3 Hacks for Marketing Measurement Problems

Perfect is the enemy of good – at least, in marketing measurement. Demands from a CMO to know “exact ROI” can be just as inhibiting as an analytics department head with zero tolerance for estimates. Hacks are a shortcut, and like all shortcuts, they represent a kind of trade-off between the marketer’s need for a rapid result and their desire for accuracy, completeness or rigor.

We’ve picked three common issues that can give you a feel for the hacker’s way and inspire your own thinking. (For more on these and two more top-secret bonus hacks, clients login here.)

#1 – “I Can’t Measure Brand Campaign Impact”

Challenge: You run a TV, radio, print or out-of-home campaign – or one that uses a combination of these and online tactics – but there is no direct-response message: no “Call This Number” or “Input This Code on Brand.com” to help you measure impact. How do you know what kind of impact it had?

Hack: The most easily available measurement signals you have come from platforms you use every day. In this case, you can use web and mobile analytics tools, search platforms like AdWords, and social listening tools to estimate the incremental impact of campaigns.

• Web and mobile traffic – compare unique visitors and engagement around the time of the campaign and for some period after. Pay attention to engagement with pages and content that are linked to the campaign’s messages.
• Branded search volume – track the volume of searches in Google, Bing or other engines that use your brand’s name, product or campaign tag lines explicitly.
• Branded social volume – track brand, product and campaign-specific posts on social channels, blogs and other public platforms and compare to norms.

#2. Forecast Campaign Spend

Challenge: You – or your CMO or (gulp) CEO – have a specific expectation for a campaign, e.g., a percent return or actual sales number. Perhaps you need to make sure your campaign or total marketing efforts contribute X% to a sales or market share goal. How do you know if your budget is enough to do the job?

Hack: It’s a good idea to sanity-check a campaign by forecasting its impact both forward and backward. What does this mean? Campaigns are built forward: they start with a budget, which determines a media plan; the media plan may (or may not) be used to estimate response rates based in historical values.

It is also useful to start with the desired outcome and forecast backward to see how much you need to spend or do to meet the goal. This method can be used for individual tactics, like a social ad campaign or email coupon promotion, or for larger efforts, like a major product launch. You start with the sales goal and estimate how many conversions you need to meet it, using benchmarks for purchase size and profit margin. You then divide by a conversion rate – again using appropriate benchmarks – to determine how many ad impressions you’ll need. At this point, you can apply impression costs from your media plan to back into a required budget. (Thanks to Nancy Smith of Analytic Partners for this one.)

#3 – Compensate for a Lack of Sales Data

Challenge: You’re in a business where you are driving at least some of your audience into a real-world location, like a car dealership, a restaurant or insurance-broker franchise, or a retail store. The problem is, for whatever reason, you don’t have access to sales data at the level you need (e.g., by store or city), or for the right time periods. Or it isn’t provided rapidly enough. Or your product has very long sales cycles, like cars and mattresses, and you’d like an early read. What can you do?

Hack: The hack here is to find something you can measure that is a reasonable proxy for sales impact. The proxy itself will depend heavily on your business, and it often involves measuring engagement with a particular part of your website or mobile app, or tracking actions on social media, publisher websites, and other external platforms. For example, brands that run campaigns that try to drive people into retail stores will track engagement with the part of their website/app that lists “Store Hours” or their “Store Location Finder” tool. (One U.K. electronics retailer even went so far as so assign a relative weight to a click on a “Where To Buy” button based on the value of the product the consumer was browsing.) If they have any data from surveys or previous studies that indicate what percent of people who look up a store location actually visit, even better.

Do you have any of your own favorite hacks? Let us know in the comments below. Happy hacking!