5 Content Marketing Measurement Don’ts (and What To Do Instead)

Imagine a world where you could prove content marketing’s long-term value in a way the CFO would understand, accept, and believe.

Avinash Kaushik is working to make it possible.

The two-time bestselling author (Web Analytics 2.0 and Web Analytics: An Hour A Day) understands the content world. As chief strategy officer at marketing agency Croud, he gets the marketing angle. And given his 16-year stint at Google, where he was part of the Google Analytics launch team, Avinash understands the data side, too.

Today, the triple-threat expert helps executive teams, marketers, and data analysts use digital strategies and emerging technologies to outsmart their competitors. You can learn from Avinash in person at Content Marketing World in September.

Recently, he joined CMI’s Ask the Community livestream, where he shared five don’ts (and their corresponding do’s) to improve your content marketing measurement today. You can watch it or read on for the highlights:

These words come from Avinash with light editing and condensing. The headings come from me.

1. Don’t measure content performance against inappropriate goals

Marketing lives up to its fullest glory when you can figure out the short-, medium- and long-term things you do that drive value.

How do you put all of these things together? It requires bringing together the art and the science.

You want to do the kind of marketing that allows you to meet quarterly revenue and profit numbers. But at the same time, you want to build this expansive relationship with consumers who might consider buying in the future – or with people who may never buy your products and services but who influence a much larger pool of customers.

What is difficult is all related to figuring out, such as, “If I run a bunch of advertisements on TikTok, should I think about whether they’re driving revenue now? Or should I consider them an extension of my brand that would allow us to create value for the company over a longer period?”

The thing that goes wrong in our space is perfectly captured by one of my favorite metaphors: Never judge a fish by its ability to climb a tree.

We do that all the time. And that’s what creates suckiness in our life. Because we’ll say, “If TikTok isn’t producing revenue, it stinks.” Or “If paid search is only driving revenue but not extending the number of new customers, then it stinks.” Both these questions involve judging a fish by its ability to climb a tree. So spend time figuring out what kind of fish and what it does best, and then judge its ability to swim.

2. Don’t track too many KPIs

Don’t think of data pukes as a solution to the problem. Most tools out there just puke lots and lots of data.

I’m a big fan of an approach I call the “digital marketing and measurement model.” It’s a simple framework that asks: What is the purpose of the marketing you’re doing? Then, “If this is the purpose, then we should focus on this kind of data. And that means we should use these KPIs.”

I recommend (whether you do owned, earned, or paid marketing) that you coalesce around two KPIs: one efficiency KPI and one effectiveness KPI.

For example, if you’re doing paid marketing, the effectiveness KPI is usually revenue or profit, and the efficiency KPI is the cost per order. Between these two KPIs, you can find and focus your attention. You can have other metrics underneath that but only two KPIs.

Now, let’s say you’re producing lots of content on YouTube. For your YouTube content, the number of net new subscribers per video is the effectiveness KPI because it shows you were able to get people to pay attention. On the efficiency side, you can measure reshares because if someone reshares it, you go from the first-level network to the second-level to a third-level network, and you expand your audience.

3. Don’t waste time on useless data

Analytics used to be a world where having more data meant you were smarter. That was 20 years ago. Now we have more data than God wants anybody to have. Being smart is all about figuring out what data to ignore.

I think we should form strong opinions. I hate the metric impressions. It’s useless. It’s not worth even a penny. If you report impressions, I’m going to get mad at you.

But you have to understand the landscape enough to say I’m going to ignore, ignore, ignore this data because it doesn’t have enough value. And that’s what makes your approach to data smarter.

4. Don’t prioritize psychographics and demographics over intent

For a long time, marketers didn’t have enough data. So they said, “OK, we’re going to think about this as a funnel – and our job is just to shove people down the fricking funnel.”

The problem is none of us behaves in a way that follows the traditional funnel.

But at the same time, we need signals. For example, a marketer in the past might look at Amanda and think, “She is 22 years old, lives in the Midwest, and has a very nice home, so let’s sell her, blah, blah, blah.”

The reality is your demographics and psychographics reveal very little about what you’re thinking, what kind of person you are, what your values are, and all of those things. So you get idiotic and irrelevant advertising because all the marketers know about you is that you’re 22, live in the Midwest, and have a very nice home. And now, out of a hundred things they sent, maybe one would be relevant to you.

But marketers don’t have to do that anymore because we can discern intent through a consumer’s behavior. The simplest example is that you type a query into Bing about a new hybrid car. You’re expressing intent, and Bing will use that to deliver the right advertising to you.

Or, if someone follows certain brands on Facebook or writes about a certain thing, we can discern intent from that. That’s a much better way to deliver advertising or marketing to you, whether a paid ad or a piece of content.

5. Don’t fear AI in analytics

I talk a lot about data – what you should ignore and what you should pay attention to. The machine-learning solutions built into analytics tools now let you avoid hunting and pecking into the data to figure out what you should look for. You just get a report that shows things you should pay attention to.

If you log into tools like Google Analytics, for example, or many other analytics tools on the market, there’s usually a report called “intelligence” that gives you this insight faster. You don’t have to pour through data to figure out what’s important. It finds hidden things inside your data and surfaces them.

Another example is intent. It’s hard to figure out how to infer someone’s intent in a sea of data. And algorithms are so fantastic at analyzing data at scale automatically to help you find the known unknowns and the unknown unknowns.

So every paid ad or piece of content someone sees might be relevant to them. AI solutions now help us figure out how to do one-to-one marketing in a way that was unfathomable a few years ago.

I’m very excited about AI’s potential to help companies balance brand and performance advertising. How much money should we allocate to things that drive revenue right now versus brand (development)? And how do we measure brand with more than touchy-feely metrics like unaided awareness, consideration, intent, or (please don’t use this KPI) brand love?

The most bleeding edge use of machine learning right now is to figure out how to understand the impact of brand advertising. How do all the emails, television commercials, stories in catalogs, and so on work together to identify marketing’s incrementality?

For our clients, we can go to the CFO and say marketing drove 32% of all sales incrementally, meaning if you hadn’t given the team the budget to do their marketing, (the brand) would not have gotten these sales. I call this the God KPI for the CFO.

I’m using machine learning to identify marketing’s incrementality and then say, “This is the long-term impact of email marketing that has nothing to do with sales. Or this is the long-term impact of content marketing.”  

At the moment, it’s tough to justify content marketing over the long term. But by using machine learning, you can. Machine learning is making us smarter about being able to find the data and insights we can activate and to do incredible imaginative marketing that wasn’t possible in the past.

And we may be able to go to the CFO and say, “Here’s the God metric. Now gimme another $20 million.”

HANDPICKED RELATED CONTENT:

• Overwhelmed by Marketing Analytics? This Advice Will Help Create a Workable Strategy
• How To Use Your Website Data To See If Your Buyer’s Journey Really Works [New Template]
• Why Learning GA4 Is So Hard – And What To Do About It
• Why You Struggle To Prove Content ROI – and How To Settle Up (or Down)

Cover image by Joseph Kalinowski/Content Marketing Institute