In my decade working in interactive advertising I’ve noticed a disturbing trend in the use of the term “Return on Investment” or ROI. The problem isn’t really an inaccurate use of the term -- the definition has softened over the years -- but more of a disregard of the equation’s potential due to sheer laziness. For instance, in the past week alone, I’ve read two different electronic books that promised an “increase in your ROI” by using their methods. However, once I got past the promising title, I found out that the author doesn’t even bring up the equation for determining this important measure of profitability.
I can’t get too upset about this abuse; math has always been a point of contention for a wide range of people, including myself. In fact, I even switched majors during my undergrad days from computer science to advertising because I couldn’t get through Discreet Mathematics. But since the ROI equation is really just about division, I’m going to take a minute or two to go over the details of what the real ROI is about and how you can use it in the optimization of your SEM and other interactive advertising activities.
Be warned here, kids, we’re going to be talking about mathematical equations today. But don’t panic, it’s nothing close to John Nash level equations, just simple division and multiplication.
The Power is in the Purity
The real ROI equation is one of those rare mathematical formulas that is pure poetry in its simplicity yet still produces powerful results. In fact, according to the “Dictionary of Modern Economics” from the MIT Press, the definition of ROI is "A general concept referring to Earnings from the Investment of Capital, where the earnings are expressed as a proportion of the outlay."
Basically, this means that ROI is a measurement of profitability that takes into account the profit from an activity in reference to the capital invested in the activity itself. Or, in an equation style format:
Isn’t that beautiful? I know, I know, it’s nothing you would want to take to the prom, but you really can’t ask for a more simple way to measure your profitability. The results are easy to interpret as well: Just like when looking at a result for profit, the closer your ROI is to 0 percent, the closer you are simply to breaking even. The farther the number is away from 0 percent, the better the return. Think of ROI in the same way you would a regular bank savings account’s interest rate.
Alternatively, if your ROI is less than 0 percent, then you’re losing money, as you probably already knew from the negative profit number. So, for example, if you partake of an activity of some sort that results in $100 in Profit that only cost you $100 in Invested Capital, then your ROI would be 100 percent, that is, you get an additional $100 back from your investment of $100, thus doubling your money.
Now, since a lack of knowledge with these terms is usually one of the reasons the ROI equation is abused, I would be remiss if I didn’t fully explain the two major parts of the equation in detail. Profit, in its simplest form, is the gain from a business activity after subtracting all of the expenses used in the activity. Invested Capital refers to all capital (that is, cash or goods), used to generate the profit in the aforementioned activity, such as advertising costs, physical plant, and so on.
It is important at this point to bring up the value of accuracy and consistency. In both halves of the equation, to get as accurate an ROI result as possible, you need to make sure you are really taking into account all of the expenses and capital used in the activity. Otherwise, your results will be a lot better than reality, and you stand the chance of possibly continuing an activity that does not warrant continuing while your company slowly bleeds cash. Additionally, you should make every effort to be as consistent as possible when calculating your ROI value; that is, you need to use the same methods to determine profit and invested capital each time. This little precaution will ensure that a comparison from one activity to another is not biased because in one instance you counted one aspect of the activity -- say a set-up fee or other expense -- and the other you did not.
By the way, while we’re discussing things like profit and invested capital, I should bring up that the financial world is a tad wishy-washy when it comes to what to call the variables used in this equation. So, if you see terms like “Net Income,” “Net Profit,” or “Net Earnings,” in this instance it is the same as “Profit.” On the other side of the equation, if you see terms like “Assets” or “Total Assets,” in this instance it is the same as “Invested Capital.” There are instances where these terms do have alternative meanings, but when calculating ROI for something like advertising, just stick to these definitions and you’re golden.
One more point that should be made: the ROI metric is specifically designed for comparison to other ROI measurements and not to be used as a stand alone measurement. Remember the "Click-Through Rate (CTR)" metric? Remember what a bad rap it got for a while? Well, that bad rap came from people misusing it. The problem arose when people started comparing their CTR for a campaign against things like "industry averages." Often, they neglected to take into account variables like sales or ad placement. In the end, these folks would cut perfectly effective campaigns that were making them money because they misunderstood the metric.
