Marketing is an investment, not an expense — or at least, it should be.
A lot of people approach marketing like they’re throwing darts blindfolded: they know roughly where the board is, but they don’t know if they’re hitting the target and can’t evaluate afterward.
That’s not effective. You need to know exactly where your money and resources are going, why they’re going there, and what return you’re getting.
“What gets measured gets managed,” goes the old business axiom. But in some cases, marketing can be difficult to measure. How do you measure your return on investment (ROI)?
1: Decide What Return You Want
First, you need to define what kind of return you’re looking for.
Return isn’t always monetary. Sales benefit from any marketing effort, but sales are not always the aim. Every marketing campaign needs a clear goal. The most common goals include:
- Driving Actions: Action campaigns try to get the audience to do something. Actions could be downloads of a whitepaper, clicks on a link, signups for an event or anything like that. This type of goal is often tied into a longer term strategy.
- Boosting Sales: The most common goal businesses ask marketers to pursue. These campaigns can take many formats, including remarketing, shopping cart followup, touching base with former customers and more. They tend to be short-term and tactical.
- Building Engagement: How long do people spend on your site? How often do they read, like, and share your social media posts? Campaigns that focus on engagement look to increase metrics like these.
- Generating Public Relations and Goodwill: This is the hardest category to measure. How do people think about you? Do they care about your company? Brand sentiment is a term you’ll often hear applied here. These campaigns look to blunt negative brand sentiment or reinforce positive perceptions within the target audience.
2: Set a Baseline
Figure out how you measure success. What type of return are you looking for?
Set a baseline, run the campaign and then measure against that baseline. For example, you might look at your Facebook page and see that you’re averaging 1,500 engagements every month. That’s your baseline. You decide to run an engagement campaign to increase engagements. You think you can boost engagement with a paid social campaign with a goal of 2,000 engagements per month. You run the campaign for a month and reach 2,100. The campaign is a success and based on that, you can adjust for next month.
3: Determine Your Yardstick
Once you define the baseline, you need a yardstick to measure success. There are two types of measurements you’re going to be looking at: quantitative and qualitative.
Quantitative measurements include efficiency metrics like cost per click, cost per engagement and average customer value. They also include things like follower count, hits, conversions and time on site. Your yardstick for these will be a concrete number. Note: as Disraeli once said, “There are three kinds of lies: lies, damned lies, and statistics.” Even if the numbers are accurate, they need to tell the correct story and be relevant to the goal.
Qualitative measurements include intangibles like goodwill, market position and brand sentiment. These goals don’t have a ready-made measurement built in (you can’t ask Google Analytics to measure goodwill). The yardstick for these has to be based around surveys, customer feedback, social media monitoring and other methods you can use to quantify feelings.
4: Choose an Attribution Model
Quantitative
Some quantitative goals may be straightforward — for example, follower count, likes and time on site are all one-number metrics that you can pull without much effort. If you’re measuring something like conversions or traffic through a particular source, though, you have to dig deeper.
To calculate return on investment, you need to be able to take credit for driving an action. This is called attribution. As Rand Fishkin over at SparkToro has pointed out, it’s hard to figure out where your customers actually come from. It could be search, email, dark social, actual social … and even if you have “last touch” attribution, it may not be where the conversion came from originally.
First- and Last-Touch Attribution
Most models use first or last touch attribution, which is simple enough: either the first or the last touchpoint before a customer takes action is why they did it. It’s simplistic, but easy to set up.
First-touch attribution emphasizes the top of the funnel: where did that customer come into contact with you first? If you can’t get people interested or your organization is relatively new, first-touch attribution can help you decide which audience-attracting techniques work.
Last-touch attribution emphasizes the last step that led to action. If your top of funnel is well defined, last-touch attribution can help you figure out whether you’re efficiently turning prospects into customers. Google Analytics defaults to last-touch (though it can be used for other types).
There is more to attribution than just the first or last touch, though.
Multi-Touch Attribution
Most of the time, there are more than two touches that you can attribute to an action.
Take this example: Some of the plants on my property are getting a little out of hand, so I decide to buy some clippers. I do a Google search for “hedge clippers” that leads me to AcmeHedgeClippers.com. I look at the site and leave. A few days later, a retargeted ad catches me on another site and I click through again. Then I browse the blog. Later that day, my brother sends me a note about a sale on AcmeHedgeClippers.com; I type the website into my browser, go straight to the site and buy my tools.
Which touch is the one that made me purchase?
Multi-touch attribution tries to assign value to touchpoints throughout the user journey. Some models weight every point the same. Others may weight some data points higher and others lower (like U- or W-shaped attribution). This type of attribution helps you find “holes” in your funnel and patch them up.
Qualitative
Some returns on investment aren’t easy to measure — brand sentiment comes to mind. How do you measure those?
Take inspiration from your forebears. Before the days of Google Analytics and social media dashboards, you had to get creative. There were no cookies, no IP addresses, nothing like that. You had to put in the legwork.
Surveys are a great way to track your ROI on something like brand sentiment. A simple survey example you can use is a popup; have it show up for logged-in users and ask “How likely would you be to recommend this service to a friend or colleague?” Offer a scale from 1 to 10; 9s and 10s are promoters, 7s and 8s are passives, 1–6 are detractors. Subtract your detractors from your promoters and you get what’s called an NPS score: a quantitative measure of your brand sentiment among users.
Focus groups can be a good way to measure qualitative return, too. If you’re trying to make inroads into a new market, pull a focus group from that market and ask them questions.
Social media monitoring is often thought of as quantitative, but it can also be used for qualitative analysis. Some monitoring suites let you track overall tone of conversation or velocity of comments, which can give you a better picture of intangibles like brand sentiment or a general impression of a particular product, service or campaign.
These methods are becoming more critical as governments and platforms crack down on tracking and user privacy. Changes like iOS 15 mail privacy (which always counts a delivered email as an open) or Europe’s GDPR make attribution harder. Surveys, focus groups and enhanced data modeling will be important for your attribution moving forward.
Don’t Throw Darts in the Dark
Marketing should be an investment, not an expense. But it will be an expense if you don’t measure ROI. Figure out what return you’re looking for, set a baseline, decide how you’re going to measure, find an attribution model. It’s simple, but critical. And if you need a hand with your marketing ROI, feel free to reach out. We’re here to help.
Best Regards,
David Brandon
Copywriter
Rainmaker Digital Services