We are writing the benchmarks as we go. Nobody has the answer yet.

I studied finance and was looking for a job in investment banking or commercial banking. The job I got was a small ad-tech start-up. I walked in to ask about an account-manager role, and the CEO asked if I knew anything about advertising. I said no. He told me to study everything I could over the weekend and come back on Monday. I came in on Monday and at the end of the interview he told me to bring my laptop the next day and start. The transferable benefit of a quantitative background is that ad tech is all about data, and a CEO's job is broader than marketing alone, so understanding how the business fits together comes back to the original training.

I worked in this industry before building StackAdapt and firsthand experienced the pain of not having the platform we ended up building. So the product was born from personal frustration and, in many ways, is still a work in progress ten years later. Our vision for where we need to be has only accelerated over the last couple of years.

The B2B space is changing rapidly. I reflect back on the start of COVID and how dramatically B2B marketing changed once conferences disappeared and the budget moved to digital demand generation. Right now the marketer's job is harder, because economic uncertainty remains and both consumer and B2B buyers are still holding back on spending. It is, arguably, what the job was before COVID. 2021 gave us a temporary frame of reference, and we mistook it for the new normal. The truth is we're back to how difficult marketing was in 2016 or so.

The space is so crowded and the barriers to entry are so low that the demand on marketers to stand out is much higher. The opportunity is to think more like a B2C marketer than a B2B one. Take risks. Build something people will enjoy interacting with. I saw an ad from ZoomInfo recently, a Western-themed piece called Finding Gold. ZoomInfo is, for all intents and purposes, a fairly dry B2B brand, but the creative was clever and genuinely fun to watch. The more progressive B2B brands are leaning into the playbook B2C has built up over decades: personalisation, creative optimisation (DCO), multi-channel reach across digital out-of-home and connected TV. The most progressive ones are seeing real results.

Every company is now trying to see where AI fits into their business. The overwhelming majority of generative-AI applications I've seen mapped in B2B sit in creative or content generation. It's an obvious application, plugging into the model and producing text, but it's a small fraction of what the technology will do. Every business is still in the early stages of figuring out how to create real value for customers. At StackAdapt we are using generative AI across many platform capabilities, but only a small fraction has been introduced to customers. The rest is still in experimentation or development. Until we put it in front of users, we won't really know how much value it adds.

For a platform that serves many different customer types and many different problems, AI is not going to be a flip-of-the-switch transformation. It is going to be incremental improvements: better quality of life, slightly better retention, slightly better efficiency. For companies solving a single narrow problem, the impact may look more dramatic. Enterprise software rarely jumps from zero to one.

As soon as you pause on marketing, everything becomes harder. The thing we have learned is that you should never go dark. Always stay top of mind. In times of internal or external uncertainty there can be a knee-jerk reaction to say we're not sure, so let's be quiet. In most cases you should do the opposite and lead with a message of some kind. Staying dark is the thing not to do.

We have a team of about fifty people in marketing doing a great many things. The way I think about great marketing, and what I've seen work for us, is anchoring on a specific customer problem and a concrete solution. When we have been able to articulate the problem precisely, the response from the customer has been they really understand my pain point, and attention follows. The evolution for us has been from saying we're a programmatic advertising platform, marketers across any industry can work with us, which is hard for a buyer to attach to their own pain, to verticalising and asking what marketers in B2B specifically struggle with, what marketers in retail or e-commerce or healthcare struggle with. We now have separate engineering and solutions teams building products against each of those vertical pains. It's still one product underneath, but the way each customer uses it speaks to a specific pain.

At a demand-generation level we measure cost per acquisition and have account executives score the leads coming in so we can add a qualification layer on the channels they came from. At a brand-advertising level we want to have fun and produce something people will notice. Two years ago we made an ad with an agency drawing parallels between business and competitive sport. The agency proposed using a toaster as a symbol of StackAdapt. At first we said a toaster, this is random. We went with it. When it went live people asked why a toaster? and we'd say we didn't really know, we just thought it was fun.

The follow-up we produced in-house. We 3D-printed a physical toaster. Even though there were only two ads, people started asking why is StackAdapt always obsessed with toasters? It wasn't an obsession. We had only done two of them. But if it gets people talking, that is success.

You have to experiment. The discipline is understanding what could realistically backfire badly and being cautious about those things. For the rest, just see what happens. Understand the consequence between different decisions. What is genuinely hard to undo? What is a small spend on an experiment we can afford? Experiments don't need to be large. If something feels right, follow the gut. Not everything can be supported by data, and sometimes the only way to generate data is to run the experiment in the first place. Most of what we try doesn't carry enough downside risk to overthink it.

Historically StackAdapt was focused on distribution. Clients arrived with creative and we ran it across the internet. We realised many clients struggle with making good creative; they lack the resources or the know-how. We started with a small creative services team to help our clients build better ads. It grew substantially. Generative AI then forced the question: how do we scale this beyond a hands-on services team to the wider client base? We built a destination where clients can build ads themselves, with generative AI as the playground. Going down that path gave us an end-to-end view across building, running, and measuring creative, which then unlocks insight across the entire customer chain, from where a user is, what they're seeing, how they interact with the ad, the site, and the purchase, and the value of that purchase. It's early, but the path is clear.

Fairly standard in many ways. Digital marketing, event marketing, marketing operations, the latter is one of the keys to our success because the team can really dig into the data and follow how leads flow through the system. One area that is not common is the investment in design. We have always valued great design from day one. It started with product design, because we wanted to build a product people enjoy using, and it extended naturally into marketing. It helped raise the profile of the company and made us look bigger than we were. We have also regionalised teams across Europe and Asia-Pacific so we have local coverage. Beyond that, work structure isn't an area where we need to innovate. There's higher-impact low-hanging fruit elsewhere.

We try to measure at different fidelities. At the executive level we look at revenue targets, growth rate, bottom line. We spot trends and zoom into specific areas. How are new customer cohorts expanding? How are they retaining? If a cohort retained worse, why? Have the lead sources changed? We can then ask marketing whether something changed in the source of leads last quarter. It's an ongoing conversation across the company. The revenue operations team and FP&A team are constantly trying to understand the business better. I'd be lying if I said we know our business 100 percent. As much as we invest in analytics and data warehouses, there's so much data. Knowing the right questions to ask is a challenge of its own.

It would be fun to see more creativity from B2B brands, breaking the dryness, making the work more human. COVID brought it very forward: the decision-makers in those buying companies are people like everyone else, with lives and hobbies, consuming the same content. The discipline is figuring out how to reach those people through consumer channels. Targeting is evolving from organisation-level to one-to-one user identity, so you can reach the same user across whether they're at work or at home equally effectively. And we'll see more B2B companies move into emerging channels: connected TV, streaming audio, digital out-of-home. Programmatic has lowered the barrier enough that this gets democratised for smaller brands.