Variations on a theme
While the power of the ROI equation may lie in its simplicity, that same simplicity can cause some issues when it comes to collecting all the necessary data to provide the most accurate results. For instance, finding the true profit of an activity can prove difficult in some instances due to a lack of data on other aspects of the business. Omissions in this data can, as previously mentioned, provide inaccurate and costly results. With this in mind, I will present a collection of variations on the ROI equation that may prove useful.
To begin with, the equation itself can be broken into two separate parts that incorporate sales (a.k.a. revenue) into the equation without undermining the quality of the results.
If you remember back to your algebra days, this equation is made possible by the fact that since the sales variable is both above and below, it cancels itself out and therefore simplifies the equation down to the previously discussed form. However, for now, let’s concentrate on the expanded version above.
As you might have noticed, the right half of the equation (profit over sales) is the same equation that provides you with the Net Profit Margin (or just Profit Margin) percentage. This rate can often be found in all public companies’ annual reports and on the balance sheets of most private companies, if the company shares information with the rest of their employees. (It should be noted that the left side of the expanded equation, sales over invested capital, is another metric known as Turnover. This metric is also often found in most annual reports; however, its use would only complicate things here, so we won’t present any versions of the ROI equation with it in place.)
Once you convert the right side of the equation to just profit margin, the above equation can be simplified further to:
In this form, it is much easier to use the company’s overall profit margin in your ROI calculations, rather than just what you can dig up as expenses used in a particular instance, such as just the cost of goods or advertising. Some may argue that this variation places an undue burden on sales revenue by multiplying the profit margin for the entire company against the sales from this one activity. Conversely, not using this number in place of a more lenient profit estimate may produce an ROI result that keeps a money-losing activity alive. When this decision is up to me, I tend to lean towards a more conservative answer that could save the company money. However, if you feel you can stand the risk, feel free to use a broader profit variable in the equation, such as Sales less the Cost of Goods, etc.
Tomorrow: So what does all this math have to do with interactive marketing?
Jeff Ferguson is an internet marketing consultant and president of TheGag.com.
It's one thing to have high awareness, but quite another to have high familiarity. If someone cites your brand unaided as part of the consideration set of brands in the category, do they know what you do or what makes you distinct? Qualcomm is a well-established brand that has fairly high awareness scores, but a lot of people think it's a sports brand because of its prominent exposure at Qualcomm Stadium in San Diego. Qualcomm is one of the largest manufacturers of semiconductors for mobile devices. It's an ingredient brand, and familiarity is probably the most important measure of health it should track.
Familiarity is another measure that doesn't require expensive studies. It can be facilitated online quite nicely, and you can even uncover useful insights from informal qualitative research. Statistical significance is nice to have, but the prescription for poor familiarity is often better informed by the data you generate in deep discussions with targets. What do they think the brand is or does? What makes them think that?
The sine quo non of brand building is the "higher order" of brand drivers. Al Ries and Jack Trout famously claimed [in their book "Positioning: The Battle for your Mind"] that positioning is the battle for the mind of your target. The war is won when you link your brand to emotional and self-expressive associations. People link Disney to magic, Nike to performance, and Apple to creativity. What do people associate with your brand, if anything? Are the associations the ones you want them to have, or are they a byproduct of the past?
Linking studies are critical for repositioning projects, but they are often poorly designed. The brand community suffers from a nasty plague: rational inquiry. Too many studies attempt to measure brand associations by using rational inquiry. Hedonism plays a role in Virgin Atlantic's brand equity. It's a brand that glorifies life in upper class. But if you tried to measure the link between Virgin and hedonism by directly asking respondents how much they value Virgin's ability to let them indulge their sensual side, you might be fooled by the results. A lot of brand equity derives from irrational associations. When you attempt to rationalize it, you tend to get three kinds of erroneous answers. First, you get people who are too embarrassed to say what they really feel, so they answer falsely to conform to a social norm. Second, you get people who answer in a very logical way. The problem is, in the moment of brand use, they aren't thinking logically. Finally, we tend to forget our irrational impulses. Since most studies are backward-looking (the study asks respondents questions about past behavior), they sometimes just remember wrong.
One solution to the problem is to simply design better descriptive studies. The old rule about garbage in/garbage out applies. Many managers skip over discussions about the attributes or questions to include on the survey. Survey design is both an art and a science. The artful part is often neglected.
