From billing complexity to control with ZoneBilling and Zoe

Summary

Complex billing models create challenges across the quote-to-cash lifecycle, particularly when finance teams must manage usage-based pricing, contract amendments, revenue recognition, renewals, and enterprise account structures. Even when the underlying data is available in NetSuite, answering operational and strategic questions can still require saved searches, spreadsheets, reconciliations, or support from technical teams.

This ZoneUnplugged session demonstrates how ZoneBilling manages complex billing and revenue workflows natively in NetSuite, while Zone Control Center and Zoe provide conversational access to billing insights. The speakers explore practical use cases for billing teams, controllers, and CFOs, compare Zoe with general-purpose AI tools, and preview a forward-looking roadmap that includes customization, anomaly detection, scheduled work, and human-supervised actions.

Transcript

Bringing complex billing into NetSuite

Haley Ashe

Hello, everyone, and welcome to ZoneUnplugged: From Billing Complexity to Control.

I’m Haley Ashe, Director of Customer Marketing at Zone & Co. I help bring our customers resources such as the ZoneUnplugged webinar series, our monthly customer newsletter, The Zone, and more.

I’ll be joined today by Becca Byers, Patrick Norton, and Anthony Dixon.

Becca Byers

I’m Becca Byers, Director of Product Marketing at Zone & Co.

One of my favorite parts of my role is working directly with customers as they use our products, find new ways to succeed with them, and pilot new capabilities. Today, we’ll discuss several ways ZoneBilling is expanding.

Patrick Norton

I’m Patrick Norton, the Global Billing Subject Matter Expert at Zone & Co. I’ve worked with ZoneBilling for approximately six years and have spent many years in the NetSuite ecosystem.

My background is in accounting and finance, and I’m looking forward to serving as a resource throughout the session.

Anthony Dixon

I’m Anthony Dixon, Platform Manager for AI at Zone & Co. I joined the company to help introduce AI capabilities across our products.

Today, I’ll introduce Zoe and explain how we are bringing AI into ZoneBilling.

Haley Ashe

Zone & Co provides a NetSuite-native software platform designed to improve operating margins.

Our solutions extend NetSuite across processes including quote to cash, procure to pay, treasury, and reporting. Inside NetSuite, we help finance teams streamline billing, payroll, AP automation, cash reconciliation, and reporting without requiring them to leave the ERP or manage separate integrations.

We are also adding AI capabilities intended to accelerate those processes.

Today, we’ll look at ZoneBilling, walk through a live demonstration, explore the AI roadmap, and answer audience questions.

Becca Byers

ZoneBilling is designed for subscription and complex billing inside NetSuite. It was originally built in response to NetSuite customer needs, and it has continued to evolve around the different ways customers structure and manage billing.

It plays an important role in the quote-to-cash lifecycle because it can take information from a CRM, CPQ system, proprietary platform, or usage source and translate that data into billing outcomes.

Customers use ZoneBilling to support goals such as shortening the close and reducing revenue leakage. Some have also used the platform while preparing for major company milestones, including acquisitions and public offerings.

Patrick will now explain how ZoneBilling supports the broader billing process and demonstrate the product.

Managing the full quote-to-cash lifecycle

Patrick Norton

ZoneBilling supports the full quote-to-cash process.

On the input side, a customer may use a CRM, CPQ platform, or another system where the sales process takes place. ZoneBilling can ingest the relevant data from those systems and automate the downstream billing and revenue recognition processes.

The subscription contract serves as the control center for the customer agreement. It captures the complete structure of what was sold, including SaaS licenses, professional services, hardware, usage commitments, and other contract components.

That contract data becomes the foundation for everything that follows.

ZoneBilling can also manage complex account hierarchies. This is particularly important for enterprise customers with parent-child structures, multiple bill-to entities, consolidated invoices, or billing that must be divided across several entities.

Rating and charge calculations are often where billing complexity becomes most difficult. ZoneBilling can support prepaid balances, tiered pricing, data mediation, minimum commitments, usage-based models, overages, fixed fees, one-time hardware charges, and upfront implementation fees within a single engine.

During my time as a solution architect, it was uncommon to encounter a scenario that ZoneBilling could not calculate.

The platform then automates invoice generation and presentation. Teams can group charges, separate them, display subtotals, and attach detailed usage files when customers need additional transparency.

