It’s probably not possible to work in the digital marketing or SEO sphere for more than 5 minutes without hearing about Rand Fishkin. He’s the founder and former CEO of Moz.com, whose tools, articles, and Whiteboard Fridays continue to provide a massive amount of value to agencies and consultants alike. He’s also the author of Lost and Founder, which is a must-read for anyone who wants to found their own company.
He recently launched a new company and a new product: SparkToro. SparkToro is an exciting new market research product that helps marketers discover who their audience actually is.
Learn all about it, and about market research in general, on today’s podcast. As a bonus, you’ll get some great thoughts about entrepreneurship as well.
Here’s the highlights:
- (1:00) SparkToro in a nutshell.
- (1:25) The challenges SparkToro is designed to fix.
- (3:49) Why market research is so important.
- (5:43) Some surprising ways that users are incorporating SparkToro into their workflow.
- (7:03) The difference between using SparkToro and going to an existing marketing research firm.
- (8:55) Recent changes in audience behavior.
- (11:35) Advice on audience checking and using the data to inform strategy.
- (14:11) Pitfalls to avoid while using SparkToro.
- (16:59) Sorting through marketing opportunities.
- (18:40) Insights into the decision making process in building the tool.
- (20:32) Venture capital vs. angel capital and bootstrapping.
- (23:21) Rand’s causes.
What is SparkToro, what does it do, and what problems does it solve?
SparkToro helps marketers understand who their audiences are and what they do. It gives users “behavioral data about what they pay attention to, who they follow and listen to, what publications they read, what podcasts they listen to, and what YouTube channels they subscribe to.”
Rand says marketers were really flying blind in the past.
“There were three ways people were solving this problem previously. #1, they were serving and interviewing their audiences. But self-reported data is terrible, suffers from all kinds of bias, and you don’t get a very complete picture.”
The second way people would try the problem was to try to Google around, or search on social networks to see the popular accounts and blogs. What ranked well, what had lots of followers?
“Unfortunately,” Rand says, “while that data is semi-useful, the problem is it doesn’t tell you whether your audience listens to or pays attention to it. It just tells you whether many people do or few people do. Or in the case of Google, whether it was good SEO or not.”
The third way they did it was to completely ignore the problem.
“In a lot of cases it was: we’ll just throw money at Google and Facebook and let them figure it out. That can work for you if you have a very high-margin business and a really big brand and you can play the ad game well. But for a ton of companies, this is a very difficult ROI situation.
You’re right at the margin of where you’re spending money on ads to seek growth, because so many people are seeking growth, vs. what you’re actually turning a profit on.”
Why is market research so important?
“The core of the problem,” says Rand, “is where should I go that actually reaches my audience?”
“Whatever kind of marketing you’re doing, right?
It could be organic outreach. It could be guest posting. It could be pitching on a podcast. Could be a co-branding relationship or sponsorship. Could be, hey, this company’s putting together this webinar, let’s see if we can get a guest spot there, maybe we can sponsor it, maybe we can amplify it.
Maybe I’m just reaching out to people on social to try to get them to link to my blog posts, right?”
The targeting is often very off-base.
“If we can just tell [marketers] you know, 18% of architects in Los Angeles, CA listen to this podcast, suddenly someone in an agency who is helping an interior design firm reach the architects they need to reach is going to be so much better at their job.
Everything they do is going to be so much clearer if you just know that number and source of influence.
And so that’s all SparkToro is. At the core of the product. There’s lots of other features, but the core of the product is you just type in an audience, a way people describe themselves, something they already follow and pay attention to, and then we tell you: here are the websites they visit in order of how much they are visited, shared, linked to, engaged with.
Here’s the social accounts. Here’s the podcasts. Here’s the YouTube channels. Here are the sources of influence.”
Were there any surprises in how customers began using the product?
“So they’ll see, oh, the audience I’m going after frequently uses these hashtags, follows these social accounts, and they take it right over to Twitter, LinkedIn, Facebook ads, and they just run ads to target those audiences.”Rand says he was surprised by this because they didn’t see it during the beta.
“It’s been a fascinating use-case and it’s been central to a few of our customers. About 10% of them have told us they use it in that way.”
What's the difference between using SparkToro and going to an existing marketing research firm?
Rand says you’ll get a lot more data from a marketing research firm, but it comes at a price.
“It was pretty wild to us that folks were willing to go and pay $50,000 for a market research report that would take six months to assemble a panel, and then they’d get the report back and and that big report would include data like: here’s how we think people describe themselves, and here’s how we break them into buckets.
But the statistical soundness of it is based entirely on the sample of people that you get, and the sampling could be all over the place.”
Rand says SparkToro was built with just one bias, which was:
“We only sample your online audience.
If your audience doesn’t have Twitter accounts, and LinkedIn accounts, and Facebook accounts, and Reddit accounts, and YouTube accounts, we have like ten networks that we look at, then we can’t get data about them. But everyone is so on, almost every audience is so online today that it works pretty well, right?
The numbers will be in the tens of thousands, hundreds of thousands. A market research firm will often use election-poll sizes of 1200 or 1300 people.”
Recent changes in audience behavior
“We’ve seen more significant change in what people follow, listen to, and pay attention to among the cohorts that we follow in the last three months than in the previous six-to-nine months.”He’s also seen changes in the words and phrases people use. It suggested a new, obvious feature that is not currently built into the product.
