Power Your Growth
hello@propensity.com
Elona Mortimer-Zhika and Sumner Vanderhoof joined ForgeX’s Davis Potter to discuss customer relationships, sales and marketing friction, and the practical work of building with AI.
Ask a marketing team why pipeline is falling short, and you may hear that sales isn’t following up. Ask sales, and you may hear that the leads aren’t worth pursuing.
Elona Mortimer-Zhika has heard both explanations. Her assessment: both sides are probably a little bit right.
That familiar disagreement became a useful starting point for an October 20, 2026 fireside chat hosted by Davis Potter of ForgeX. Joining him were Mortimer-Zhika, a chair, non-executive director, investor, and former CEO of IRIS Software Group, and Sumner Vanderhoof, CEO and founder of Propensity.
The conversation covered what AI is changing inside software companies and revenue teams. Much of it came back to a problem that predates the technology: departments working from different information, pursuing different targets, and losing sight of the customer somewhere between them.
Mortimer-Zhika brings the perspective of an executive who has led through growth, acquisitions, and international expansion. At IRIS Software Group, she oversaw a business that more than tripled in size during her tenure.
Asked what distinguishes successful companies, she started with customer understanding. Businesses need to know which problems they can solve, why customers would choose them, and what a successful outcome looks like from the buyer’s side.
If a customer is trying to save time, the conversation should be about time saved. If the priority is reducing compliance risk, that is the outcome the team needs to address. A list of product features does little to explain either.
That understanding also needs to be specific to the market. Drawing on her board experience in Europe, Mortimer-Zhika described buyers in Germany who expect depth, local references, and evidence that a supplier understands their regulatory environment. A successful approach in London may need substantial work before it succeeds in Munich or Milan.
The test continues after the contract is signed. Mortimer-Zhika said net revenue retention is one of the first numbers she examines when reviewing a software business. For her, it helps reveal whether a company understands its customers well enough to keep earning their business.
Winning an account gets the relationship started. Renewal and expansion show whether the company delivered on its promise.
Potter compared go-to-market teams to sections of an orchestra. Each has a distinct role, but they need to be reading the same music.
Mortimer-Zhika described the requirement as:
“One set of numbers and one view of the customer.”
That sounds straightforward until teams examine how they are measured. Marketing may be accountable for lead volume this quarter. Sales may be accountable for revenue from opportunities that began several quarters earlier. The definitions differ, the time horizons differ, and ownership of the handoff can be unclear.
Her advice was to agree on what a good account looks like, share a pipeline target, and build account plans together. Those decisions give teams something concrete to work from when disagreements arise.
This is consistent with ForgeX’s approach to account-based marketing, which focuses investment on defined accounts, engagement across the buying group, and account-level results.
For leaders, the practical question is whether sales and marketing would give the same answers about which accounts matter, what makes an opportunity worth pursuing, and who acts next.
Vanderhoof uses an engine analogy to describe the same challenge. Audience building, campaigns, qualification, scoring, and sales follow-up all contribute to a process that needs to keep moving.
When customers drop out, teams need enough visibility to understand where it happened and investigate why.
He sees AI helping businesses make that investigation part of their everyday operations. Campaign engagement can inform targeting. Sales feedback can improve qualification. Retention and expansion can help teams understand which customers were the best fit in the first place.
Vanderhoof described this as a recursive learning loop. Each cycle supplies information that can improve the next.
The distinction matters because a business can automate a great deal of work without learning much from it. A campaign may launch faster, yet the next campaign may still begin with the same assumptions and the same gaps in information.
His ambition for autonomous ABM is that results will feed back into decisions about audiences, messaging, channels, and sales activity. He explores that direction further in Propensity’s article on autonomous ABM and recursive learning loops.
For teams considering the idea, a useful place to start is the information they already collect. Does sales feedback reach the people planning campaigns? Do renewal outcomes influence the ideal customer profile? Can anyone trace a promising account from its first engagement through to its experience as a customer?
Those answers reveal where a learning loop could make a difference.
Mortimer-Zhika sees an immediate opportunity for AI in customer research and preparation.
She described research that once took days becoming possible in minutes. Knowledge that lived in someone’s head or an old presentation can be refreshed and made available to the next person speaking with the customer.
That could mean a salesperson arriving with a better understanding of an account, or a customer success team recognizing a concern before renewal discussions begin.
“That personal touch at scale used to be a contradiction.”
In her view, AI is making it more achievable. But the quality of the human interaction still matters. Knowing more about a customer is useful when it helps someone ask a better question, offer relevant advice, or respond with an understanding of the relationship.
Vanderhoof made a related point about the information supplied to AI systems. He described customers choosing relevant context from Salesforce, website activity, and product usage. People decide which information matters to their business and which can be left out.
That judgment affects what the system has to learn from.
Propensity AI applies this broader approach to buyer and customer journeys across acquisition, expansion, and retention. The opportunity discussed in the chat was to make each interaction more informed, with the customer’s history and circumstances available when they are needed.
Vanderhoof offered a specific example of how AI has changed product development at Propensity.
Customers can vote on features they want. He explained that when a request has enough support and is sufficiently well scoped, the company’s internal engineering system can build it and move it into a staging environment, subject to safeguards.
Human review is required before it goes into production.
That gives customers a more direct influence on product priorities while shortening the path from request to working software. It also makes the point at which a person must review the work explicit.
Vanderhoof described his approach to automation this way:
“The person is automating it. The people are guiding it.”
He also spoke about controlling the tools used in internal AI workflows. His concern was practical: the organization needs to understand how its systems work and be able to support them when the person who built a workflow is away.
Mortimer-Zhika’s advice to executive teams followed a similar line. Choose a business problem worth solving. Understand the data being used. Invest in people who are curious and willing to keep learning.
ForgeX’s AI in Marketing practice addresses those organizational questions through prioritized use cases, named owners, tool evaluation, and team training.
Looking toward 2027, Vanderhoof expects demand to grow for systems that support more of the customer journey and learn across its stages. He sees that as an important direction for the market.
Mortimer-Zhika’s contribution supplied a way to judge whether those systems are helping. Are customers getting what they need? Do teams share a useful view of the relationship? Can employees use the information available to make better decisions?
Leadership teams can begin with a single customer problem or a handoff that regularly breaks down. Agree on the outcome, assign ownership, and decide what evidence would show improvement. Then use that evidence to guide the next change.
The conversation was optimistic about what AI makes possible. It was equally clear about the work companies still have to do themselves: understand their customers, settle on shared goals, and take responsibility for the experience they deliver.
%201.svg){kind=icon}