My Blog

This week: transactional models.

Return on investment has been the mantra of marketing communications for several years. All the major marketing companies are now chanting it, and even some less-sophisticated marketers are starting to hum along. But there's not much harmony, because ROI means very different things to different companies.

Data presented at the Association of National Advertisers' 2007 Marketing Accountability Forum put the confusion into perspective. Only 55% of the top marketers surveyed said that their marketing ROI goals were closely aligned with their overall corporate goals!

That's just the goals, not the metrics.

Things get worse. What metrics were factored into the marketing communications ROI equation?

• 81% said: changes in brand awareness.
• 79%: changes in market share.
• 73%: changes in consumer attitude.
• 59%: changes in purchase intent.
• 36%: return on objective.
• 23%: lifetime customer value.

That's not a typo. Return on objective is in fifth place, and lifetime customer value in sixth! That may be because only 37% of companies surveyed report involving their financial departments in marketing ROI metrics and analysis.

So who is keeping score?

• In 31% of companies, the marketing department does it alone. (Nice for a team to
keep their own tally without pesky referees or scorekeepers horning in.)
• 24% report "informal efforts." (How can data gathering and analysis be "informal?")
• 20% use a team of marketing, IT and finance.
• 17% marketing and finance without IT.
• 8% don't even bother to try!

What's a marketer to do? First, set some meaningful goals and metrics for marketing ROI based on the nature of the business and the available data. One that's right for any particular company depends on the nature of its business.

Immediate results.

A retailer who runs an ad or commercial on Friday can measure the results at the cash register on Saturday. Done right, this method delivers a roughly accurate marketing communications ROI. How do you do it right?

• Establish a benchmark level of sales with no marketing communications.

• Measure increase in sales generated by marketing communications. (That is, the amount over sales which would have occurred without any marketing communications support.)

• Calculate the incremental cost of those increased sales to arrive at incrementally increased profit. The incremental costs include additional product, commissions etc. They don't include the amortized cost of the rent, electricity, insurance etc. There are two reasons to ignored fixed costs.
They're sunk costs. They would have been sustained with or without the incremental increase in sales.

Second, if part of fixed costs are attributed to the incremental sales, that reduces the fixed costs allocated to the sales that would have occurred without advertising. That reduction in fixed costs would be an additional contributor to profit attributable to advertising. That reduction should approximately balance the fixed costs ignored in the original equation.

• Deduct the cost of marketing communications from the incremental profit on sales to get the bottom line profit generated by marketing communications.

• Divide the bottom line profit by the marketing communications cost to get marketing communications ROI.

One important caveat: This model doesn't allow for significant changes in factors other than a marketer's own communications. A great commercial would look bad if a snowstorm closed roads the day after it aired. A weak print ad could seem potent if a major competitor was burned to the ground the day before it ran.

Despite that, this equation can be useful for companies which measure success on a week-to-week basis or, perhaps, quarterly. However it ignores the cumulative effect of marketing communications and any possible future value to acquiring a new customer.

Sales cycle analysis.

Although an immediate "ads equal sales" analysis might work for products or services with a short or instantaneous sales cycle, marketers with longer purchase-decision cycles may need another metric. For example, auto dealersmight be well-served by an "immediate results" approach to ROI metrics and analysis. Their customers decide on a dealership in a relatively short decision window. But auto manufacturers need a much longer-term metric. The automobile brand decision cycle is generally reckoned to be six months, and brand preference and attitudes are frequently formed years before purchase.

Messages that reach prospects at the bottom of the funnel – the brief period immediately before purchase – are increasingly web-based. Until recently that "last click" tended to be credited with the entire purchase decision. Now research has shown that the majority of internet shoppers or purchasers initially decide to visit a particular web site or search a specific product because of marketing communications in traditional media: TV, print, radio, outdoor and direct mail. And furthermore, that last click is often just the final click in a series of visits to multiple sites – visits that may stretch over days, weeks or even months.

There are three principal ways to establish ROI for a product or service with a long purchase decision cycle and multiple channels of communication:

Multivariable testing. If the product or service is marketed in a number of distinct media markets (increasingly difficult as web becomes a more important part of the mix, but, even then, not impossible), if there's a six- or seven-figure research budget and if the data are scrupulously collected and analyzed, this methodology can measure the bottom-line impact of a number of factors simultaneously. Generally the cost and complexity of multivariable testing means that only very large marketers can take advantage of this methodology.

Total program measurement and analysis. John Wanamaker said "I know half the money I spend on advertising is wasted. The trouble is, I don't know which half." (The British attribute the quote to Lord Leverhulme.) Measuring the whole ball of wax won't help determine if any parts of your marketing communications programs are ineffective, but it will give you a reading on the program's overall ROI. The same benchmark-and-tracking process used to determine ROI on immediate-purchase programs can be adapted to work here.

• Establish a benchmark of sales without marketing communications support.
• Measure increase in sales generated by marketing communications over time. Track both time-specific and cumulative increases.

• Calculate the incremental cost of those increased sales to arrive at incrementally increased profit.
•Deduct the cost of marketing communications from the incremental profit on sales to get the bottom line profit generated by marketing communications.

•Divide the bottom line profit by the marketing communications cost to get marketing communications ROI.

• Extrapolate from key factors. This involves a lot of work up front, but may be the simplest way to track long-purchase-decision-cycle ROI.

The Net Promoter Score revolutionized customer satisfaction studies by isolating one key question and one set of metrics to determine how good a job a company is doing at satisfying its customers. (Not surprisingly the Net Promoter Score correlates very closely to a company's growth within its category.)

Just as the Net Promoter Score seems to be the only significant metric for customer satisfaction, there are key factors that can presage marketing ROI. Unfortunately, none has been proven to be effective across all industries, so discovering which factor is the one key metric for a specific company can take a bit of investigation.

Start here:

• Examine existing benchmark-and-tracking study awareness, specific attribute association, preference and purchase/usage intent data. If the data do not exist, begin an ongoing series of benchmark-and-tracking studies immediately. (And hope no one finds out you weren't already collecting this data.)

• Graph awareness and attitude trends and sales trends to identify possible causal relationships. Look for two kinds of patterns: those in which sales coincide closely in time with the awareness and attitude data and those that follow the data by a relatively constant interval, perhaps as much as a year or two later. (When a BrainPosse principal worked on the Coca-Cola business, Coke had identified a key metric that led sales by a year. Tracking that metric allowed us to adjust marketing communications programs to address challenges a year in advance.)

• When a key metric or metrics have been identified, design marketing communications programs to impact those metrics specifically.

•Track pre- and post-program sales and quantify revenue attributable to the changes in the metric.

• Calculate the incremental gross profit attributable to each increment in change in the metric.

• Deduct the marketing cost per incremental increment of change in the metric from the gross-profit-per-increment of change.

• Divide the result by the marketing cost to determine ROI.

The immediate and longer-term models outlined here work for and companies which base ROI on a transactional model. Next week, ROI metrics and analyses for companies which base ROI on equity models.