Contract amendments are another area where many billing systems struggle. Midterm swaps, upsells, downsells, cancellations, rate changes, billing-address changes, and segmentation adjustments can disrupt a billing schedule.

ZoneBilling handles those amendments without requiring teams to rebuild the schedule manually or create separate change-order records.

Those changes also flow into revenue recognition. ZoneBilling can automate merge rules, recalculate total contract value when the agreement changes, and split or combine revenue elements. This helps keep revenue recognition aligned with billing without requiring a separate spreadsheet-driven process.

If renewals are not managed through the CRM or CPQ system, ZoneBilling can support those as well.

The complete process remains inside NetSuite. That reduces handoffs and reconciliation because the data is managed within the same financial system.

Inside the ZoneBilling subscription record

Patrick Norton

I’ll begin the demonstration inside NetSuite, because that is where ZoneBilling operates.

From the ZoneBilling subscription list, I can open an individual subscription record. This record functions as the contract’s control center.

At the top, I can see the customer, contract dates, billing cadence, and other key information.

The subscription items appear below. These may include one-time hardware fees, fixed recurring subscription licenses, and usage-based charges.

The record also displays forecasted charges, customer hierarchy information, bill-to overrides, usage details, revenue recognition data, renewal information, and related transactions such as invoices and sales orders.

This provides a consolidated view of the contract and everything connected to it.

From there, I can move into the billing dashboard within Zone Control Center.

The dashboard allows finance teams to interact with their billing operation from one view. It can sit alongside other financial processes, including procure-to-pay, treasury, and reporting workflows.

The billing funnel shows where each subscription currently stands in the process. It also identifies contracts approaching renewal so that the team can act before revenue is at risk.

Users can drill into individual values, manage approvals, and update relevant processes from the same interface.

That is what it looks like to run ZoneBilling natively inside NetSuite: the data and workflows remain in one place rather than being managed through disconnected spreadsheets.

Asking Zoe questions about billing data

Patrick Norton

Zoe also lives within the dashboard.

Finance teams already have large amounts of data, including invoices, payments, customer balances, and revenue schedules. The challenge is often getting useful answers from that information.

Traditionally, that may require building a report, waiting for assistance from IT, or exporting data into a spreadsheet. Zoe changes that experience by allowing users to ask questions in plain language.

Zoe is built specifically for finance operations within Zone products. It understands the relevant data, workflows, and role-based context.

The information a billing specialist needs at the beginning of the day may be very different from what a controller needs before the month-end close. Both differ from what a CFO needs before a board meeting.

We can look at examples for each role.

For a billing specialist, I might ask Zoe for additional insights into expiring transactions.

Instead of searching through saved searches or spreadsheets, Zoe returns a prioritized analysis of renewals that may be at risk over the next 90 days.

It can show the total value at risk, identify important patterns, and offer recommendations. In this demonstration environment, it also recognized that several records used generic demo names and suggested that the data could be cleaned up.

The underlying information can be exported into an external document for distribution or further analysis.

Zoe does not only display data. It highlights what matters, what appears urgent, and what the user may need to do next.

A controller may be more focused on accuracy, close readiness, and the audit trail.

For example, I can ask for a summary of credits issued during the past year. Zoe returns the total credits, supporting currency information, and a more detailed analysis.

It can also identify credits that remain pending and have not yet been reflected in the financial results. That gives the controller real-time visibility into potential liabilities without requiring a separate reconciliation spreadsheet.

A CFO may ask a broader strategic question, such as which customers represent the largest revenue concentration and whether those customers have overdue balances.

In this example, Zoe immediately identified that one customer represented almost 24% of revenue. It then provided a breakdown of concentration, credit risk, and recommended areas for review.

Together, Zone Control Center and Zoe provide different levels of the finance organization with information suited to their responsibilities.

The controller can understand credit exposure. The CFO can assess concentration risk. The billing team can see which accounts require action.

Instead of spending time finding data, finance teams can spend more time acting on it.

Where Zoe fits alongside other AI tools

Haley Ashe

Thank you, Patrick.

We asked the audience which AI tools they currently use alongside NetSuite.

Becca Byers

More than half of the respondents indicated that they were using Claude through the NetSuite Model Context Protocol.

