Q&A: The $130B AI debt and how to fix AI’s biggest revenue leak

While Meta is being called out for its exorbitant AI spend used to power its advertising engine ($130 billion), there is a glaring disconnect across the rest of the marketing stack.  Brands are certainly leveraging AI to generate demand, but still failing to capture revenue. If tech leaders are pouring billions into AI efficiency, why aren’t they using it to fix the most basic leak in the marketing engine? For example, new data from Typeform exposes this massive execution gap across the tech stack. For instance, 70% of teams still leave warm leads waiting 24 hours or more, and waste 6 or more hours a week managing them manually anyway. 

Furthermore, over half of marketers say leads often get routed to the wrong place, and 43% blame marketing/sales misalignment for delayed responses and around 28% of marketing teams use AI to trigger workflows for lead routing and follow-up, and just 24% use it to synchronise info into their customer relationship management.

Malinda Sandman (VP & Head of Global Marketing at Typeform) works closely with marketing and sales teams to realign responsibilities and automate speed-to-lead. She tells Digital Journal about tangible steps to help revenue leaders stop treating AI as a buzzword and instead to deploy workflow automation, and deliver verifiable ROI.

Digital Journal: Your research suggests that many companies are investing heavily in AI-driven demand generation, yet warm leads are still waiting more than 24 hours for a response. Why do you think organizations are prioritizing customer acquisition over customer conversion?

Malinda Sandman: Most organizations aren’t intentionally choosing acquisition over conversion, but acquisition has historically just been much easier to see and measure. You have a campaign, a lead target and a dashboard telling you how you performed, but once an actual person raises their hand, things tend to get messier. That lead might move between marketing and sales, through several different tools and across multiple owners before anyone actually responds.

That’s what stood out to me in our research. 97% of marketing teams are already using AI for at least one part of inbound lead follow-up, but 69% of organizations still leave at least a quarter of their leads waiting more than 24 hours for a response. Marketers aren’t stalling on AI adoption, but the bigger opportunity is connecting all of those individual pieces so the information a customer gives you can translate into action quickly.

We’ve spent a lot of time thinking about how AI can help us generate more, but marketers need to spend just as much time making sure we’re doing a good job with the demand we’ve already earned.

DJ: More than half of marketers report that leads are frequently routed to the wrong person or team. Is this fundamentally a technology problem, a process problem, or a human accountability problem?

Sandman: I think it’s a combination, but the process has to come first. 51% of marketers say leads frequently get routed to the wrong person or team, and 43% point to marketing and sales misalignment as a reason responses get delayed. It’s important to first establish clear accountability throughout the process, then invest in technology to automate it.

Teams need to understand where responsibility is unclear and where information or leads are getting held up in the workflow. As a marketing leader, I want my team to know what makes a lead a priority and when ownership moves to someone else. There also needs to be a clear plan for what happens when nobody follows up.

My advice to marketing teams is to get the foundation right, then use AI to make the process more efficient. Once ownership and next steps are clear, AI can help move information through the workflow without requiring someone to manually push it along.

DJ: AI is often marketed as a productivity tool, but your findings suggest many organizations are not applying it to workflow automation. What are the biggest misconceptions businesses have about where AI can deliver the fastest ROI?

Sandman: One of the biggest misconceptions I see is that the most valuable AI use case also has to be the most sophisticated one. 

There’s obviously a lot of excitement around automation, specifically around some of the more advanced things AI can do. I’m excited about those things too, but sometimes the biggest opportunity is much less glamorous. It’s making sure information gets where it needs to go or that the next step happens without someone manually pushing it along.

Marketers are already using AI across a lot of individual lead-generation tasks, but only 28% use it to trigger workflows for lead routing and follow-up, and just 24% use it to sync information into their CRM. We’re getting really good at making individual tasks faster without always looking at what happens before and after them.

That’s where I would start if I were looking for ROI. Find the places where work stops and waits for somebody to do something manually. Those may not be the most exciting AI use cases, but removing that friction can have a very real impact on the business and the customer.

DJ: Marketing and sales alignment has been a challenge for decades. In what ways can AI genuinely improve collaboration between the two functions, and where is human leadership still essential?

