Open up an analytics program and you’re immediately hit with a mind-numbing amount of data.
Eyes glaze over as you quickly scan for the green, up-trending stuff to report back to bosses and clients.
The problem is, that while your analytics might tell you a lot of stuff, they don’t always tell you the full picture.
It’s similar to garbage in, garbage out, where you’re given a ton of black and white data even though your world is varied shades of gray (and that’s assuming you’re not making any obvious, simple mistakes).
Here are five Facebook analytics mistakes sabotaging your campaigns (and how to fix them).
Mistake #1. Incorrect Conversion Tracking
Let’s start with the most simple, basic and straightforward. Your conversion tracking is off.
I recently started working with a client that spent a healthy ~$30k /month. When I began an initial assessment of their campaigns to see how each was performing and compared, I noticed that almost all of them had 100%+ conversions.
Awesome, right? Unfortunately, no. Because those numbers are impossible.
More likely, is that the conversion tracking pixel is applied to all pages (instead of the confirmation or ‘thank you’ ones) – tracking clicks as conversions.
This also commonly happens when the pixel is placed on the beginning of an opt-in or purchasing process (think checkout process or landing page), instead of the final destination URL.
Sure, some campaigns (especially top of the funnel ones where you’re simply generating site visits) won’t have conversions per se. Fortunately, there are still a plethora of potential ‘engagement’ events to track to see if that traffic (or more specifically, that audience you’re targeting) is of sound quality or not.
For example, you can set ‘ViewContent’ as an event to track specific PageViews.
This script is supposed to be located before the closing </head> tag on the specific page in question (the ‘product’ or key page for a top of the funnel, or the ‘confirmation’ and ‘thank you’ pages for a bottom of the funnel goal).
To double check your efforts, head on over to the Chrome store and download the Facebook Pixel Helper extension.
When installed, head over to the page in question and click on the helpful new icon in your browser to get some instant feedback on whether the pixel is properly loading or not.
In an ideal world, you make strategic decisions based on your analytics to dictate the next week or two’s iterative, day-to-day marketing activity updates or changes.
You’re looking at leading indicators (like Cost Per Click) and key metrics (like Cost Per Conversion) to determine which campaigns get more budget, and which get less. Which need new ads, and which need audience targeting adjustments.
Incorrect numbers, based on simple mistakes like not tracking the right page conversions, throw all these decisions off.
Mistake #2. Over-Optimizing Your Campaigns
This next problem is a little more subtle.
There’s one thing all clients and bosses love: graphs that go up-and-to-the-right. Preferably, also if they’re the color green.
Nobody likes sitting in a meeting, trying to explain why certain numbers are stagnating or (God forbid) declining. And red.
It’s uncomfortable.
Beads of sweat appear on your forehead as you try to explain complex, sophisticated topics like the importance of ‘lead nurturing’ prior to seeing conversion increases. But it’s falling on deaf ears (or flying over their head).
Good execution is insanely difficult; often requiring a number of variables to line up properly with tons of iterations along the way. Taking shortcuts can make this much easier and less stressful in the short-term.
For example, a quick five minutes and $five bucks on Fiverr can get you all the likes you want.
But you already know buying likes is dumb.
The stuff you get is junk. Your ‘audience’, will never buy anything.
Many of your ‘bottom of the funnel’ campaigns should be specifically targeting recent site visits and these audience members too. So now you’re wasting money a second time, targeting low-quality people en masse.
Ideally, we’d be focusing not just on CPC or Cost Per Conversion, but actual, honest-to-God revenue and the lifetime value of each customer.
Attract the wrong people from the start, and you’re already screwed.
My favorite example of this comes from Rand Fishkin and Moz, which dove deep into their customer analytics to separate the differences between the customers who churn quickly, from those with the highest lifetime values.
You would think that customers who converted on just their first or second visit are great. Logically, you pat yourself on the back and show these awesome numbers to said stakeholders.
The problem? For Moz, those customers were also the fastest to bounce.
Surprisingly, the people who visited many, many times (like 8+) were also the ones who paid the most and worth the most to the company.
That’s a common scenario for companies with technical products or expensive services, where it often takes a while for people to move at their own pace through their own customer journey.
The lesson? Don’t over-prioritize, or over-optimize for initial conversions at the sake of the ones that matter.
Mistake #3. Focusing on Leads vs. Conversions
Piggybacking on the last mistake, another age-old variation of this problem comes when you’re optimizing campaigns for lead conversions (instead of paying conversions).
Tracking eCommerce is easy. Product sales happen within a few minutes, and you can instantly pull Point of Sale data to see exactly what people spent.
But what about service companies, where a ‘sale’ might take place over the course of two-to-three months after initially becoming a lead?
On top of that huge time delay, service companies also commonly deal have little-to-no connection between sales conversions, accounting or finance, and marketing channels. (We’ll try tackling this a little later on down below BTW.)
Almost every single company on Facebook advertising can tell you the amount of (lead) conversions they’re getting. However, only a tiny sliver can tell you how many buyers (and even more – the revenue) they’re generating.