That creates an important opportunity to explain where a general-purpose AI tool may be useful and where a finance-specific tool such as Zoe may be more appropriate.

Historically, many accounting principles have been difficult to reconcile with generative AI.

With a general-purpose AI tool, the organization must carefully create its own security model, guardrails, and instructions so that the system understands the accounting and business context.

Zoe is designed to provide those controls out of the box. It respects user permissions and regional data-sovereignty requirements, and customer data is not used to train the underlying models.

Reliability is another difference. General-purpose AI may hallucinate or produce an answer without a dependable connection to the underlying records.

Zoe uses deterministic data retrieval to pull information directly from billing records. It is also designed to provide verifiable answers and show the data behind the response.

Consistency is equally important. With a generic AI tool, different users may receive different results from similar prompts.

During the Zoe pilot, customers reinforced the importance of ensuring that everyone calculates metrics such as monthly recurring revenue and annual recurring revenue in the same way.

That need is shaping the short-term roadmap, including capabilities that allow organizations to define consistent calculations and have Zoe use them across the finance team.

We conducted the pilot because we want customers to receive capabilities that have been thoroughly tested in real billing environments.

The pilot also helped clarify how Zoe fits alongside other AI tools.

Zoe is intended to serve as a subject-matter agent for ZoneBilling. It has billing-specific context and understands the custom records and structures used inside the product.

Ask Oracle, which is being introduced with NetSuite Next, may be particularly useful for general navigation, help content, and support with NetSuite configuration.

Claude may be useful for general-purpose tasks and external calculations. However, organizations should carefully monitor token consumption and associated costs. One pilot participant reported that a single query through Claude cost approximately $50.

Zoe is designed with finance teams and cost control in mind. Initial usage is included with ZoneBilling rather than billed unexpectedly in arrears.

Our recommendation is to use Zoe when the question involves ZoneBilling data, including subscriptions, charges, deferred revenue, and approvals.

Claude may remain appropriate for general-purpose work outside NetSuite, while Ask Oracle may be useful for broader NetSuite navigation and assistance.

One pilot customer described Zoe as more likely to provide the correct answer the first time and more consistent than Claude when prompted correctly. The customer also noted that Zoe included more guardrails, while Claude could move beyond the intended scope.

The goal is to provide an AI capability that supports billing operations, controllers, and CFOs while keeping NetSuite as the source of financial control.

The Zoe roadmap

Anthony Dixon

You have now seen Zoe operating within ZoneBilling, but there is more planned.

The roadmap is organized around several priorities.

The first is addressing specific pain points in billing processes.

One area is anomaly detection. We are evaluating different points in the billing workflow where Zoe can surface information that helps teams identify and resolve problems.

During the pilot, one customer needed a better way to review monthly invoices and identify records that did not look correct.

We helped the customer establish a workflow in which Zoe compared the current month’s invoices with the prior month and highlighted invoices that changed significantly, either by value or by percentage. The finance team could then investigate those differences and understand their cause.

That is one example of the specific billing pain points we intend to address.

As Zoe expands through ZoneBilling and into other Zone applications, we will continue identifying similar use cases.

The second priority is making Zoe adaptable to each organization’s processes.

AI has what is sometimes described as a jagged frontier. It can perform extremely well in one situation and less effectively in another.

For AI to provide reliable value, it must fit the process in which it is being used.

Finance organizations follow common accounting principles, but every team also develops its own operational practices. Zoe needs to understand those practices or allow the organization to teach them to the system.

One capability intended to support that is the customization agent. It will allow authorized users to tailor Zoe to their finance team and apply those decisions coherently across the organization.

The third priority is repeatable work.

Finance teams have tasks that occur daily, monthly, and in the background. We want Zoe to support asynchronous and scheduled work so that a user can define a workflow once and have it run repeatedly.

The system could perform the analysis in the background, identify potential problems, and bring those results to the user at the appropriate time.

The roadmap therefore focuses on resolving defined pain points, understanding each finance team’s processes, and completing work at the times when the team needs it.

The capabilities I am about to describe are forward-looking and remain subject to change.

Subscription Intelligence and organizational customization

Anthony Dixon

The first release was Subscription Intelligence, which Patrick demonstrated.

At this stage, Zoe is read-only.

A user opens the chat and asks a question. Zoe uses available tools to retrieve the relevant information from the customer’s NetSuite instance and formulate an answer.