The other solution is to consider experimental study formats. Most people think experimental research is reserved for academics, but experimental studies can be cost effective and highly predictive. Online brands have relied on experimental research more than offline brands, largely because it easier to conduct. Amazon.com frequently tests new features and new user interface designs in real-time by pushing the new material live for a small random sample of its customers. It then compares the behavior of the sample with the behavior of the general population. It's a safe and effective method for product testing. With a little ingenuity and bravery, brand managers can use similar designs to test changes in brand positioning.
Also, it is important to consider that brands are not static entities. They change over time, just like consumer tastes. Longitudinal studies that track changes in brand associations over time are a useful approach. While the first year of the study might be the most costly, incremental waves of the study are generally easy and economical to execute. Just as your doctor keeps a record of your vital signs, an ongoing linking study provides a simple diagnostic of your wellness.
Preference and loyalty
Do your customers prefer to use your brand even when prices go up? Are they willing to recommend your brand to a friend? Do they give you permission to take your brand into new markets or product categories? Brands with loyal customers can and should ask these questions and consider them in strategy formulation. Brands on the cusp should weigh the potential costs and benefits.
When budgets are tight, loyal customers are like a rainy day savings account. It's less expensive to keep a loyal customer than it is to win a new customer. However, like contributing to a rainy day fund, you really need to study customer loyalty in advance of when you need to draw upon it. To gauge the strength of your customer's loyalty, you really need to compare two sets of data. For many of the reasons stated above, it is hard to gauge loyalty with just one wave of a study. Also, some of the best loyalty measures are not descriptive surveys but sales and customer satisfaction data gathered over time.
If you're planning a baseline loyalty study, there's another reason you should be cautious now, in a wounded economy. When there are macro-environmental trends that cut across the entire market, you may be measuring loyalty during a low tide. Just like your company, your consumers are also pinching pennies. In this context, most people are willing to sacrifice brand loyalty for pocketbook savings. That will often surface in a loyalty study, and even though it may be relative in regard to your competitors, you're gathering bad data.
Preference, which is a closely related cousin of loyalty that measures how much customers prefer your brand to a competitor, may be a more effective area to study. There are many simple and effective quantitative techniques to gauge preference, conjoint being a favorite of many researchers. Keep in mind that you may need qualitative data to supplement what you find in quantitative studies. It's one thing to know that customers prefer your brand to a competitor, even if the price is a few dollars more. It's another thing to understand why that is.
This is not to say that you should ignore loyalty measures and focus solely on preference. For some brands, loyalty is a critical indicator of brand health, and now may be just as good a time as any to take its pulse.
One size does not fit all
When doctors order lab work, there are many panels of tests they can order to assess your health. They rarely order all. Instead, they assess your age, your family history, your previous health issues, and their surface observations about you after an exam. You must do the same in assessing your brand health. There is no silver bullet brand health study that will serve every brand on the planet. You have to assess what you know about your brand and fit the study to the patient.
If your brand is young, growing, and owns limited market share, you should probably focus your efforts on awareness and familiarity. If your brand is an established and dominant player with scores of legacy data, you may want to focus on linking and preference. As we've heard so often in the national healthcare debate, the key to keeping brand health costs down and generating data that will improve your strategies is to focus on the desired outcome, and then work backwards.
Laurence Vincent is a writer and seasoned brand strategist.
On Twitter? Follow iMedia Connection at @iMediaTweet.
You've seen countless examples of this, and maybe you've even been a participant. A brand delivers a bad customer experience, so the customer turns to social media to vent (hopefully going easy on the outrage and profanity). You can find plenty of articles talking about the recent intersection of customer service and social media. But for all the ink and pixels spent on that topic, I still hear a lot of agency staffers looking to delineate the practice of responding to comments in real time as either a PR or marketing function. Frankly, it's both. Although I'm not sure the debate matters much beyond assigning workload and allocating budgets, because the consumers don't make that distinction at all.
"Consumer-facing brands have used social media as an opportunity to turn consumer complaints into marketing opportunities," says Fischman. "Next time your American Airlines flight is delayed, try tweeting a message about it. Within minutes you will get a response from one of the many AA operators that man the AA war room around the clock. AA does an amazing job at turning consumer complaints into RTM experiences."
Writing about a recent travel mishap, Alan W. Silberberg detailed exactly what Fischman is talking about.
But it's not just American Airlines. A lot of commercial carriers work the customer service RTM continuum particularly well (perhaps because with so many flights in the air at any given moment, there's always going to be some issue somewhere).
