Type to search

The CMO’s newbie guide to data-driven online advertising


The CMO’s newbie guide to data-driven online advertising


By Jeremy Crooks, managing director Australia and New Zealand for Criteo.


Today’s ‘smart’ data management techniques can make your online display advertising much more efficient than it’s ever been. Fact. This approach allows your ad budget to work better, harder, faster and stronger, with returns on your display ad spend comparable with those of Search. By injecting customer intelligence into your ad buying strategy and carefully crafting the message, you’re much more likely to reach the right user at the right time and for the right cost.

Regardless of whether you embrace data-driven advertising or not, know that your competitors are already doing it. However, newbies beware: before you take the big leap, there are five things you should know… just five.


1. Real-time is not for the faint of heart 

Be sure to use the right data quickly and at scale, as missed ad impressions due to delays translate into wasted opportunities for profit, as does buying less than optimal ad impressions. The required computing power is massive: in peak times, our servers process up to 500,000 requests every second, and analyse 230 Terabytes of data per day.

Survival tips:

  • Make a CTO your buddy. Even though you drive the strategy, the techies will do most of the heavy lifting,
  • bet on partners with both a sufficient capacity for and a strong commitment to R&D, and
  • don’t be shy about asking potential partners for the percentage of engineers within their total workforce.


2. You want the right data, not just more data

Big data doesn’t have to be considered synonymous with fat data – in fact, it shouldn’t be huge because it needs to be actionable. Agile. Sprightly, even. The real value isn’t found in the data per se, but in the cross-pollination of datasets formerly in silos. It’s even more important that people in your organisation can actually extract value from your data.

Survival tips:

  • Invest in technologies that will allow the most people in your company to crunch and visualise your data,
  • partner with vendors who provide high-value, meaningful reporting, and
  • clean your existing in-house data aggressively so it can be used widely, both internally and externally. (Oh yes, your data could also be a source of revenue if you can sell some of it externally. Be careful however, not to cannibalise your own revenue if you sell your data.)


3. Surround yourself with the right people (this means data scientists)

As you embrace ‘smart’ display advertising, hire the talent that will allow you to make the most of all this new data. The people you need – the people with the skills required to do this right – are called data scientists. They are quickly becoming as indispensable to CMOs as financial controllers are to CFOs.

Survival tips:

  • Get sophisticated about how you analyse the data. By this, we mean that you should forget about medians and averages. As you move deeper into one-to-one marketing, things like variances and quintiles will become as important – maybe more so – as aggregate numbers,
  • beef up your testing. With so many players in the space, it is crucial that CMOs develop a strong internal infrastructure that allows them to invite multiple companies to come and test their solutions, and
  • run, don’t walk. Find your data scientists. The supply is likely to run dry within the next few years.


4. Embrace your display vendor. Literally. 

Data-driven display advertising relies on algorithms which feed on your data so it’s important to keep in mind that algorithms work better the more time you give them to learn your clients’ and prospects’ behavioural patterns. The more data you share with your display vendor, the more profitable the relationship will be for you. Data-driven advertising means closer and longer-term relationships with display vendors than ever before.

Survival tips:

  • Forget about campaigns. Advertise constantly and let the algorithms work their magic,
  • don’t shy away from sharing as much data with your display vendor as possible,
  • the best ROI goes to those advertisers who stay with the same vendor for long periods of time, and
  • look beyond your comfort zone. Be prepared to make a change with your current ad-tech partners, even if making the switch seems difficult.


5. Make friends with the legal department (those people know the best lawyer jokes)

With big data comes big responsibility – data driven advertising puts your company in contact with sensitive consumer information that could be subject to regulations, and those regulations change faster than you can say ‘Of course we’re compliant!’ Take this one seriously. You’ll be glad you did.

Survival tips:

  • Get proof from your display targeting partners that they are compliant with any privacy laws and industry practices.


Some parting thoughts…

I’d like to stress that the injection of intelligence and real-time has changed advertising forever. It’s now more relevant to the user, more efficient and ultimately offers far better returns on advertisers’ advertising expenditures. By following these five simple rules outlined above, you now have what it takes to make data-driven advertising work for your brand. So what have we learned?

Wimpy? Don’t be.

Data? Quality, not quantity.

Data scientists? Get some.

Vendor? Hug.

Lawyers? Yup.

Ready? Go and reap the rewards of data-driven advertising!


Jeremy Crook is managing director Australia and New Zealand for Criteo.


You Might also Like

Leave a Comment