The user can download the supporting data into an Excel document and review the queries used to retrieve it. This makes it possible to understand where the answer came from.

The next area of development is customization.

Anyone who has worked with general-purpose AI knows that it can be capable but may require repeated instructions. A user may need to explain which information to use, which data to ignore, or how a calculation should be performed.

The customization agent is intended to allow NetSuite administrators or senior finance leaders to define those instructions once and apply them consistently across the team.

Consider annual recurring revenue or monthly recurring revenue. Organizations may calculate those metrics differently.

One company may include every product, while another may include only subscriptions and exclude one-time charges.

The customization agent is intended to let an authorized user define the organization’s preferred calculation. From that point forward, when any team member asks Zoe about ARR, the system would use the calculation established by the organization.

That creates consistency across users and over time.

The initial focus is on metrics, but the longer-term plan is to expand customization to other elements such as workflows and system notes.

In customization mode, an authorized user could select an existing metric or create a new one. Some common metrics would be provided as a starting point.

The user could then explain the required change in plain language. For example, the organization may want an ARR metric that includes only a particular product.

Zoe would review the customer’s NetSuite schema, propose the required changes, and show how the change affects the resulting data.

The authorized user could then publish that definition to the broader system. From that point forward, the entire organization would use the same metric.

Human-supervised actions and scheduled work

Anthony Dixon

The next major step is moving from read-only capabilities to actions.

The intention is for Zoe to complete certain tasks on the user’s behalf while keeping a human in the loop.

Examples may include creating subscriptions, finding errors, and performing mass updates.

If Zoe identifies an error and the user knows how it should be corrected, the user could instruct Zoe to address it.

Zoe would determine the proposed change and then return to the user for confirmation. The action would occur only after the user approves it.

This phase is also expected to include asynchronous and scheduled tasks.

For example, the invoice-variance workflow described earlier could run automatically each month. It would complete in the background and surface the relevant exceptions without requiring the user to launch the analysis manually each time.

Detecting billing anomalies before invoices are sent

Anthony Dixon

Another area of development is anomaly detection.

The goal is to identify potentially incorrect invoices between the rating process and the point when the invoice is sent to the customer.

The planned approach uses not only generative AI but also more traditional machine-learning methods.

We are working with TimeGPT-2, a foundation model designed for time-series analysis. Similar to how a large language model can interpret many types of language-based tasks, this type of model can adapt to different time-series patterns.

The anomaly-detection process would run in the background and surface unusual invoices for review.

Users would have a sensitivity control that could be adjusted to change how many potential anomalies the system identifies.

Over time, agents could also examine an anomaly before it reaches the user. They could review historical data and current activity, then suggest why the invoice appears unusual.

The user could determine that the change is expected or confirm that the system has found an error and decide how it should be corrected.

Expanding Zoe across Zone products

Anthony Dixon

Zoe is not intended to remain limited to ZoneBilling.

AP Intelligence is being developed for ZoneCapture and approvals customers, while Cash Intelligence is planned for ZoneReconcile customers.

The experience will be similar to the ZoneBilling demonstration. Instead of a billing dashboard, the user would access Zoe from the relevant capture, approvals, or reconciliation dashboard.

The broader value of using several Zone products is that customers could eventually ask questions about AP, billing, and cash within the same conversation.

The roadmap also includes a broader agent-based experience.

The concept is that users would be able to launch specialized agents as needed. Because those agents would use the organization’s customizations, they could complete workflows according to the finance team’s established practices.

Tasks could run while the user completes other work or on a predefined schedule.

A finance team might create a collection of month-end close tasks and have the agents run those tasks at the appropriate time. The agents would then surface the results.

The user could review those findings and initiate another group of agents to correct identified issues or complete approved actions.

This represents a larger planned expansion of Zoe. The current goal discussed in the session was to bring this experience to ZoneBilling in early 2027, although the timing and functionality remain forward-looking and may change.

Audience Q&A

Haley Ashe

How much of Zoe relies on third-party large language models, and how much is developed internally?

Anthony Dixon

We use third-party large language models because the leading frontier models perform extremely well across many measures.

However, those models still have the jagged frontier I mentioned earlier.

A useful analogy is to think of a general-purpose model as a very intelligent intern. It has a great deal of capability, but it needs the correct context and training.

