3 Founders Reshaping the Marketing Funnel with AI

Ed Lee, a product-led growth leader at Adobe, says that just five years ago, he and his team spent most of their time growing sales by optimizing keywords and “target parameters” on Facebook. But with the meteoric rise of artificial intelligence today, the team finds it much better for business to defer to the social network’s algorithm-based recommendation engine. That’s freed up time for Lee and his team to focus on the human side of marketing.

“I spend my time working with the creative and product teams,” Lee says. “We can take the time to really optimize an ad’s messaging and visuals. And this has the greatest impact on lowering costs and improving the bottom line.”

Lee’s case is demonstrative of the marketing industry as a whole, which most believe is undergoing a revolution in standard practices and operations due to AI. According to a report published in the Wall Street Journal, most marketers believe AI will, in the end, make their work more efficient, freeing their time for higher order endeavors. The technology is now at the fingertips of everyone within an organization, from analysts to customer service to project managers, helping them reshape how they work and what they work on.

A focal point for this change is where Lee sits in the online sales process—where data-driven decisions are fundamental. As a performance marketer, Lee works to optimize the marketing experience up and down the “funnel,” the cascade of steps that triage sales leads from interested buyers to purchasers and then, hopefully, to return or expanded customers.

In some organizations, AI tools are being deployed for the entirety of the funnel. In others, where perhaps the product is more nuanced or expensive, AI helps match potential buyers to the right salesperson. All of these tools have appeared—in a business sense—”overnight,” which means that, like any innovation, professionals are having to find ways to integrate new tools into their processes.

Still, the addition of AI tools has left many in the marketing industry wary of the old saw that “the robots are coming for their jobs.” So far, this has only really played out where new technology usually takes jobs: at the most entry-level work, where the task is repetitive or menial. But despite the rise of text-generating tools like OpenAI’s ChatGPT or image-generating tools like Dall-E or Midjourney, Lee does not believe copywriters and graphic designers have to worry about their jobs. They simply move more into the role of editor or curator, using the engines to create “seed” content with a few basic keywords rather than spending hours brainstorming.

“We aren’t cold-starting meetings anymore,” Lee says. “We have warm leads. We come into the room prepared with concepts in mind—before we engage a team. Things I would have spent four hours on, I now spend fifteen minutes. That’s been the paradigm shift.”

The Beginning of an Evolution

The marketing tech stack is undergoing a rapid evolution. This is driven in part by many disparate factors: changing consumer expectations; privacy regulations; infrastructure consolidation; the continued rise of omnichannel; and the infusion of artificial intelligence into every layer of the stack from automation and personalization to sales and analytics.

While GenAI is all the rage in streamlining marketing tasks, its older, slightly less-buzzier cousin, machine learning, continues to also deliver tangible advancements. We spoke to the CEOs of three Dell Technologies Capital-backed companies, Aidaptive, Zingly, and Pecan, to understand how their platforms are delivering transformative tools to marketing professionals right now.

Aidaptive: Better Retail Sales through Personalization

A recent report from Deloitte notes that one of AI’s key applications is creating dynamic and personalized customer experiences. By parsing together huge amounts of data into specific insights and actions, in theory, these applications can provide each customer with a tailored experience and targeted product choices. Until now, however, that level of customization was only available to the biggest platforms with the most data to use.

Aidaptive is innovating in this space, creating customized shopping experiences on the fly for mid- to large-sized retailers alike. Before founding the company, CEO Rakesh Yadav spent fourteen years building ad products at Google. Now, he and his team have set out to “democratize” access to applied machine learning (ML) with the same level of sophistication that the biggest tech platforms use. They’re accomplishing this by using ML to create specific customer groupings and hierarchies—known more technically as “taxonomy graphs.” These graphs allow Aidaptive’s algorithms to shape every detail of the shopping experience for individual customers.

“This is critical in recommendation engines and search,” Yadav says. “We also have a separate

class of models which do event or intent understanding on the consumer side based on what they’ve searched. From there, we build out more specific sequence models.” Sequence models take into account dependencies in order or context and are used to make much more accurate predictions.

Yadav says one challenge for “etailers” is that about ninety-five percent of visitors to direct-to-consumer retail sites are not logged in or are first-time visitors. Still, those customers are producing valuable data from browsing patterns, geolocation information, device types, and search terms used. Aidaptive can then parlay that data into defining a taste model for that individual new shopper. From there, the platform will predict and reveal the three to five products they’re most likely to buy.

Aidaptive’s influence over a customer’s experience doesn’t stop when they bounce from a website. It can inform the content, timing, and offers sent through other marketing channels including targeted ads, emails and SMS-based marketing.

Yadav says, “This level of personalization improves click-through rates, average order value, and LTV performance. Really, AI can improve every sales metric for retailers today. And we’re just in the nascent phase here. There’s an insane amount of innovation happening daily.”

Zingly: Increasing Sales Productivity

We’ve all had the decidedly unfulfilling experience of interacting with generation 1.0 of chatbots. You’re invited to ask a question about the company or a product and in return, you get a rote response that’s probably not relevant followed by a “Did that answer your question?”

