The Attribution Game
Smell the coffee (public domain image from pixabay.com)
There are many digital marketing platforms in the market that wrangle data and wring out insights. Two firms decided to take that a step further by sharpening their answers to “why” things work the way they do.
Phoenix-based Proof just released its platform, which actually computes the effectiveness of a sales and marketing campaign, breaking it down segment by segment and showing the relationship of all parts to producing an outcome. Today, Menlo Park-based Quantifind released SIGNUM Impact, which aims to quantify the ROI for ad and marketing campaigns, looking for effective factors that are buried in the data but still produce an outcome.
What both companies have in common is a quest to simplify the complicated, rendering that answer in a graphically simple fashion. In effect, one good visual is worth a thousand data points.
The Proof is in the Outcome
“We created something normal people could use without giving up accuracy and fidelity,” said Mark Stouse, CEO of Proof. “We have not reinvented the math wheel.”
Rather than constructing a new model, Proof relies on “algorithms known and accepted by data scientists,” Stouse said. Then Proof puts that data into context, displaying each step in a sales and marketing process. Each point in that process is shown with a simple circle graph with a single-digit number to express effectiveness.
The most important data point in the graphic is the “time shift”, which shows how many weeks or months have passed between an action taken and a result showing.
Explanatory graphics can include 20 or 25 of these “blocks”, all laid out in a sequence to show how each step feeds into the next, Stouse explained. Users can spot the places where leads are turned into prospects, handed off from marketing to sales, and converted into final sales, noting the drop off from step to step over time.
“This is a huge pain point for marketers and business people. Everyone wants the problem solved,” Stouse said.
The Signal is in the Noise
Quantifind's approach to gauging effectiveness comes from its heritage—its data approach was once used to predict terrorist attacks, as well as hit movies. “Once the signal is found…we make the signal explorable,” said Joshua Reynolds, head of client consulting and marketing. Hence the introduction of Quantified's SIGNUM Impact, which can help develop insights into why campaigns work, and why they don't.
Consumers express a preference for a product, but that expression can be scattered across of wide range of sources—social media, blogs, credit card transactions, forms, surveys. “These are all pockets where consumers express themselves,” Reynolds explained. “It's messy.”
About 15-20 percent of this data correlates to revenue, Reynolds said. Quantifind filters out the dross to get down to that correlative data, and then screens it again to remove coincidence. Once a user can spot the key piece of data that drives revenue, one can then calculate the ROI of an ad spend, or change the campaign to improve results. Reynolds referred to this as “exploratory analytics.”
Despite all the math, humans must be kept in the analytic loop, Reynolds pointed out. Curiosity and intuition can tease out a relationship between seemingly unrelated pieces of data to yield an insight.
Reynolds offered as an example an ad campaign aimed at teenagers for a fast food outlet. The food tested well with the demographic, but the campaign had limited success. Why?
The answer was “coffee”. The teens weren't drinking it. But their moms were, since they drove the kids out to get breakfast. Improve the coffee for the moms and you literally drove more sales of the teen-friendly breakfast, Reynolds noted.