mobilise it effectively for sales and marketing purposes?

The answer is something the sector has always relied on perfectly happily, for all our old-school image of marketing being about creatives enjoying long boozy lunches: mathematics and statistics.  'General Linear Models' have been used in Marketing since 1972, as it’s especially helpful around marketing mix modelling, as they allow for an additive approach (sales = baseline + TV spend + radio spend + other channels). In combination with clear model specifications, stats have always provided marketeers with great ways to identify incremental contributions of marketing channels to sales.

How Machine Learning helps win back they will lose from the end of cookies

Now, mathematics is getting even more useful for our sector, thanks to the introduction of advanced software techniques to make the most out of vast heaps of data and numbers. Yup: step forward Machine Learning (ML)—and my prediction is that ML can, and will, help Marketeers balance out (to some extent) what they will lose from cookies going.

This will provide better experiences for customers and improve performance and conversions. The opportunities emerging out of Machine Learning deep dives into customer behaviour range from hyper-detailed segmentation of customer types, allowing brands to offer truly dynamic pricing, hyper-specific/accurate ad targeting, tailored messaging and performance measurement. 

ML is already being used in the sector to:

  • predict customer needs and future trends
  • derive deeper understanding of the customer journey via attribution modelling
  • better budget allocation to improve ROI
  • and allowing campaigns to dynamically flex to customers’ next actions and experiment with personalised experiences (‘Next Best Actions’ work).

Right now, too much of this excellent work is being done in larger companies with big teams of data scientists: there is a definite gap between the data and ML specialist and the frontline Marketeer right now. Companies like mine are trying to close that gap via what we call ‘Explainable AI’: systems set up to be so user-friendly that there is as little ‘black box’ mystery and lack of explanation as we can manage. This way, teams like yours can access amazing tools that can crack very large amounts of data and work with very tough statistical models with ease.

‘A higher accuracy in out of-sample predictions’

Our client Allergan, a big US pharma company and the maker of Botox, has been working with us to move beyond older models for estimating the impact of promotion campaigns, and instead starting to optimise the marketing mix of literally dozens of products across a variety of marketing channels. A Machine Learning technique called gradient boosting is proving particularly beneficial, as it typically shows a higher accuracy in out-of-sample predictions than the General Linear Models I referenced above. 

What Allergan’s marketing team have found: better results, more conversions, a better marketing budget mix—quicker and cheaper than before. This is what your team can and should be achieving, too. 

So, yes, you’re going to lose cookies. But if you pivot early to a smart use of ML, you can avoid wasted months ahead as you try to work out how to live without cookies and ad tracking. 

Marketeers, the bottom line is data is still your friend; you’re just going to source it differently.

Replies (0)

Please login or register to join the discussion.

There are currently no replies, be the first to post a reply.