No. It often did not.

Websites are 24/7 store fronts. For SaaS companies, a website is also the top of the sales funnel. It’s often the medium that’ll entice a customer to interact more deeply or to instead bounce to a competitor. On the other side of the website is a team of sales development representatives (SDRs) eager to receive, score, and route new leads.

This can be tricky for quickly growing software companies. SDRs tend to have fewer years of experience, can have a high rate of turnover, and are often run as lean teams. And yet, once that contact form is filled out, it’s the SDRs that are first in line to respond. This is the perfect opportunity for AI to have impact says Gaurav Passi, founder and CEO of Zingly.ai.

“Before now,” Passi shared, “much of the automation being used on websites or in sales and customer support came at the expense of the end user experience. With GenAI, we can create more useful interactions on both sides of the equation.”

Zingly’s model ingests a company’s customer and product documentation from across myriad sources. It then layers in interactions between customers and sales or support reps to better understand questions customers had and what about the offerings at hand appealed to them.

From there, Zingly’s platform informs every interaction. As a potential customer browses a site, Zingly is formulating a lead score. A visitor deemed to have “high intent” can be funneled directly to a senior sales rep. One with lower intent or more questions can be connected to a more junior SDR. But that SDR is also fortified with Zingly.

“Once AI is integrated, it’s essentially ‘born’ smart,” Passi says. “Rather than a representative spending fifteen minutes researching and thinking through the wording for a response to an inquiry, Zingly cues up the text and lets the rep customize it.”

The goal once again isn’t to replace sales or support but rather to aid in prioritizing leads and help customer or potential customers get the answers they need through AI-supported human interactions.

In fact, Passi cautions sales leaders against cutting staff in favor of AI investment just yet. Especially while the industry is still trying to figure out pricing models for AI platforms. He suggests:

1. Start by leveraging AI and your team together. How can you do more with the team you have using AI? Or how can you use AI to better understand individual performance?
2. Spend a good amount of time deciding which metrics to track right out of the gate and determining what “good” or “great” look like. Start small and then expand your use cases.
3. Dive into the technology. Learn about how it works and more specifically, how it works with your data.

Pecan: Smarter Marketing Spend Planning

Imagine being able to understand how a marketing campaign will perform before committing spend against it. And further, imagine being able to do that across myriad types of campaigns and channels in a matter of minutes without involving a whole data science team.

While Aidaptive focuses on bespoke shopping experiences and Zingly streamlines B2B sales interactions, Pecan.ai uses ML to help marketers of all stripes get surgically smart in the forward-looking planning process.

The co-founders of Pecan CEO Zohar Bronfman and CTO Noam Brezis have three PhDs between them. They came upon the idea for the company back in their academic days when they entered a data science competition. In that competition, they built an algorithm with the goal of making customer behavior predictions for a bank. Bronfman told The Globes for a profile piece on the company, “We realized there was a Miguel in Madrid who didn’t know yet that he was going to take out a mortgage, while we and our algorithm knew it before he did.”

Armed with Pecan for forecasting, marketers can simulate campaign outcomes to better understand where to spend their energy, dollars, and discount offers. They can avoid offering discounts to customers who would have bought anyway or spending against customers who’d hard churned and had an exceedingly low intention of buying again. Pecan can also help marketers better inform ad platforms by illuminating the strongest predictive behaviors in potential new customers.

This level of predictability in outcomes is something that has become increasingly important as budgets have slimmed down. Every marketing dollar needs to perform.

“If you need to cut ten percent of your acquisition budget, you can’t just cut it across the board and hope for the best. You have to know where exactly to cut and target for the best results,” says Bronfman. And he added that good news is, most companies have the staff and data to get these results with Pecan.

The innovation here isn’t just the specificity in customer spend forecasts but how the marketers access those predictions. In the past, deeply mining a company’s data for these types of insights might take a data scientist several days to formulate the right queries and then analyze the response to make a single prediction. With broad data source integrations and ML onboard, Pecan cuts out the data science middlefolks and gives marketers a direct line to the predictions they need to plan campaigns and deliver against sales goals.

In a time when data professionals are scarce and come at significant premiums, Pecan’s approach is yet another example of how AI continues to be a complement – not a replacement – for marketing and sales professionals.

AI’s Continual Evolution

It will be these sorts of companies who can distill the under-the-hood complexities of AI into simple tools that bring in the new wave of AI marketing tools.

It will also be the marketing executives and managers who are dynamic. The resounding message, when looking across the AI landscape, is that the era of buying a tool and letting the software run for a decade is long over. To stay ahead of change, a team must remain curious and willing to reconsider their role as the tools evolve. While the media can debate “how much” AI is changing work, the one constant—for a while, if not forever—will be rapid innovation.

“When calculators came out, everybody said, ‘Hey, we’re going to start losing our mental acuity,’” Passi at Zingly points out. “But people realized that you can leverage calculators and use your brain somewhere else. The fear went away, and embracement started to happen.”

“It’s going to be very similar in the next five years for AI in marketing. The fear is going to be gone and everyone will be implementing it.”