Starting Out With Data-Driven Marketing If everyone could afford to invest in an ad tech stack to support a comprehensive data-driven marketing strategy, we'd be laughing all the way to our analytics dashboards

Advertisers spending less than $1 million per year in media often have a hard time justifying the cost of implementing and maintaining a data management platform (DMP). Why? The Year 1 cost alone can be six figures as well as ongoing monthly licensing agreement, making the investment uneconomic, even to the most optimistic CMO.

`Combine that with the difficult task of finding and retaining data marketing talent, and it is no surprise that many marketers and medium-sized business owners baulk at the idea. That is, of course, not to say that the benefits will likely outweigh the costs. Even if the payback is there over the first couple of years, it is often hard to justify the initial cost.

To invest in DMP or not?

We know there is upside in leveraging our data to buy media, we already have a CRM, but outside of email, our data is not connected to any other media platforms. Most businesses keep a good database of customer details, products purchased personally identifiable info, which they segment and address via email. We know most about these customers because they have already been acquired, we are first party to this data.

Then there are those who visit our stores, online or in person, we know something about these prospects as they have left a digital footprint. We capture their movements via analytics, we may cookie their device, we may even re-market to them based on their browsing behaviour and intent profile using a third-party platform.

If display remarketing is the first step in this journey, we begin to really get somewhere when we combine our first party customer data with our largely anonymous browsing data, and anything available to buy from third-party data distributors.

DMP alternatives: What and how

When marketers get it right they may see material improvements in their marketing efficiencies. But if you are not ready to take the plunge you can replicate some of the functions of a DMP within your existing tech stack, which may be simply your CRM and Google Analytics.

If you've already segmented your audience, and have an email or mobile phone number, you have the basis for a framework to extend this segmentation across your Web traffic via tagging and pixel rules which fire when a person does something on your site.

For example, in a hotel context, if a person searches for a hotel with a short lead time, midweek, one night stay, it is highly likely they will be a business traveller. Armed with this high intent indicator you can both exclude and include them from various creative execution. You may exclude them from seeing creative with messaging around late checkouts, or weekend event based packages.

On the other hand, when the messaging does suit, you may bid more for their eyeballs, given you know with a high level of probability they are a high value, possibly repeat guest. Of course, this is a simplified view, but you can build complex audience segments, track and address them with a high degree of confidence… some way approaching what you may see via a DMP.

OK, so what next?

So you are now collecting data and building actionable insights based on first party and browsing data. You now need a DSP to find these audiences and buy in real time. DSPs can be accessed via agency trading desks, but some DSPs do work directly with advertisers and offer a managed service.

The largest publishers/networks, Facebook and Google act as DSPs, providing you access to their online inventory. Start with the platforms you know, leverage the data and begin building the case for further investment in data-driven marketing.