A while back, I flew Virgin America from Los Angeles to New York. Around midnight, we had to divert to Chicago because the bathrooms had become inoperable. We spent about two hours on the ground in Chicago before we eventually made it to New York. None of the passengers were in a good mood when we landed. In fact, we were a tired and cranky lot, even though the airline had acted quickly to remedy the issue and the pilot had been diligent about keeping us informed.
Now, I know what you're thinking. I tweeted something about the experience, and Virgin America stepped up with an apology. But that didn't happen. The fact of the matter is, a smart brand with a capable social team doesn't always need a social media prompt to get in the game.
Just after landing in New York, all of the passengers turned on their phones and saw an email from Virgin America apologizing for what had happened. The company even tossed in a $50 credit because of the hassle. So what began as a potential customer service problem ultimately ended up as a marketing win because Virgin America leveraged a potential PR problem into an opportunity to build brand loyalty. And yes, I liked Virgin America before that trip, but it's safe to say that I'm now officially a huge fan.
These days, it's just not a media event if a slew of brands haven't geared up for an RTM blitz. And why not? Events like the Super Bowl, The Academy Awards, and even Shark Week are huge media opportunities. But successful RTM campaigns vary wildly.
"The degree of success the brand has is dependent on the resonance they can create between themselves, the event, and the consumer," says Ming Linsley, senior partner and practice lead of social at MEC.
But don't mistake that as a simple equation. Resonance is very subjective, and RTM isn't as simple as hooking your brand to the coattails of a big-deal media event. In fact, the event doesn't even have to be all that big to be a win. According to Linsley, what really counts here is thoughtful planning.
"The most common misunderstanding about real-time marketing is that it is serendipitous," says Linsley. "While there is a certain degree of luck involved, there is a much higher degree of discipline. Brands need to establish a team empowered to make decisions and act in real time; this involves clear guidelines and processes. Brands need to give prior thought to what events or topics are relevant and appropriate for them to engage with in real time, and what communities and influencers are the most important."
Consider how GE handled the anniversary of Thomas Edison's birthday, which just so happens to be a global holiday known as Inventor's Day.
Last Inventor's Day, GE asked its Twitter followers to share their ideas for inventions using the #IWantToInvent hashtag. But that was just the beginning. GE selected the most interesting ideas and turned them over to a design studio it had hired for the event. By the end of the day, GE had produced 70 original blue prints and tweeted them back at the amateur inventors.
But while the campaign came off as spontaneous, nothing about it was accidental.
"For all brands...the imperative is to find ways to act with more agility, which, ironically, would require a lot of practice and planning," Giselle Abramovich wrote in Digiday. "At General Electric, that means coming up with a long-term editorial calendar of events to focus its content strategy around. Then, GE brainstorms a production schedule and gathers the right resources: a strategist, producer, designer, and a lawyer. These four people are in the same room during the event to make it easy to make fast decisions on both content and distribution."
Some people don't like the term RTM. Instead, they'll tell you that RTM is really just news-jacking, which is what happens when a brand exploits a news story for marketing purposes. Obviously, I'm not one of those people. For me, news-jacking is a subset of RTM, and a rather risky one at that.
"The worst mistake you can make in RTM is appearing inauthentic," says Rachel Farrell, senior content producer at Imagination. "When a brand is trying to news-jack something that doesn't make sense to their brand, or when they execute something as 'real time' but it's very clearly not, fans will call you out on it."
But high risk often means high reward, which probably explains the temptation to news-jack.
Consider the case of The Illuminati, a clothing brand that felt compelled to weigh in on the Trayvon Martin story in the wake of George Zimmerman's acquittal. Here's what the brand tweeted just after the verdict: "The Only Justice for Trayvon Martin is to take the Life of George Zimmerman. #EyeForanEye #ZimmermanTrial."
The brand took that tweet down, but not before The Washington Times made a screen grab.
A little later, The Illuminati had this to say on the subject: "We do not support any violence in reaction to the #ZimmermanVerdict however we do believe in #JusticeForTrayvon."
OK, so this is a clear example of overreach. It's also a clear example of stupidity. (Although arguably, the content of the message did resonate with some, even if there wasn't a clear connection to the brand's values.) But while it's easy to dismiss The Illuminati's tweet as an outlier because it's just so damn dumb, the fact of the matter is that all brands need to be prepared to respond to mistakes, which can happen often in the news-jacking context.