Here’s how this plays out:
Campaign A is driving 10 conversions in August.
Campaign B is only driving 5.
So what do you do?
As we’ve discussed already, we’re all biased to big numbers.
But that’s potentially catastrophic when they’re leading indicators, because you’ll shift more attention (and spend) to the first campaign.
Even though:
Campaign A only generates 1 buyer, and
Campaign B generates 3.
Not only that, but:
Campaign A’s average sales value comes out $1,000, while
Campaign B’s average sales value is $2,000.
So even though Campaign A generates more ‘conversions’ on the surface – these are still only leads. Campaign B might be generating more, and more profitable customers.
We’ll explore how to fix this in the next two sections, however the critical point here is to be careful of making big decisions (like which campaigns get more or less budget and attention) based on surface level information.
Mistake #4. Losing Offline Conversions
Some people might pick up their phone and call the phone number in your ad. But they’re probably in the minority.
More likely, is they click on the ad and browse around your website for a bit. They might even bounce, and come back at a later date.
Then and only then, after doing their own research (which 77% of B2B buyers do prior to speaking with a salesperson) do they pick up the phone and call the phone number listed on your landing or home page.
The problem? Attribution.
Once a purchasing process starts to go offline, it goes into a black hole. You no longer have a pretty report or graph tracking all of the activity and aligning it back to the channel that generated it.
Tracking in-store sales with a coupon code is table stakes. But what about phone calls, where 70% originate from digital channels?
So what happens?
Your ‘intake’ people (or department) notes a ‘Direct’ phone call. Even though it ain’t direct. That sh!t came from YOUR Facebook campaign. The one you’ve been toiling away at trying to explain to your bosses and clients the value of. And someone else is getting (or taking) the credit.
Let’s close that loop.
Enter call tracking services like CallRail. They have a feature called ‘Dynamic Number Insertion’ that lets you create a pool of available phone numbers that will refer the call back to your primary line. Each new session on your site gets assigned a phone number that now follows them around all the pages they browse.
NOW we’re getting somewhere. You can set these up to run automatically across all channels, or you can even tailor it to focus on a specific channel (like, um, Facebook advertising!).
The result, is a complete individual profile of that person’s session data, along with the pages they viewed, where they’re calling from, and most importantly which session drove them.
You can make this even better by integrating your call tracking services with whatever CRM or marketing automation platform you’re using. For example, CallRail integrates with Salesforce and HubSpot so you can start sending data back and forth seamlessly.
Which brings us to our last, yet most important, mistake.
Mistake #5. No Closed Loop Analysis
Call tracking can help you start to close those pesky little holes or ‘leaks’ in your marketing funnel.
But you still need to attach that phone call back to a specific person for your records (and to reconcile with any Point of Sale or invoicing systems).
The simplest method I’ve found is to set up a good marketing CRM and automation tool. Look: This isn’t a sales pitch. But it’s worth the cost. Even if you gotta splash out the ~$10k on something like HubSpot.
If you’re spending anything significant on advertising, and you don’t have this, you’ll make your money back quickly (I’ve seen it happen as fast as a month or two in some cases) by simply spotting:
Which campaigns you’re overspending (thereby cutting costs), and
How to increase campaigns that are performing well (thereby increasing revenue)
Win, win.
Then, you’ll be able to finally (aaatttt laaaaasssttt) line up real, paying customers against your marketing KPI’s.
If you’re on a tight budget (or just stubborn), you can get a fraction of this by using something like the excellent (and free) LeadIn.
This handy WordPress plugin is like a CRM-lite that pulls in all form submissions from common sources (like Gravity Forms) and begins assigning their session data.
The key piece of information to look for is the traffic source that generated new customer Betty Sue. Pull up the individual contact in LeadIn, and look at the lower right-hand corner.
If you’ve set-up UTM codes properly (like a good little marketer), you should see not only the Campaign that generated this new customer, but even down to the specific search term.
Is that expensive-looking keyphrase with a $50 CPC worth it? How about that campaign with a cost of $500 Cost per Conversion?
Sounds expensive on the face of it, but now you should know. With cold. Hard. Data.
The Holy Grail.
Phew. We did it.
Conclusion
Analytics aren’t meant for reporting.
Instead, the value comes from the analysis, drawing insight, and making day-to-day tweaks, changes, and updates based on those new findings.
The problem is, that there’s an abundance of potholes and potential minefields in waiting for you if not careful.
Obvious problems like incorrect conversion tracking can be easily solved by some software and some technical help. However, the common mistakes from there get a little more subtle and nuanced; becoming harder to be aware of your blind spots and how to address them.
A combination of focusing on business metrics (over leading indicators), and making sure you know as much about someone as possible (including which specific ad in what campaign generated this buyer who’s worth how much), can help.
It’s not always easy. But shoring up these issues is the only logical way to know exactly how to tweak your ad campaigns, targeting, and creative at the end of the day.