This article will provide some sound advice about getting data-driven marketing right—from think big but start small (and with the basics), to take an action-based approach, pick your battles, beware of small sample sizes, and more...
1. Start with the basics
It's easy to become obsessed with pinpointing the ultimate ROI-based marketing metrics; but, rather than hunt for 'the killer' indicators, it's best to identify a few easy-to-implement, basic metrics that can provide a sound assessment of both performance and ROI.
Those can include number of leads, number of Marketing-qualified leads (MQLs), number of MQLs pending sales review, and the number of MQLs recycled/accepted. All of those will give you an idea of how to ensure leads have the proper number of touchpoints to be qualified and get to the right salesperson for conversion.
2. Think big, start small
Your data-driven strategy should be a combination of small wins and long-term goals. For example, it's relatively straightforward to track the conversion of leads into opportunities. It's much harder to track the ROI on search engine marketing campaigns if you want to track ROI all the way to closed deals that were touched by 10-20 different campaigns.
Both are important, so get your quick wins where possible and recognize that other data points will take much longer to realize.
For example, in search engine marketing, you will need to tie Google, Yahoo, Bing, and others to your website to begin, then connect your site with Marketo, Salesforce, and your data warehouse. After that, you'll need to do attribution across marketing campaigns. There are various places where tagging and data sharing can go wrong, so be prepared—and allocate time for setbacks.
3. Take an action-based approach
It's easy to come up with a wish list of data... and much harder to find actionable data. All too often, a marketing analysis will take days of crunching numbers, only to discover that there is no smoking gun or major opportunity to optimize the business. The best data-driven marketers will find ways to minimize the time they spend "exploring" data and accelerate their time-to-insight.
When you're considering the data you want to analyze, ask yourself two questions:
- If I had this data regularly, how could I improve my efforts? If you can't answer, there is probably a better analysis to run.
- What is the quick-and-dirty version of this analysis? If a cursory analysis of the data doesn't yield much, it may be an indicator that the data isn't as important as initially thought—or, at least, not as actionable.
4. Pick your battles
There will always be an infinite number of metrics you can analyze; but, most often, less is more. Whatever the major needle-movers are for your business, put 80% of your effort into optimizing those rather than trying to focus on every possible metric.
Once you've picked your battles, the real analysis begins to determine how to improve tracking, fine-tune efforts, etc.; then, over time, you just need to wash, rinse, repeat.
With big metrics, the low-hanging fruit often isn't there, but with focus and dedication these key drivers can improve significantly over time with small, steady enhancements.
5. Beware of small sample sizes
One can quickly go blind trying to track down the cause behind every dip and spike in results. All too often, it will be impossible to pinpoint the underlying cause of those peaks and valleys. Instead of obsessing over them, stay focused on the big picture and overall trends.
It's often better to look at trends or campaigns quarter over quarter rather than week over week if you want draw more accurate conclusions on overall success. That's especially true for international data, because every country has its own ebb and flow due to cultural nuances, market drivers, and societal biases.
6. It takes a village
To have a truly data-driven marketing organization, you need every marketing employee to be goal-oriented and diligent about performing analysis. There are too many nuances and exceptions in marketing to have just one marketing analyst and a CMO who makes all data-driven decisions. As they say, "It takes a village": To ensure the data used to enhance marketing efforts is accurate and trustworthy, you need for the entire market context to be taken into account by all your analysts. Collaboration is therefore key!
7. Create a shared-learning organization
In marketing efforts, it's usually better to have a "long tail" to each campaign, meaning it would runs for many months or years. That approach gives you and your team the ability to message-test, refine, launch, analyze, and adjust as needed to increase the success rate and traction.
Being able to complete the review cycle quickly and incorporate those learnings in near-real-time is key. Having an organization trained in the art of launching, benchmarking, and re-visiting all cross-marketing efforts will pay the biggest dividends over time.
8. The devil is in the details
Having reliable partners on both your operations and sales teams is key to executing any data-driven marketing program. Otherwise, many great marketing ideas will be poorly executed and leads will go untouched.
Thinking through the entire marketing process, understanding the cast of characters involved, and taking steps to ensure all are aligned around a common goal are critical to success.