In a word, that reaction should always be about transparency. But if you're having trouble figuring out what that means and how to execute in the wake of a screw-up, Ed Lee, senior director of social media at Tribal Toronto, has some advice.
"Social media is simple, and we make engaging with it far more complex than we should," says Lee. "RTM humanizes brands, and humans make mistakes. Obviously, brands are held to higher standards than humans, or at least their mistakes are more obvious because they reach so many people. Nevertheless, they should act as humans do: 'fess up, apologize, and try their hardest not to do it again. Humans are a forgiving species, and we reward honesty and transparency."
But is news-jacking a bad idea? No, not always. As Steve Hall pointed out in a post at HubSpot, a lot of brands made the most of the Oscars. True, that was a planned event, which allowed those brands to prepare content in advance, but some brands like Special K were able to news-jack, adapting on the fly as events unfolded in real time.
"When 'Life of Pi' VFX supervisor Bill Westenhofer's acceptance speech ran a bit long, he was played off with the 'Jaws' theme (planned by the organizers to cut off long-winded speeches)," Hall wrote. "The Special K marketers wasted no time jumping in and having some fun with a famous quote from the movie."
Of course, Special K's news-jack, while witty and topical (the brand's Twitter copy reminded consumers to keep their acceptance speeches "snack size") also underscores the risk of playing with news. The big story that came out of "Life of Pi" was actually a controversial one, because director Ang Lee came under fire for failing to acknowledge the contributions of VFX artists, who have been under enormous economic strain.
Consumers never did connect Special K's joke with the controversial aspect of the story. But in retrospect, I doubt the brand would have wanted any part of it if it had known about the controversy. Of course, that's precisely the point; the risk that comes with news-jacking is that most of the time you don't know where the story is going to go.
Not all news stories are a surprise. In fact, there are plenty of big and small planned news stories every year. But perhaps one of the biggest and most recent was the birth of the Royal baby. That momentous occasion was actually something of a jump ball for a lot of brands, as evidenced by Lee Newton's Engauge blog detailing the winners and losers in the Royal baby bonanza.
According to Newton, brands ranging from Bud Light to Hostess Twinkies might have gotten some viral mileage out of their Royal baby RTM, but the creative came off as a bit of a stretch. In the end, neither found a message that really resonated. Hostess was just weird and a little creepy. And Bud Light essentially reminded consumers to have a drink with its "Keep Calm and Drink Bud Light" message, which I suppose was meant to play off of a would-be father's anxiety around a delivery. Then again, I'm not sure there's much overlap between expecting dads who find humor in soothing their nerves with a cold beer and the people who obsess over all things Royal Family.
So which brands did better? Newtown has a list that includes Disney, Charmin, and yes, Oreo. But for my money, Mini Cooper is the best example of a brand that delivered, because it created a high-concept, polished ad that built on the brand's British roots.
But for all the brands that tried to get in on the Royal baby announcement, no one brand owned the day or came anywhere close to an Oreo moment, according to Linsley.
"Brands need to apply more rigor when determining 'what make's sense for our brand,'" says Linsley. "The birth of the Royal baby received an RTM blitz, [but] there was no one who stood out and created resonance with consumers, and ironically marketers were able to plan for that event. We as an industry need to do a better job of putting on our consumer hat when it comes to real-time content and asking ourselves, 'Would I share that?'"
Make your own fun
Long before Oreo sparked the current conversation about RTM, there was Old Spice's "Man Your Man Could Smell Like" campaign. The campaign didn't begin in real time, but its second act, which featured 200-plus real-time reaction videos, really was an RTM coup.
But according to Caro, marketers shouldn't fixate so much on the RTM aspect of the Old Spice campaign. What really made the difference, she says, was actually integration across all platforms.
"Each successful brand starts with a channel-agnostic brand idea and surrounds that idea with a truly integrated ecosystem (online and offline)," says Caro. "Every channel maintains the same look, feel, voice, and tone. Each channel serves a distinct role and purpose. In this real-time world, every channel must be as relevant and timely as possible."
In other words, RTM is one piece of the larger picture. Social, by its very nature, happens in real time. But for brands to really break through, they need an idea that can work across an integrated network of platforms. When that happens, the brand is in a position to hold its own conversation.
Michael Estrin is a freelance writer.