By It’s January, and that means many marketers are tasked with a project to identify new channels and sources of influence for new potential investments. This year is likely more active than most given the decline of Google search traffic, the rise of AI, the Twitter exodus, and the meteoric growth of Threads and Bluesky.
But, far too many times, we see marketers using tactics to answer this question that fail miserably. Worse, the failure isn’t obvious, so companies will double down on these methodologies for years, foolishly missing opportunities and overspending in channels that might not be ideal. This week’s 5-Minute Whiteboard is here to help.
Transcript
Howdy, SparkToro fans. Welcome to another edition of 5-Minute Whiteboard. This week, we’re kicking off 2025. The new year means that a lot of you are almost certainly gonna be tasked with a new project, especially if you’re in the marketing world or you’re a founder/entrepreneur and you’re trying to figure out what are your customers sources of influence. New year, new marketing campaigns, new channels, new places that you might go to try and attract customers, which is a good time to do it.
Here’s the big problem. This is the biggest problem that I see, which is most folks default to audience surveys and interviews, which is not a bad thing. Talking to your customers is excellent. I absolutely recommend it. Can’t recommend it enough. But… what happens is when you run these surveys, you’ll ask questions, you know, maybe from a group of people who are on your email list or, perhaps you’ll try and go out to your social channels and get, folks to take the survey that way. And you’ll ask things like: Which social networks do you use, and who do you follow on those? What publications do you read? What websites do you visit? How much are you using ChatGPT or Gemini or, you know, other LLMs?
Fine.
I want to be really clear that I don’t think it’s a terrible thing to ask these questions but you should be prepared to be very skeptical of the answers because when people answer these they are rarely accurate. Their actual behavior and their perception of their own behavior are disconnected.
I’m not saying that they’re lying to you, I’m saying that they themselves don’t know. Asking these types of questions is not gonna get you accurate answers or comprehensive answers.
Another thing people do is they ask this question after a purchase. So someone’s reached a checkout phase, you know, they’ve bought something for your website for example, and then they take a group of those folks and they say, how did you find us? How did you hear about us? This isn’t the worst thing ever, but I’ll share with you a little anecdote which is, we tested this with a company, a consumer product company, and they ran a post purchase survey. And what they did was they, you know, made some of the answers at the top, some at the bottom. The ones at the top always got the most. So, okay, you can randomize them and then control for the fact that the top always gets most of the, responses.
But they also did something incredibly clever which was to put in an answer in this checkout survey that said, I heard about you from this particular, I think it was a YouTuber in this case, who had never talked about them. They had never been mentioned on that channel. Guess what? Lots of people, because that audience was very familiar with that youtuber, said that’s where they heard about them.
I think it was like fifteen or twenty percent of the respondents gave the fake answer. Right? Said, yeah. That’s where I heard about you.
I heard about you from this person who’s never actually talked about you.
So you’re you’re getting bad answers, but that is not the biggest mistake. The biggest mistake, the biggest problem is that who are these people? Who are all these people that that we’re talking about here, right, that you’ve run audience surveys and interviews with, through your social channels, through your email list, through post purchase surveys. Who are they?
These are people you’ve already captured. These are people who already know about you.
So there’s a universe out there, right? This much bigger universe of people, and you are missing every single one of these folks. You cannot possibly capture them when you survey the people who have already purchased from you, the people who have already been in your in your email list or via your social channels. You’re just gonna miss out on these. And this is why I cannot recommend enough going out and getting broad data about, you know, groups of people from the Internet. You can do this manually, you could, you know, go crawl through LinkedIn or, other social networks, you could take a look at what people follow and share, you You could go to conferences and events and listen to what people talk about.
But generally speaking, what you really need is some sort of audience research at scale. This is obviously something we do at SparkToro. There’s a problem we hoped to solve, right, when we when we built the company.
But I just want to prepare folks for the idea that that this is is a bad way to prioritize your marketing, choose channels, choose sources of influence, and make investments because you’re going to get a lot of inaccuracy.