We put significant effort into providing that context so that the system understands finance data and Zone products.

That contextual layer is what allows Zoe to operate as a specialized ZoneBilling expert rather than as a general-purpose chatbot.

Haley Ashe

How do you measure and track Zoe’s hallucination rate?

Anthony Dixon

We maintain an evaluation set that is used whenever we make a change to the system. That allows us to measure and track performance over time.

We also expose the data used to produce the answer. Pilot customers have found that transparency particularly helpful because they can verify the underlying information.

In addition, we use deterministic components and precompute certain calculations rather than relying on the language model to perform those calculations independently.

Because the system is dedicated to finance and billing, we can build specialized protections that would be difficult for a general-purpose tool covering many unrelated use cases.

Haley Ashe

What is the underlying model behind Zoe? Is it proprietary or built on a third-party model?

Anthony Dixon

We use several third-party large language models. We do not publicly identify individual providers because those providers may change over time.

The models are accessed through AWS, which provides additional protections around data security and tenant isolation.

It also allows us to direct model requests to specific regions to support regional data requirements. That level of regional control is not always available through a standard subscription purchased directly from a model provider.

Haley Ashe

What kinds of questions can users ask Zoe?

Anthony Dixon

Zoe can access the NetSuite data available to the user, so users can ask questions about that data.

However, the best results currently come from questions related to ZoneBilling, because that is the product context in which Zoe has been most specifically developed and tested.

As AP Intelligence and Cash Intelligence are introduced, the same approach will apply to data and workflows in those products.

Questions tied to Zone product functionality are therefore the strongest current use cases.

Haley Ashe

Will Zoe respect NetSuite permissions? Could a user see information that they should not be able to access?

Anthony Dixon

Zoe uses the NetSuite role under which the user is currently signed in.

A user cannot retrieve data that is not available to that role.

The same principle applies when a user changes roles. The conversation history is tied to the combination of the user and the active role, helping prevent information from crossing between roles with different permissions.

Haley Ashe

How much does Zoe cost?

Anthony Dixon

At the time of this session, the first 100 Zoe queries were included with ZoneBilling.

Usage beyond that amount followed a tiered model. The specific commercial structure may evolve as adoption expands.

Haley Ashe

Is Zoe available only for ZoneBilling?

Anthony Dixon

At the time of the session, Zoe was generally available through ZoneBilling.

ZoneCapture and approvals customers interested in participating in an early-access pilot could contact the team.

The roadmap discussed during the webinar included extending Zoe to ZoneCapture, approvals, and ZoneReconcile customers over the following months.

Haley Ashe

Can Zoe remember how an individual user works?

Anthony Dixon

Yes. Zoe currently includes two relevant forms of memory.

Within a single conversation, it retains the context of the earlier questions. A user could ask for a report and then follow up by asking Zoe to limit the same analysis to a particular month.

We have also introduced a user-level memory layer.

A user can tell Zoe about a personal working preference, such as which data source should be used for a recurring question. Zoe can then apply that preference for that individual user.

The customization agent is intended to expand that concept from personal preferences to organization-wide standards.

Organizational memory would be controlled by authorized administrators, allowing the finance team to maintain consistency while limiting who can define shared rules.

Customer resources

Haley Ashe

Zone customers have several resources available to help them get more value from their products.

ZoneUniversity is the hub for product training and education. It includes on-demand content for new users as well as more advanced material for experienced teams.

Zone Support is available through the customer support portal. Zone has also introduced an AI support chatbot to help customers find answers at any time.

The Zone knowledge base provides searchable articles, how-to guides, and resources based on customer use cases.

Customers can also subscribe to The Zone, our monthly newsletter featuring product updates, customer stories, upcoming events, webinar invitations, and other company news.

Finally, the Zone Insider program gives customers opportunities to help shape the future of finance technology while earning recognition and rewards.

Customers can choose how they participate, whether by sharing a testimonial, leaving a product review, responding to a survey, or contributing to a customer spotlight. Participation is flexible, and customers can select the activities that are most relevant to them.

Thank you to everyone who joined us for this session. We hope the discussion provided a clearer view of how ZoneBilling, Zone Control Center, and Zoe can help finance teams move from managing billing complexity to acting on trusted information.

‍