“Big Data” in Marketing: 3 Prep Steps
Guest blog post from Fan Foundry. Original post here.
In preparing client case studies for my talk titled “Be a Big Data Voodoo Daddy” at the October 2012 FutureM conference in Boston, I noticed that almost half of our client projects over the recent years have evolved from “Implementation” projects to “Readiness” projects – equally valuable, and absolutely necessary. How’s yours going?
Is your marketing automation, CRM, analytics, email marketing or other automation project going to deliver your desired payback? Here are my top 3 warning signs that it may take longer to pay off than you think.
Stated differently, here are 3 must-do’s to ensure near-term ROI.
1. The Right Stuff.
Often we find that you are not gathering useful, relevant data to help you accomplish your stated strategic goal. This may stem from having broad, imprecise goals. For example:
“Grow revenue” is a great goal, but the paths are varied and nuanced.
“Increase Partner Channel Revenue” is, well, getting warm.
“Double Partner Channel Service Contract Revenue” is more like it. Now you have a specific channel, identified players, and a specific product/service element attached to a numeric goal.
Having specific, measurable goals and then measuring the right things are both essential elements if you are to to yield any meaningful process automation and relevant data analysis. No matter how efficiently you automate the wrong data, you risk stretching out the time horizon for any meaningful payback or, worse, running in multiple or wrong directions and wasting effort.
2.) The Stuff, Right.
Typically, your data is not homogeneous. It ofte exists in a variety of formats ranging from locked spreadhseets and various departmental databases to unstructured documents, such as paragraph text and visuals.
Significant effort is involved in standardizing and preparing data for upload into your new automated solution, as well as selecting the right tools to enable you to access and mine insights from unstructured information. We are familiar with an array of powerful tools, and can develop custom, reusable upload frameworks to help clients address current and future needs for unstructured data.
This is where the scope of a project almost always expands, as additional valuable information repositories become included. Upside: more data. Downside: more expense. Ultimately, though, it proves worthwhile to expand the project and include the data, because we often discover additional formats, sort fields, and other new requirements that deserve consideration and inclusion.
c) The Players.
The talent shortage is legendary. If you are inadequately staffed or trained to assume the role of data manager, analyst and strategist, let along carry on administratively after implementation, you shouldn’t start the project. The time to assign roles is up front. Get any necessary hires onboard first so they can be involved in the project. The single most effective way to stretch out the payoff time horizon is to not involve its eventual owners and primary users.
The full list of must-do’s is extensive, but if you tend to these three first, most of the rest will fall in line, and you’ll enjoy a successful implementation.
Toward a “Measurement Culture”
You’ll know you are succeeding when you have established a “culture of measurement” in which the right things get measured, the data supports meaningful analysis, all meaningful data is reflected in a single, integrated, centrally accessible “record of truth”, and you are achieving your strategic and tactical goals.
Even if you are not achieving as intended, your data and measurement will help you understand why so you can confidently pivot, adjust and redirect. Finally, it must be stressed that human judgment is not taking a back seat here. Interpreting analytics in light of pragmatic experience and taking calculated risks are the hallmarks of innovators.
Check out the last post in this series, by Subu Desaraju.