“The number one benefit is that we're able to aggregate and look at what is happening around the world.” – Harris Thamby, Microsoft
There’s just no way around it: measuring marketing ROI is a challenge.
Even with sophisticated technologies in place, it takes strategy, communication and process to tackle the many moving parts involved. No two companies approach marketing ROI in exactly the same way.
That’s why we wanted to share the very interesting — yet very different — journeys two of our customers undertook in their pursuit of excellence in marketing results measurement. We invited Microsoft and Red Hat to share their stories on a webinar.
We learned a tremendous amount from the time we spent with Red Hat’s Ryan Danner and Microsoft’s Harris Thamby.
This blog summary focuses specifically on Microsoft's story.
When Microsoft’s marketing operations teams came to the conclusion that they wanted to improve the way they measured and optimized their marketing spend, Harris Thamby led the charge.
As Senior Program Manager for Marketing Planning within Marketing Engineering, he had the perfect skillset for this job. Introducing process change into an organization at Microsoft’s scale requires a mix of project management, technical and leadership skills, and Harris combines them all.
Harris began his story by pointing out the two most important factors that were the catalysts for change in this area at Microsoft:
1. A lack of common performance metrics
While all of Microsoft’s marketing teams had access to data about the results of their marketing programs, the way the teams viewed this data was inconsistent. Some folks were taking a very granular view, evaluating specific metrics such as impressions, clicks or attendance. Others took a high-level approach at the very other end of the spectrum — evaluating performance by program or multi-tactic level, for example.
“Each team and the groups within those teams were measuring the results in very different ways and different levels,” Harris noted.
This made any kind of benchmarking or comparison across marketing groups impossible. “When it came to ROI, we had lots of different “r”s and lots of different “i”s, Harris said. “That inconsistency made it very hard to compare and very hard to make a decision on future investments.”
2. The operational burden of ROI measurement
Measuring ROI took time and effort on the part of the marketers. Because of the number of disparate marketing technology systems they used, there was a large amount of back-end operations and procedures that had to be followed in order to get all their systems talking to each other.
”We were operationally burdening our marketing teams to manually instrument almost everything we did,” reflected Harris.
It’s worth noting that of course, as a very analytical company, Microsoft had been doing ROI measurement for a long time. The challenge was being able to do it with consistency and efficiency at scale.
“The scale of our operation right now is approximately 500+ [marketing] users across approximately 70 countries,” Harris said. “This is where the standardization and the kind of common taxonomy becomes the most critical part.”
Harris’s Approach to Marketing ROI Measurement
For Harris, meeting this ROI measurement challenge was about zeroing in on a very specific set of smaller steps. He knew it would take a combination of strategy, technology, and careful change management.
Microsoft needed a way of structuring and linking together all the various volumes of marketing data that were being gathered every day across the world. They needed a single system of record for marketing.
Here’s what he and the teams chose to focus on in particular:
1. Defining the level of granularity for measurement
With the massive amounts of marketing results data being generated at Microsoft every day, it was crucial for Harris and his team to be very strategic about their choice of “level” at which to analyze ROI.
They had to strike a delicate balance, not wanting to measure at too high a level (ROI by country, for example), nor too low a level (like examining the ROI of every link on every web page).
2. Prioritizing a few key “slicers” of data
Throughout the many internal discussions at Microsoft that preceded the changes that Harris would set in place, marketers came to him with requests for “slicers” of ROI they wanted to be able to measure.
Examples of these were Engine, Campaign/Program and Country. They differed wildly from each other — which is natural, given the size and scope of Microsoft’s many business units.
Harris was able to broker an agreement. He and his stakeholders settled upon three to four common slicers of ROI data that would be made available to all marketers, no matter their business unit, region or function. This would allow for standardization across the marketing organization. The more standardized things became, the more useful the ROI data would be.
3. Intense communication and continuous preparation
Harris and his team spent several months evangelizing and communicating the data model they developed before setting anything up in Allocadia or other technologies.
They held weekly calls and briefings to explain to the various marketing teams exactly which data and metrics they would now have access to as a result of this project, and exactly how they would get it.
A framework for ROI measurement at Microsoft
After the work mentioned above and an exhaustive study of the marketing technologies in use at Microsoft, Harris and his team came up with this framework:
- Allocadia became the system of record for field marketing activities. Every marketing activity in the field is tied to an ID number in Allocadia. “We made the activity the central hub of metadata that we were going to use to connect all these pieces together,” Harris said.
- As the green circles below show, each activity can be connected to different kinds of technologies used for marketing execution. A single activity (a marketing campaign, for example) could have components executed in Marketo, Certain, and Marin Software (for advertising).
- Every purchase order or dollar spent is also tied to an ID number. Allocadia is also used as the central system to manage field marketing spend.
- This way, every marketing activity has both the “r” in ROI (performance measurements) as well as the “i (execution costs).
Creating this framework centered around the activity (ID number) in Allocadia gave Microsoft a common structure and taxonomy, paving the way for useful and consistent ROI measurements.
Results of this ongoing journey
Why has Harris’ team’s mission to bring ROI data to Microsoft’s marketers been worthwhile?
Visibility into all global marketing activities
“The number one benefit is that we're able to aggregate and look at what is happening around the world, he said. “Not just what’s happening in general but what’s happening at the same grain.”
His team’s very structured approach to ROI data means that members of the various marketing teams have access to benchmarks; they can easily see what’s happening across a particular product or across the world.
“It gives us information on how much stuff we're doing, when are we doing, what's landing, what's planned, what's out there.”
“This helps us in terms of decision-making. It helps us to understand: are we over-investing? Are we under-investing? What can we do about it?”
Everything’s connected, including the dollars
Allocadia is now the central point connecting Microsoft’s activities, as well as its marketing spend.
“100% of our marketing dollars are connected,” Harris says. “We now have a central place to connect marketing activities and spend to other parts of the lead flow mechanism.”
Harris’ slide summarized the results to date of his journey towards better marketing ROI measurement using Allocadia.