"I Know Half My Advertising Works, But Which Half?"

If John Wanamaker* had a big data strategy he may have been able to answer his own question; which half of my advertising works? He may then have refined his advertisng strategy and called it 'revenue event planning' in order put return on investment at the heart of every stage of his allocation and optimization process within and across marketing channels.

The desire to influence and measure both short and long-term marketing effect is the goal of ‘revenue event planning’ and requires the identification and valuation of outcomes that, in turn, set the optimization priorities for campaigns and individual placements and even impressions. The opportunity to do that is now real and its execution is supported by the granularity of the digital delivery of communication and the data created by it.

What is a revenue event? Most obviously a revenue event is a sale or some other kind of completed action, like paying for a flight online or a sign up for a CRM program. Businesses can ascribe specific value to these events, not just to the event but to the avoidance of higher transaction costs on other channels and sometimes to lifetime customer value.

Not all revenue events are both easily isolated or valued, and it is not always the most productive route to seek to isolate them to the last click. Consider the case of an automotive company. It is not quite so straightforward to establish perfectly how many visitors to vehicle.com actually purchased that brand, but it is fairly easy to model. If there was a ‘buy now’ check-out page on vehicle.com that would also be of only so much use as currently relatively few people complete such purchases online. In such cases it is better to identify proxies for revenue, such as time spent, test drive bookings and the use of finance calculators and the incidence of abandonment at various steps in the process. The analysis will reveal those actions that combine the likelihood of driving sales volume with highest conversion from action to transaction.

Other analysis can reveal even more textured and valuable insights. For example in consumer electronics there are goods that sit at the heart of a product, peripheral and service ecosystem. Simply, selling an Xbox can sell games and an Xbox live subscription and maybe additional devices and subscriptions creating an aggregate value far exceeding the original purchase.This can help answer the question; do I advertise Xbox, Xbox Live or PGA Tour 14?

Higher up the purchase funnel (or vortex) the process becomes more complex still. To what degree is awareness or engagement a proxy for a sale? What is the appropriate multiplier of a view or a like to a share?

If we are successful in understanding these values and their interdependence, it will be possible to create better and more accountable ways of approaching communication investment strategy and execution in all paid, owned, earned and shared media platforms and optimize overall performance.

The complexity then is not in the concept but in the requirements for execution:

1. The analyses required to isolate revenue events and their proxies

2. The analyses required to identify the one-time and lifetime value of those events

3. Some mechanism for assessing avoided cost

4. Tracking of event occurrence via on-site pathway analysis and tagging

5. The ability to drive queries and discovery to the site location most likely to generate the event

6. The ability to identify value and to set bid prices on relevant inventory in reserve, auction and real-time environments.

All of the above depend on the availability of data and the willingness to use it, to say nothing of the legality of its use.

If the promise of big data is to be realized the practice of ascribing value to events must become a core discipline for advertisers and the agencies and data partners that serve them.

* John Wanamaker was an American retailer; the statement is often also attributed to Lord Leverhulme, the soap magnate.

@robnorman

Photo: isak55/Shutterstock

Figo Luo

T12GROUP -Cofounder

10y

It is great to read. Everything surrounds business is always about creating revenue. And big data gives us a better potential to get better revenue. So before we start to dig something from big data, we must make sure it could make something valuable. And I do believe data is just one kind of reference, marketing, included advertising and so on, is more than data cause customers are all humanities.

Like
Reply
Larry Smith

Madison Ave Branding x Main Street Communications for big ideas with a hometown touch.

10y

If only there were half that was effective. Response rates, lifts and other attribution models for DR and brand work in the single digit percentages.

Like
Reply
Lee Latham, PMP

Senior Lead Project Manager | Commercial Program Management | Digital Product Management

10y

Great Question: "To what degree is awareness or engagement a proxy for a sale?" Better understanding of these values and their interdependence will optimize overall performance.

Like
Reply
José Zulmar Lopes

Owner RGL /Siemens 25 years/CFO/Renewable Energy/Engineering /Sustainability 

10y

Big Data is not inherently bad or good. Our use of the sheer volume and variety of data that makes it good or bad situation. Predictive algorithms can alert us of a beginning of the epidemic and give us time to take preventive measures. Can uniquely identify a person in a group and thus prevent stereotypes are used to judge certain person in an ethnic or religious group. Helps us make better decisions correlating data that they had not imagined relationships between themselves and this can change our opinions, eliminating false beliefs.

Like
Reply
The Sparks Technology Group Combining experience,talent, work and kindness.

Pursuing remote and hybrid opportunities. I am calm, kind and learn concepts very quickly. Please, only USA based recruiters.

10y

Word of mouth is the most effective and still the best.

Like
Reply

To view or add a comment, sign in

Insights from the community

Explore topics