This year Clay invested over $1MM in billboards. That kind of marketing spend without clear ROI usually makes finance people nervous.
But we make a ton of big and small marketing bets like this at Clay, to the point we wanted to share some of our decision making process. Our co-founder Varun sat down with our head of finance Karan Parekh to talk about what guides our marketing motion without demanding perfect attribution and killing our creativity.
Some of what we cover:
How brand and performance marketing are linked and why B2B companies invest in brand too late
How we use attribution modeling, customer-defined attribution, and incremental analysis to measure the unmeasurable
What marketing works as you move from Series A to scale-up to public company (and why Figma nailed it)
Guerrilla marketing tactics from Clay’s early pre-PMF days
At the end of the video, Varun and Karan share the strategy behind Clay’s recent billboard campaign and real examples of how we’re measuring impact and ROI.
Brand and performance go hand-in-hand
Most finance leaders think about marketing in two independent buckets: brand (the fluffy stuff you can’t measure) and performance (the measurable stuff that drives revenue). We think that’s the wrong way of looking at things.
Brand and performance are linked. Brand enables performance. If nobody knows who you are, you can buy all the Google Ads in the world and you won’t get conversion.
Think of it this way: you don’t invest in brand to get direct ROI from the brand investment itself. You invest in brand so that when you put $1 into performance marketing, you get more than $1 out.
Most B2B companies miss out on this, though. You need to invest in brand early. Why? Because B2B brands suck. No one invests in them. So if you do, you’re already different. And half of marketing is just being different.
Consider Figma. They built an entire community and ecosystem around their product to the point that designers have an identity around using Figma. That’s not performance marketing with direct attribution, but that community is now so embedded that they’re almost impossible to disrupt. When Figma runs performance campaigns, they’re building on a foundation that makes every dollar work harder.
The real question isn’t “what’s the ROI of this billboard?” It’s “how much should I invest in brand such that when I invest in performance, I get a higher net ROI?”
What to do at each stage
Series A / early stage: focus on inputs, not attribution
If you’re pre-product-market-fit or just finding your footing, your bias will be to measure everything. Resist that urge. Your dollars are too small to run meaningful incremental tests. Your data is too noisy to trust attribution models. What you actually need to do is create moments that people remember—a splash that people can rally around.
At this stage, it’s not about “did I get one customer for this $1 spent?” It’s about “did I create something memorable that starts spinning the flywheel?”
For Clay in the early days, this looked like:
Taking customers to Dreamforce dressed as “Clay-gents” in full suits
Producing absurd videos like “V for Viglietta” showing our team building Clay tables while skydiving, on the ocean, or rappelling off mountains
Doing whatever felt creative and different, even if we couldn’t measure it perfectly
These things rely on creativity, ingenuity, and scrappiness. What they don’t rely on are big budgets. Any company can do them. They weren’t measured by pipeline dollars per creative asset. They were measured by whether people started talking about us.
Growth stage: get sophisticated about measurement
As you scale, you can and should get more rigorous. There are two lenses to apply:
Customer journey analysis: For any given customer, are you doing the right things at each moment to drive conversion? If a customer got in touch with you on LinkedIn, came to an event, attended your conference, and then was reached out to by sales at another conference—all of those stack up to a total cost relative to the lifetime value of that customer. You want those two sides of the equation in balance.
Incremental testing: Once you have a baseline, you can start testing, e.g. if you run a billboard campaign in SF but not in New York, does SF see a 10% lift in traffic? This is where you start optimizing the engine you’ve built.
The three-method approach to marketing measurement
There are three ways to understand what’s working:
Attribution modeling Multi-touch, last-touch, first-touch—whatever model you choose, you’re trying to understand which touchpoints contributed to a conversion. This is helpful but imperfect.
Customer-defined attribution Just ask customers: “How did you hear about us?” You’d be surprised how often they remember a billboard, a video, or a specific event. This qualitative data is gold.
Incremental analysis Test a channel in one market but not another. Measure the lift. This is the most rigorous approach, but requires scale to do well.
You need all three working together to build a complete picture. Don’t rely on just one.
A real example: how—and why—we’re spending $1M+ on billboards
We recently made a big investment in out-of-home advertising by spending seven-figures billboards, ads on the New York City subway, and other OOH surfaces. Here’s a little bit about how we’re thinking about it:
Know where your customers are
Spray and pray with this sort of advertising is exactly what you don’t want to do. You should be using your your own data to map exactly where your customers work. (Yes, we use Clay for this and you should too.) For something like a billboard, you should be buying space near their offices, along their commute routes, and the places they actually go.
When you go in person to verify placements, you’ll find that some billboard spots don’t even exist. Some face bathrooms. Some are behind doors. Some are in the wrong airport terminal. The lesson here is to do the homework yourself, don’t rely on someone else to do it for you.
Pro-tip: for airports, which are an extremely valuable part of the OOH landscape, you should know which terminals your customers use. Are they mostly international or domestic travelers? Do they have relationships with specific airlines or favor certain airports in a given region? Figuring out whether you’re better off spending money at LaGuardia vs. JFK or Heathrow vs. Gatwick is vital to spending this money smartly.
Don’t play it safe
Any OOH campaign has to be different. You need great creative talent, but you also need to do something no one else is doing.
This year, we filmed the entire process of installing our billboards and turned a physical representation of Clay into highly distributed digital content that reached a national audience. We got twice the value: the billboard itself and the content about making the billboard. We’ll be doing a deep dive on the actual process of creating and launching a billboard campaign with our schemer-in-residence Jessica Jin who came up with that idea.
Measure what you can
People say “you can’t measure a billboard.” That’s not true. You can’t measure it perfectly, but you can do meaningful attribution:
Control markets: Did the SF market perform better than NYC where we didn’t run ads?
Self-reported attribution: Are customers mentioning the billboard in conversations?
Geographic lift: Are we seeing more product-led growth signups from areas around the billboards? More deals created in those regions?
Traffic analysis: Can we see a lift in new customers coming from the area around the billboard?
Perfect measurement isn’t the goal here. You should be focusing on gathering enough signal to make better decisions next time, just like you would with any process.
The bottom line
Your job as a marketer isn’t to measure everything perfectly. It’s to understand the relationship between brand investment and performance returns, and to optimize that relationship over time.
Focus on inputs and creativity early one, and try to create moments worth remembering. As you scale you can start layering in attribution, self-reported data, and incremental testing to understand what’s working. You should always be investing in brand, because that’s what makes your performance dollars work harder.
And most importantly: be different. B2B marketing is boring because everyone is doing the same thing. Half the battle is just standing out.
