Closing the loop: when reality meets the forecast

Establishing customer lifetime value (CLV) is crucial in building accurate models to measure the success of marketing investments. Grant Halloran, Orbis CEO, explains.

In keeping with the 'new accountability' of marketing, a number of marketers are beginning to treat marketing investments like capital investments. This involves development of quite sophisticated models that enable a marketer to forecast a hard ROI number. Understandably, a lot of background work needs to be done to support validity of these models, which I will touch on. The next big challenge, however, is for marketers to demonstrate the actual returns versus the predicted returns of these individual investments.

At the heart of many of these models is the concept of 'customer lifetime value' (CLV). CLV is a measure of the economic value the business will derive from an individual customer whilst on the books of the business. For example, the acquisition of a customer by a bank, insurer or telco will result in the opening of a new customer account, and this account will generate cash at various points in time. The value of this cash discounted to a present value drives the ROI calculations. Depending upon the industry, the other variables in the model can be quite complex.

In all cases, a 'return', expressed as a dollar figure in today's terms, can be compared to the investment. It's exactly the same investment model used by the finance department for modelling other business investments, so marketers, if they're not already using these methods, just need to get themselves educated.

Behind the modelling are many assumptions. Analytics tools can help marketers develop these assumptions, such that they have reasonable confidence about the model. For instance, the response rate, acquisition rate, and economic driver will be different depending upon the campaign's location, audience and other defining characteristics.

Arriving at a prediction of the returns is one thing. Actually getting them is another. If marketers want to 'close the loop', then they need to be able to track the ongoing cash stream back to that original campaign. It gets more complex once portfolio managers get their teeth into customers, but the underlying principle still holds firm. If our model predicts a Net Present Value (NPV) of $2000 per customer account and 1000 accounts are expected to be acquired, then we need to know (over time) if that two million dollars in gross returns is actually generated. If we do that, then the assumptions underlying the models can be constantly tweaked, resulting in a significant increase in marketing effectiveness. Add fingertip tools for marketers to do the modelling and view intermediate and long-term results and you have a high-performance marketing team with which the competition struggles to keep pace.

Effective CLV modelling is a long-term initiative. It takes several years of data build and campaigning before the models can be tinkered with. If you haven't yet started, beware: your competition is already at it and this capability will be exploited in very powerful ways over the coming years. Strength in 'precision marketing' will continue to grow as a defining characteristic of high-performance companies operating in competitive markets and I would expect any corporate strategist to believe in this.

So what does a marketer need to do to get started? Firstly, getting a person on your team with a finance/accounting background is proving to be useful to many marketers I know. This person can add a lot of value in operational support of the marketing function, and can also drive the analysis needed to establish investment models.

Secondly, you'll need to invest in systems. Designing 'a process', telling everyone to follow it and sticking a process diagram on the wall just isn't good enough. Marketing distinguishes itself from the rest of the business by being a 'projects-based' department - lots of cookie-cutter projects with all the necessary ingredients that productivity improvement experts look for: many people involved, repetitive tasks and processes, standardised documentation, measurable costs and results. Your key marketing management system can enable you to increase your capacity and dramatically reduce process administration costs. Investment modelling, results analysis, cost tracking, project management, sign-offs and forms processing (such as briefs and business cases) can all be done from one window. Basically, if you haven't got a specific budget line for systems to support your department, then get one.

Last but not least, embedding the 'measurement discipline' in your team is vital. As it is for any type of cultural change, success or otherwise is driven right from the top. Making marketing analytics part of your regular communications, business plans and reporting is a good entry point. 

Back to Articles