Using Marketing Analytics to Measure the ROI of Marketing Activities
Unlike sales, R&D and operations, the marketing department’s contribution to the company’s bottom line is still a field that is not fully understood. Although CXOs do have a basic understanding of the qualitative value that promotional campaigns bring to the table, whether it’s an email series or a social media contest, they are always just one bad quarterly earnings report away from trimming down the marketing department’s budget. In fact, one out of every three company executives hold the CMO accountable if the enterprise fails to meet its growth targets. This becomes even more evident when you consider CMG giant Coco Cola’s choice of doing away with the CMO title altogether and appointing a new Chief Growth Officer (CGO) instead.
All these indicators point to the fact that marketing is now directly answerable for an organization’s growth. The executive heading marketing is still responsible for building and maintaining a brand’s image, generating leads, cultivating customer loyalty, and ultimately driving growth from the bottom up. Each step of the way, this individual will be tasked with establishing, qualifying, and continuously demonstrating the causal relationship between building the brand and expanding the bottom line.
In the wake of the digitization and data revolution, it has become considerably more complex to provide a satisfactory answer to the question: “What kind of return on investment (ROI) do my marketing campaigns deliver?” The increasing dominance of unconventional distribution, sales, and service touch points like e-commerce channels and social media, compounds the complexity, leaving marketers wondering at what point, in which way and by which influencer was the campaign performance affected. In addition, organizations even need to consider the impact of external factors like the state of the stock market or the weather.
The Benefits and Limitations of ‘Touch’ Analytics
This is where data and marketing analytics come in to play a critical role. As challenging a proposition as it may be to measure the success of a marketing campaign, there are a number of widely accepted methodologies which leverage algorithmic attribution models and advanced analytics services to do just that.
By and large, the single touch attribution (first touch/last touch) method is preferred by as many as 45% of companies owing to its low cost and relative simplicity in implementation. In this scenario, the revenue credit attached to a closed lead is merely credited to the first or the last marketing program that the prospect came into contact with before converting. This method provides an easy way to gain directional insights into the early stages of the revenue cycle, but it neglects to take into account the contribution of other touch points. More often than not, the results are considered not totally reliable, especially in the context of large deals and protracted sales cycles.
Since the first/last touch attribution model does not account for time-to-investment, some companies choose to adopt the single attribution methodology with an added revenue cycle projection tacked on to the first touch. This helps marketers estimate what the total future revenue impact of a specific campaign would look like. However, this approach suffers from a vital flaw – just like the single channel attribution model, it fails to account for the effect of subsequent touch points. The multi-touch attribution modeling method is a significant improvement over this and considers the value added by multiple touches from multiple people that contribute to close a deal. It generally involves analyzing data associated with a particular action like sales pipeline creation and requires the analysts to work backwards and identify each touch point that a prospect came in contact with before converting. This method allows businesses to gain deeper insights in terms of channel attribution, and helps them identify which campaign has delivered the maximum ROI using weighted factors such as the time elapsed between a touch, the action that delivered value or the prospect’s role as a purchase decision-maker in the target organization. However, using weighted factors can add a degree of bias to advanced marketing analytics and therefore often increases the risk of over-crediting certain touch points.
A good attribution model should not only consider all the touch points involved to make a conversion but also the advertising exposure on the Adtech side and the content/offers used in those campaigns.
Bringing Statistical Rigor to ROI Measurements
The age-old test and control group campaign analytics approach has proven its worth time and again, to help marketers measure the incremental impact of marketing efforts on revenue. In terms of design, this model has greater success when it comes to revealing the actual impact of a marketing campaign. With a decent control group in place, it sets itself apart as a relatively cheaper but far more robust model to measure the impact of marketing activities. Using two distinct but homogenous test and control groups, you can measure the effectiveness of a particular marketing campaign by introducing it to the former and not to the latter. Provided all other things are equal, this methodology can help marketers discover any positive difference in customer buying behavior and link it to the relevant campaign. However, we recommend proceeding with caution while analyzing the results of such a ROI measurement strategy, particularly in terms of factoring in natural statistical variance before comparing responses from the test and control groups.
If extensive statistical rigor is what you seek, marketing mix modeling (MMM) analytics is the ultimate tool for measuring and validating campaign ROI. The relatively high cost of implementation means that currently only 3% of marketers use it. Don’t let that put you off though as it has been shown to improve marketing campaign effectiveness by as much as 40%. Essentially, this method leverages machine learning-based statistical techniques like multivariate regression and vector auto regression to correlate a campaign’s outcome with independent marketing touches and other non-marketing factors. It uses this data to develop a holistic view of how each factor has contributed to a campaign’s performance and to what extent. There are only a couple of problems when it comes to embracing this ROI measurement method on a large scale – MMMs need to be fed with large volumes of historical data for them to become effective. They also require advanced analytics models designed by expert statisticians who work closely alongside marketing experts in that business domain. But fear not, for help is at hand. Even if the existing enterprise IT infrastructure is not designed to support MMM analytics, third-party analytics service providers can help fill the gap with their own set of tools and expertise.
Measure Performance to Maximize Results
It’s no easy task to sift through the varieties of models available today to measure the efficacy of your marketing programs. But, as marketing becomes a key driver of growth (some would say ‘the’ key driver), a few well-thought out measurement strategies that rely on sophisticated marketing analytics solutions, could go a long way towards fine-tuning your marketing strategy. This will in turn, enhance your ability to predict the outcomes of future marketing programs and give your department the confidence they need to invest strategically.
- App Engagement
- App Push Notifications
- Artificial Intelligence & Machine Learning
- Campaign Management
- Email Marketing
- Email Security
- Growth Marketing
- In-App Messages
- Marketing Automation
- Mobile Marketing
- Powering Smart Conversations
- Product Update
- Push Notifications
- SMS Marketing
- Tech startups
- Web Engagement