The false economy of digital econometrics
Data, data, everywhere – but none to help me think.
The term ‘Big Data’ is the jargon du jour, but little operational excellence has come about in the pharma industry to justify the buzz. Partly this is due to the fact that leveraging ‘Big Data’ is an internal and operational challenge, one that the typical agency is unqualified to surmount. The other is that, and let’s be real here, the ‘Little Data’ around digital marketing is mostly bunk, and you don’t get permission to do the big things unless you do the small things well.
What do I mean by ‘bunk’ you ask? I believe that pharma has been blindly living in an accepted false economy of digital metrics, and unless a concerted effort is made to correct the problem, digital will continue to remain a second or third class citizen in the overall marketing mix. These challenges center around three specific issues: The mythology of ROI, a lack of integrated analysis, and the continued ignorance of the industry to pay anything but lip service about wanting to fix the problem. I’ll unpack these one by one.
The ROI Myth
Somewhere Don Bartholomew is probably cringing about what I’m about to type, but I’ll plow ahead anyway. You can’t define ROI in digital without having a program that includes last–click attribution which directly ties to a sale of a product. In pharma that means if a brand cannot directly associate the fulfillment of a script with a digital engagement, you can’t, under any circumstances, show a true ROI. Outside the industry, e¬–commerce and couponing do this routinely. In that world, you can clearly see if someone clicked on a banner or website and ordered my stuff. This is a clean ROI analysis. In pharma, the only possible avenue for this kind of data is usually through loyalty or copay cards. Does your brand have this? If not, no ROI for you.
If you look carefully around the industry, you will see the tangible manifestation of this reality. Agencies are talking about ROE, or Return On Engagement, as the new means of defining marketing productivity. The end result of this approach is that most of the data collected from digital programs only show the viability of the tactic, not whether it was effective or not at lifting share. Web site visits or time on site only tell you just that, they actually don’t give you any indication on whether that behavior did anything to move someone along on the continuum to request your product. For all we know, those visitors may already be taking your brand. This issue has become so acute that there’s now a pervasive movement in the industry that argues marketing effectiveness shouldn’t even be considered, because ‘customer centricity’ should preclude everything, data and effectiveness be damned.
Exacerbating this issue is the lack of experience (or desire) by clients to want any kind of substantial metrics on campaign effectiveness. Most digital marketers enter their jobs at the lowest levels of the brand team. They often have no experience in the digital arena and are only interested in showing success in order to meet their objectives and get promoted. I can’t blame them; because that’s the environment they’re operating in. You and I would do the exact same thing. Unfortunately the lack of experience in measuring campaign and marketing effectiveness has left a lot of room for snake oil data to be sold throughout the industry. Recently I sat through a presentation by a company that promised to use “data to drive the brand business.” The presentation posited that because a website had 3+ page views, those users were certainly going to ask their physician for the product. Their data was based on a correlation score from years of metrics collected across a number of brands. (They argued this held true regardless of category. So an athsma sufferer behaves the same as a schizophrenic. I remain skeptical). I had several questions. First, if 3+ pages grantees a brand request, why don’t we just make the page content shorter? Second, How can you KNOW that 3+ pages equal a product request, since there’s no corresponding data to prove it? Finally, without accounting for billable hours, media spend, internal time, out of pocket expenses, how can you calculate any kind of ROI since the data doesn’t include the costs of actually producing the marketing?
My questions were met with a number of vague and inconclusive responses, with the presenter finally claiming that, “When we first ran the model the brand had a 70:1 ROI and we thought that was high, so we reduced it to 7:1. Is it really that hoard to believe you had a 7:1 ROI?” I’m a data atheist. I don’t want to believe, I want to know, and this person basically admitted they made the number up. If your model is that far off, maybe your model is hosed.
The problem was the client loved it. “My boss will be thrilled.” Who cares if those numbers are a fantasy, right? Never let data get in the way of a good story, I suppose.
The Integration Problem
I’ve been in this business for almost 20 years. I can probably count on one hand the brands that insisted that their agencies report metrics and data in a consolidated fashion. Having also spent a great deal of time outside of pharma, I can tell you that this lack of integration would be unacceptable in other categories. Pharma has an infatuation with pitting agencies against one another, to the point where I believe they feel that we shouldn’t even share or collaborate on the data showing marketing performance. That’s especially weird, because those same clients are paying for that information.
For example, there’s no way to know if search is effective if you don’t track behavior once that person gets to your website. It’s great if your Facebook page got 200 likes this week. Any idea what they did after that? I’ve written before about the PR/Marketing divide, and that IS real. But the internal marketing silos are also real, and they prevent a brand from truly knowing how effective their marketing dollars really are.
I once had a client tell me that I wasn’t allowed to see the data from an NPP program because I wasn’t “the agency that handled that.” My reasoning, I explained, was I wanted to see if the tracking tags were working correctly (as we were experiencing a weird influx of unattributed traffic) and I wanted to segment out the behaviors of those visitors to the HCP site to see if they were doing anything more than bouncing off the pages. “No,” I was told, “You’re probably going to use the data to try and steal their work.” I called the NPP agency and got the info I needed in 5 minutes. That example is not an uncommon one and it begs the question: Are you more interested in babysitting your agencies or are you more interested in how your marketing dollars are performing? Clients need to be the arbiter of collaboration amongst their agencies. They alone set the tone for how the relationships are expected to work. Without partner integration and collaboration, your marketing data will be superficial and utterly devoid of context.
The First Step Towards Fixing The Problem Is Admitting You Have One
The agency world has been griping about wanting to fix this problem for years, but the solutions are often centered around larger fees, a custom dashboard, or some other kind of ‘solution’ that particular business can provide. Sometimes those programs have merit, but often the solution is as simple as working with your agency partners to create an integrated report detailing effectiveness. Collectively, we as agencies should stop allowing our efforts to be so easily dismissed because the data and metrics aren’t being appropriately represented, regardless of how much money we could charge by ‘doing it ourselves.’
And clients, you should be jumping up and down on your agencies to provide you meaningful insights about how your programs are performing. You’re paying a fortune for these programs; don’t you want to know how well they’re working? It’s also inherently critical that you fund the resources necessary amongst your agency teams to provide these analyses. It doesn’t have to be expensive, but it does require some leg–work. Cutting these resources will leave you flying blind.
Which brings me finally to my friends and readers in various procurement functions around the industry. Your group, more than any other, should have a vested interest in this issue. Classically, procurement groups are tasked with minimizing costs to maximize value. This is almost always centered on the billable hour and reducing those rates. But isn’t a higher performing campaign and agency worth paying more for? How do you equate value devoid of performance? Through a lens that ignores performance the idea of ‘value’ becomes completely arbitrary and is more aligned with ‘cheap’ than ‘effective.’
For all the consternation by agencies and clients that digital remains the red–headed stepchild of the industry, we’ve done ourselves little favors in proving business effectiveness. If we can solve that challenge, there’s no doubt in my mind that the channel mix will change dramatically. Are we courageous enough to make it happen? We’ll see.