Measuring the return on marketing spend by channel is essential, but done in a vacuum it can lead to misinterpretation and unintended consequences.

During a recent conversation on the topic with Steve Bonnell, director of digital analytics at unified.agency, he explained how marketing performance falls short of its potential when marketers are too focused on what he calls middle metrics: diagnostic metrics used in isolation for one part of the customer journey, such as click-through rates and landing page conversion rates. If, for example, marketers focus too much on email open rates, they may not pay the needed attention to click-through rates and landing page conversions.

Fortunately, Bonnell also explained how marketers can refocus, viewing channels and campaigns more holistically to optimize marketing efforts across the customer journey.

Why are so many marketers hooked on metrics that tell only one piece of the marketing story?

There’s so much data available today, trying use as much of it as possible is almost over-correction. It’s not so much that the metrics marketers are tracking with all this data are bad or wrong. The issue is the way they’re being view.

A lot has to do with the interpretation. It goes back to adage that you can use statistics to tell whatever story you want—for example, cherry picking data to tell a rosier story. Looking at channel metrics in isolation is one way that marketers can misinterpret results because it doesn’t tell the whole story.

Getting a clear picture of the push and pull of levers across the customer journey is critical today. Marketers need to take a more holistic, strategic view of their campaigns and communications. It’s not just about delivering the connected experience today’s customers expect; it’s about connecting the metrics, as well.

What’s an example of using data in ways that are counterproductive?

Too often businesses aren’t focused on customer lifetime value (LTV). Instead, they’re focused on acquisition and retention separately. A focus on metrics such as LTV comes from leadership; you have to want to know what customer is really telling you.

Consider this example: Company leadership at a financial services firms gives its marketers the goal of a lower cost per lead, and tells sales to increase conversions. The lower-cost leads, however, are of lower quality and less likely to convert. So that creates conflict and negatively impacts performance. Worse, those lower-quality leads could be lower-quality customers if they do convert.

Why doesn’t leadership get it? You don’t want to optimize something that will have a negative effect not only on specific parts of the customer journey, but also on the overall customer journey.

How can marketers avoid these unintended consequences?

One of first steps is reassessing your attribution model. Some marketers are still using last-click. That’s not an accurate measure. An e-commerce site I know of was giving all the credit for conversions to one coupon site, until discovering that getting a coupon was an action its customers took at the end of the process, after they made the decision to purchase.

Attribution should be about understanding the true path of the customer decision and where it was made. The availability of data has exacerbated the silo issue. When looking at data for one channel, consider the impact decisions for that channel could have on the other channels.

More broadly, marketers need to invest in doing the real analysis of the full customer journey through to conversion and beyond, including the entire lifecycle. To do this, marketers need to get cross-functional teams involved. Marketers may think paid search isn’t performing well because it brings in few leads, for instance, but salespeople may say, hey, leads from paid search are high quality and convert well.

Also, marketers should get all their cost data in once place. Marry up the cost of each touchpoint with attribution data. This approach will help reveal such insight as an early-decision touchpoint worth investing in.

Finally, slow down. There’s an overwhelming need for speed today—a “quick, do something” mentality. Speed and agility are important, but be careful that rushing to a decision for one touchpoint doesn’t negatively impact the customer experience or marketing performance elsewhere. Think through all of the pieces and the real value first. Sure, a lead may cost more, but it may also deliver more long-term value and greater lifetime value.

Final advice?

It’s essential to consider and track middle metrics, but make sure they’re not the be-all, end-all. Uncover any issues caused by adjusting for one metric, then look at how any levers pulled may impact elsewhere. Level set to make sure everyone’s aware of risks associated with those metrics.

It’s requires cultural change internally to view how all channels work together, but it’s worth the effort. Considering the impact of middle metrics holistically across the customer journey, as well as individually, will improve the performance not only of those individual touchpoints, but also of final metrics such as cost per sale.