Sandman: AI can make sure what one team learns about a customer doesn’t get lost before it reaches the next, which I think is one of the most important outcomes of cross-team collaboration. I never want a customer to feel like they’ve taken the time to tell us something about themselves and we haven’t listened. That could mean never hearing from us again, or worse, receiving a message that completely ignores what they’ve already told us about how they want to engage.

We found that 78% of marketers believe they already collect enough information to personalize a response to every inbound lead, but 65% still struggle to actually do something with it. The issue a lot of marketers are running into is getting that customer context to the right person at the right time so they can take action quickly and turn interest into revenue. 

AI can help connect those pieces by moving relevant context between systems so sales isn’t starting from scratch every time a lead reaches them. If somebody has already told us what they’re interested in, we shouldn’t make the next person reconstruct that story.

However, leadership still has to define how the teams work together. There needs to be agreement on what a qualified lead means for the business and when ownership moves from one team to another. Once everyone understands what a good handoff looks like, AI can make that handoff much more seamless.

DJ: Speed-to-lead has become a major competitive differentiator in digital business. What impact does a 24-hour response delay have on revenue opportunities, customer experience, and long-term brand growth?

Sandman: Speed-to-lead is part of the customer experience, not just a sales metric. In our research, 50% of marketers said delayed responses contribute to lost revenue and 49% said they lead to lower conversion rates. There’s a very clear business cost to waiting.

Customer intent is a lot more perishable than we’re comfortable admitting. When somebody fills out a form or asks for a demo, you have their attention at that moment. Twenty-four hours later, a lot can change. They may have already heard from a competitor or simply lost some of the urgency that made them reach out in the first place.

But there’s also a brand impact that’s harder to put on a dashboard. We see all the systems and handoffs happening behind the scenes, but the customer doesn’t. They just know they told you they were interested and nobody responded. Every interaction like that teaches someone a little bit about what it’s going to be like to do business with you.

If you’re going to ask someone to raise their hand, you need to be ready to respond while their hand is still up.

DJ: If you were advising a CEO who has already invested heavily in AI but isn’t seeing measurable revenue gains, what are the first three workflow automation initiatives you would recommend implementing immediately?

Sandman: First, I’d look for where work is still stopping and waiting for a person. A lot of companies have used AI to make tasks faster without looking at the workflow around them. In lead generation, you might be able to summarize a lead instantly, but if someone still has to manually figure out where that lead goes next, you haven’t really solved the bigger problem.

Second, I’d look at where context is getting lost. Teams shouldn’t have to keep finding, recreating it or manually moving information from one system to another if it already exists somewhere in the business. If we’re using the lead journey as an example, the information someone gives marketing should follow them when they reach sales. They shouldn’t have to tell you the same thing twice.

Then I’d look at the places where a workflow can break down. AI can be really valuable for recognizing when something hasn’t happened and triggering the next step. If a high-intent lead hasn’t heard from anyone, for example, that shouldn’t depend on somebody noticing it and remembering to follow up.

That’s where I’d start as a CEO because these are problems almost every business needs to address in some form. Find where work is waiting, where context is getting lost and where things are falling through the cracks so you can better understand where AI can remove that friction. Automation works best when it’s accelerating a workflow you’ve intentionally designed, rather than being expected to fix the process on its own.

DJ: Meta and other technology giants are spending tens of billions of dollars on AI infrastructure, yet your research suggests many organizations have not automated even basic lead-routing processes. Is there a danger that companies are chasing the most visible AI innovations while overlooking the operational improvements that actually drive revenue?

Sandman: I do think there’s a danger, and the amount of attention on the newest AI capabilities can make it easy to overlook some of the less glamorous ways the technology can create value for customers.

For example, 70% of marketing teams still spend six or more hours every week manually managing inbound leads. These are companies that are already using AI, so clearly adoption by itself isn’t the outcome. If someone tells you they’re interested in your company, the technology should help make sure their intent and context can move through the organization quickly enough for your team to respond while it still matters.

There will always be excitement around the newest model or capability, and there should be. But I’m much more interested in what that technology allows us to do for the customer that we couldn’t do before. I want every investment to help me reach the customer faster, remove manual work so my team has more time to engage with customers, or create a more seamless experience across brand touchpoints. That’s how I determine if a technology investment is valuable.