Interview: ROI and Best Practices for Digital Marketing

September 26, 2012

This month I had the opportunity to speak with a reporter from Exame, a leading Brazilian magazine and online hub covering financial markets, technology and marketing for Brazil’s top business executives. Exame journalist Mauricio Oliveira and I had a conversation around ROI and best practices for CMOs and marketing executives.

The interview was originally published in Exame in Portuguese–check out the English version below.

Mauricio: What are the main steps for a CMO or other marketing executive who wants to measure marketing ROI?

DR: The first step is to actually understand what ROI is. When we asked US marketing executives to define marketing ROI, 37% of them made no mention of financial return on their marketing spending. If your team does not know what ROI is, they won’t be able to measure it.

The second step is to define clear objectives for all of your marketing campaigns or activities – are you trying to drive adoption of a new product? Increase brand awareness in a new market segment? Generate leads in your sales funnel? Marketing can add financially to the firm a number of ways: acquiring new customers, retaining existing customers, growing your profit per customer, or reducing your operating costs. But how each marketing campaign delivers ROI will depend on its particular objectives.

Once you know your objectives, you need to try to find metrics that will gauge your success. Not everything marketers do can be measured at the same level of precision. But you need to be sure you are using some kind of metrics on most or nearly all of your marketing.

With those metrics in place, you need to try estimate the ROI for each activity: how much money did you spend on the campaign, and how much do you believe it returned to the firm, whether in customer acquisition, retention, new sales, etc.?

Finally, going forward, you need to apply your ROI analysis to your decision making. We found that only 47% of marketers apply any kind of ROI analysis when deciding where to spend next year’s marketing budget. But those who do are much more likely to be satisfied with the results.

Mauricio: Maybe this is a very expensive process and some people prefer to avoid it for this reason. Did this reason appear in the study findings? Is it expensive to have a strong marketing ROI?

DR: The money invested in measuring ROI will vary. Some marketing tactics are much easier to measure in ROI terms, such as direct response mailing or online ads where the customer clicks through to purchase online. Other tactics such as television are harder to link directly to sales; getting a clear ROI measure may require complicated statistical modeling that is expensive. A smaller organization using brand advertising may need to rely on indirect measures of its impact. These measures won’t tell them the ROI, but can give some measure of impact on factors like brand recognition, trust, and likeability in the target audience.

In our study, satisfaction with ROI measurement did not correlate very strongly with the size and revenues of the firm. What was most important, actually, was how long the firm had been trying to measure ROI. The longer, the better. This seems to indicate that marketing ROI is a long-term process, and that by starting now, you can begin a learning process that will pay off in years to come. So, smaller firms with small budgets should not wait. Start measuring your ROI today!

Mauricio: Do social networks make this task harder? How do you convert “likes”, “shares” and “comments” in US$?

Yes, they do make it harder. The rise of digital media and marketing (from social networks, to mobile apps, to online video, to search advertising) has created a tremendous explosion of new data for marketers. Digital media are inherently measurable. So many marketers assumed that would make it easy to measure ROI. They were wrong, of course, because most of these new digital metrics are not financial metrics. You can’t deposit 1 million Facebook “likes” into your bank account, or use them to pay your bills. And there is the very big problem that most of the digital metrics cannot be compared across platforms. If you are getting 1 thousand “shares” on Facebook and 1 thousand retweets on Twitter, it is not at all obvious if they are equally valuable.

The process I have developed with companies for measuring digital marketing ROI is based on collecting four different kinds of metrics. 1) Audience metrics (e.g. frequency, reach, followers, visitors) measure who you reached. 2) Channel engagement metrics (e.g. likes, shares, comments) measure how engaged they were by your marketing. 3) Universal engagement metrics (e.g. customer satisfaction, lead generation, brand sentiment) can be compared across channels to see the effect on your brand and customer relationships. 4) Financial metrics (e.g. revenue, market share, customer lifetime value) measure the actual impact to your bottom line and ROI.

The key is to develop a model for your own business that links these metrics together, so that you can understand how something like Facebook is creating ROI. But the model will be unique to your business, your customers, and your strategy for using a tool like Facebook.

Mauricio: Could you give me some examples of companies who are doing a good job in marketing ROI?

DR: Ford Motors does a good job of always knowing what their objectives are, and putting metrics in place even when they are marketing in social media, which are inherently hard to measure. For the pre-launch marketing of the Ford Fiesta in the US, they enlisted a community of young driving enthusiasts to test drive the car and share their experiences in social media channels. Ford identified its objectives in advance, to determine which metrics to use. Instead of just measuring social media engagement (shares on YouTube and Twitter), they also measured sales leads generated (number of test drives requested), the increase in awareness of the Fiesta within its target segment, and the cost of their social media campaign compared with traditional pre-launch marketing they had used for other cars.

Another great example is eBay, the online auction site. They run a lot of online “display” advertising, the kind that is hard to measure ROI, because the user may not click through to purchase a product immediately. So, eBay uses web cookies to track the people who see its ads and try to find out if they visit eBay and buy something within the next 30 days or so. But how do they know if the ads are what caused the sale days later?  eBay does something very smart. They always reserve something like 5% of their media buy to run “dummy” ads (e.g. a public service announcement on health). Those ads act as a control group. Now eBay can compare the 95% of viewers who saw its ads, versus the 5% who saw the health message, and see exactly what impact the ads had on customer behavior, and clearly measure the ROI of their marketing.

The most important thing to understand is that there is no “silver bullet”. There is no single way to measure ROI that will work for every company. So every organization will need to develop metrics, and models, that are based on their own business structure and marketing strategy. But the process I’ve outlined will enable them to develop the right model for any business.

New Study: Marketers Struggle with "Big Data" & Digital Tools
[Video] Brite '13: "The Power of (Big) Data in a Networked World"