Imagine a world where you could prove the long-term value of content marketing in a way the CFO would understand, accept, and believe.
Avinash Kaushik is working to make this possible.
The two-time bestselling author (Web Analytics 2.0 and Web Analytics: An Hour A Day) understands the world of content. As Chief Strategy Officer of the marketing agency Croud, he understands the marketing perspective. And given his 16-year tenure 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 new technologies to outsmart their competitors. You can learn in person from Avinash at Content Marketing World in September.
He recently took part in CMI’s Ask the Community livestream, where today he shared five don’ts (and corresponding do’s) to improve your content marketing measurement. You can watch it or read on for the highlights:
These words are from Avinash with slight editing and condensation. The headings are mine.
1. Don’t measure content performance against inappropriate goals
Marketing is at its most effective when you can figure out what you’re doing to add value in the short, medium, and long term.
How do you bring all of these things together? It requires the merging of art and science.
You want to do the kind of marketing that will enable you to achieve quarterly sales and profit numbers. But at the same time, you want to build that broader relationship with consumers who might consider making a purchase in the future — or with people who might never buy your products and services but influence a much larger audience.
What’s difficult has to do with figuring things out, like, “If I run a lot of ads on TikTok, should I be thinking about them now increasing sales?” Or should I think of them as an extension of my brand that is would allow us to create value for the company over a longer period of time?”
What goes wrong in our space is summed up perfectly 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 makes our life shit. Because we’re going to say, “If TikTok doesn’t generate revenue, it stinks.” Or, “If paid search only increases sales but doesn’t increase the number of new customers, then that’s terrible.” Both questions are about catching a fish judged by his ability to climb a tree. So take the time to find out what type of fish he is best at and then assess his swimming ability.
You don’t judge a fish by its ability to climb a tree. But that’s exactly what a lot of marketers do with their #content analytics, says @avinash of @KMoutso’s @CMIContent. Click to tweet
2. Don’t track too many KPIs
Don’t think of data vomit as the solution to the problem. Most tools out there just throw up tons 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 are doing? Then, “If that’s the purpose, then we should be focusing on that kind of data.” And that means we should be leveraging those KPIs.”
I recommend (whether you’re doing owned, earned, or paid marketing) that you focus on two KPIs: an efficiency KPI and an effectiveness KPI.
For example, if you do paid marketing, the effectiveness KPI is usually revenue or profit and the efficiency KPI is cost per order. Between these two KPIs you can find and focus your attention. Below that you can have other metrics, but only two KPIs.
Use two KPIs – one for efficiency and one for effectiveness – for your #ContentMarketing, says @avinash of @KMoutso’s @CMIContent. Click to tweet
Let’s say you produce a lot of content on YouTube. For your YouTube content, the number of net new subscribers per video is the KPI for effectiveness as it shows that you were able to grab people’s attention. In terms of efficiency, you can measure reshares because when someone reshares it, you go from the first tier network to the second tier network to a third tier network and expand your audience.
3. Don’t waste time with useless data
Analytics used to be a world where you were smarter when you had more data. That was 20 years ago. Now we have more data than God wants anyone to have. If you want to be smart, you need to figure out which dates to ignore.
I think we should form a strong opinion. I hate the metric impressions. It’s useless. It’s not even worth a penny. If you report impressions, I get mad at you.
But you have to understand the situation so well that you can say I’ll ignore, ignore, ignore this data because it doesn’t have enough value. And that makes your data handling smarter.
4. Do not prioritize psychographic and demographic characteristics over intent
For a long time, marketers did not have enough data. So they said, “Okay, we’re going to think of this as a funnel — and our job is just to push people through that damn funnel.”
The problem is that none of us behave in a way that follows the traditional funnel.
But at the same time we need signals. For example, in the past, a marketer might look at Amanda and think, “She’s 22 years old, lives in the Midwest, and has a very nice home, so let’s sell her, blah blah blah.”
The reality is that your demographics and psychographics tell very little about what you think, what kind of person you are, what your values are, and all of that stuff. So you get idiotic and irrelevant ads because all the marketers know is that you are 22 years old, live in the Midwest and have a very nice home. And out of the hundred things they sent, maybe one would be relevant to you.
But marketers don’t have to do that anymore because we can tell intent from a consumer’s behavior. The simplest example is that you enter a request for a new hybrid car in Bing. You state your intent, and Bing will use that to serve you the right ad.
Marketers can use intent data rather than demographic and psychographic data to assess a consumer’s behavior, says @avinash via @KMoutso’s @CMIContent. Click to tweet
Or if someone follows certain brands on Facebook or writes about a certain thing, we can recognize an intention from that. It’s a much better way of delivering advertising or marketing to you, whether it’s a paid ad or piece of content.
5. Don’t be afraid of AI in analytics
I talk a lot about data – what to ignore and what to pay attention to. With machine learning solutions built into analytics tools, you no longer have to dig through the data to figure out what to look for. All you get is a report highlighting things to watch out for.
For example, when you log into tools like Google Analytics or many other analytics tools on the market, there’s usually a report called “Intelligence” that gives you those insights faster. You don’t have to sift through data to figure out what’s important. It finds hidden things in your data and brings them to the surface.
Another example is intention. It’s hard to figure out how to infer a person’s intent from a deluge of data. And algorithms are great at automatically analyzing data at scale to help you find the known unknowns and the unknown unknowns.
Therefore, any paid ad or content someone sees could be relevant to them. With the help of AI solutions, we can now figure out how to do one-to-one marketing in ways that were unthinkable just a few years ago.
I’m very excited about the potential of AI to help companies balance brand and performance advertising. How much money should we currently be investing in things that increase sales versus brand development? And how do we measure the brand on more than just tricky metrics like unaided awareness, consideration, intent, or (please don’t use this KPI) brand love?
The most cutting-edge use of machine learning today is figuring out how to understand the impact of brand advertising. How do all emails, TV commercials, catalog items, etc. work together to determine marketing incrementality?
For our clients, we can go to the CFO and say that marketing incrementally generated 32% of all sales. That means if you hadn’t given the team the budget for their marketing, (the brand) wouldn’t have made those sales. I call this the God KPI for the CFO.
I use machine learning to determine marketing incrementality, and then I say, “That’s the long-term impact of email marketing, which has nothing to do with sales.” Or, that’s the long-term impact of content marketing .”
Currently, it is difficult to justify content marketing in the long term. But through the use of machine learning, it is possible. Machine learning is making us smarter when it comes to finding the data and insights we can activate and to do incredibly imaginative marketing that wasn’t possible in the past.
And maybe we can go to the CFO and say, “Here’s the God metric.” Now give me another $20 million.”
HANDSELECTED RELATED CONTENT:
Cover photo by Joseph Kalinowski/Content Marketing Institute