“You should be able to say: I really care about architects in Los Angeles, tell me about how their behavior changes through whatever’s going on in the world.”It’s something they’re planning on building. They’ve also seen a change in podcast behavior.
“There are more people listening to new podcasts, more new listeners, and more new podcasts. But overall listening time per listener is down because people don’t have their commutes.”
Advice on checking and using the data
“The big thing for me is identifying a source of influence as it’s on the rise rather than waiting until it gets hugely popular.”For example, he’d rather buy a year of sponsorship for a podcast that’s on the rise than one that is on top, when advertising rates will be ten or a hundred times higher.
“I think for marketers being able to see that a source of influence is growing, or growing rapidly for their audience, even if it’s only moving from oh, half a percentage of my audience was paying attention to this source last month, now 1 and ½ percent is. That’s interesting. Let’s watch it next month. If it jumps again, we should build out. Let’s build a relationship there.”Rand notes that once someone gets to that top echelon they’re much harder to build a relationship with.
“They’re very sensitive to people reaching out and spamming them. But during that rise they’re often much more targetable, much more reachable. You can build relationships, you can form a connection, and so that is, for me at least, a huge part of how and why that data is valuable.”
Pitfalls to avoid while using SparkToro
Rand notes SparkToro works best when it’s easy to describe your audience. Then there are clear patterns of behavior, clear words and phrases they use in their content when they talk about things.
“So it’s awesome for finding real estate agents. Because real estate agents call themselves real estate agents. They talk about real estate. It’s terrible for finding homeowners.
If your target market is: I want to reach homeowners, well, how do homeowners uniquely describe themselves from everyone else online?
They don’t put it in their bio. They don’t put it in their shares and content. Maybe there’s a few things they would talk about that’s different from people who rent, or people who are nomadic and travel, but eh, not so much.
So you can’t find: ‘Jewish-Americans in Seattle, Washington.’ Or data on ‘18 to 35-year-olds.’ “
Rand notes they intentionally do not show and in fact ignore personally identifiable information for privacy reasons and for legal reasons.
“But if your audience has something in their job title, if they have something they already follow, if you say, gosh, it’s very hard for me to describe this kind of health care worker in the logistic supply chain, oh but they follow this account on Facebook or Twitter, so I’ll plug that in is my source…that can be useful.”
Rand also points out that you have to figure out what to do with the data once you have it.
“Your job is to find potential things that you could do once you figure out where your audience pays attention.”
How decisions about product functionality got made
SparkToro isn’t prescriptive: it leaves “what to do with the info” to the marketer. Rand says that this was intentional.
“Early on in our research people with different job titles and roles had this same problem. They all had the problem of: I need this data.
And rather than building…for example, we could have built SparkToro to be extraordinarily focused on content marketing agencies and their outreach needs.
We could look for places that accept guest editorials and highlight those, or amplify other people’s content and highlight those, but we saw that content marketers were 5% or 10% of the people who had the problem.”
They wanted to solve the data problem for everyone rather than try to be very narrow.
“That is not often the advice you get in SaaS. We went a little against the grain there, and I guess we’ll see how it works over time. I’m keeping my fingers crossed that we can build a diverse group of potential customers.
I’d rather solve the problem for lots of people, versus only a few.”
Venture Capital vs. Angel Capital & Bootstrapping
Rand says SparkToro didn’t raise venture capital. They did raise angel money, and they do a profit-sharing distribution.
“We want to get to profitability with the business, and we’d like to grow, but not be constrained by venture mechanics.”
By venture mechanics, Rand means: “Either you become a billion dollar company, or you die trying.”
He notes he’s a big fan of the concept of “Zebras vs. Unicorns.” The idea is that there’s a lot of focus on “unicorns,” companies that will act as disruptors, and that will generate lots of money for venture capitalists. According to Entrepreneur, unicorns must spend a lot of money while racking up their own substantial losses in the first few years of operation, while continuing to parcel out equity so they can continue raising money.
Zebra companies, by contrast, do their business, are profitable, and work for causes as well, because they’re focused on alleviating social, environmental, and medical challenges. They aren’t attractive to venture capitalists, but they also maintain a great deal of control.
Rand also notes you can take a hybrid approach by raising venture capital late, like Basecamp did.
“If you get to $10 million in revenue and you’re on this crazy scale and curve and you can see the demand there, maybe it does make sense to go raise venture at that point.
Those are reasonable decisions by people who are already wealthy, already have plenty of control and ownership over their business, and they’re not going to be giving away a lot of that at that point.”
That way, you’re not setting yourself up for failure by making anything other than a billion-dollar outcome into the “win” scenario.
What’s your right now cause?
SparkToro did a donation-matching program. They donated a dollar for everyone who tried the tool for free, and the chosen charity was “Give Directly.” They traditionally operate primarily outside the US, but since the Covid-19 economic crisis, they opened a program here in the US that gives money to those most negatively affected.
“People who are food insecure, people who are economically insecure, people right on the edge of homelessness.”
Rand says he loves their philosophy.
“Giving dollars to people instead of giving them other kinds of things is really productive and extraordinarily efficient, so I highly recommend them.”
SparkToro put its money where Rand’s mouth is. A few weeks ago they donated $10,000.