NowVertical Group: AI begins with the plumbing

A gym member can disappear twice. First, they stop turning up. Later, they cancel. Between those two events lies a small commercial opportunity: a business still has time to do something useful. The difficulty is knowing whom to call, and why.

That interval helps explain NowVertical Group better than a parade of AI vocabulary. In its work for Nuffield Health, the company gathered membership details, service usage and customer interactions, built a churn prediction model, and gave retention teams dashboards identifying members at risk. Python and an AWS-hosted database were part of the machinery. People supplied the intervention.

The published Nuffield case study reports £700,000 in revenue savings over three months. That is a company-reported outcome for one program, rather than a promise for every gym. Still, the sequence is instructive. A prediction becomes valuable when it arrives early enough, in a form someone understands, with an action attached. Otherwise, it is an interesting number watching a customer leave.

01 / The work beneath the AI pitch

NowVertical occupies the space between buying technology and making it useful. Its work includes data engineering, cloud migration, governance, business intelligence, customer analytics and AI implementation. Marketing teams want to spend more intelligently. Finance teams want coherent reporting. Customer teams want to spot cancellations or recommend something worth buying. The underlying information often lives in different systems, speaking different dialects.

Consider its Telefónica modernization case. The starting problems included an on-premise Hadoop cluster, costly Teradata licensing and difficulty deploying machine learning. NowVertical describes building a Google Cloud data lake, introducing serverless data processing and creating frameworks for putting models into production. The attraction is easier scaling and a more usable foundation for analysis. Moving information is only part of the assignment; preserving its meaning is the other.

This is a business of expertise as well as software. NowVertical uses its own tools alongside platforms supplied by larger technology companies. It can therefore sell implementation work, continuing services and software, while some revenue comes from reselling third-party technology. A customer is buying a combination of capability and delivery. That combination matters when the impressive demonstration meets the less impressive database.

02 / An empty marketplace has plenty of data

Informa offers another version of the same problem. As physical events were replaced or supplemented by virtual events and online marketplaces, visitors struggled to discover relevant products and content. Some spent their time scrolling without finding anything they liked. A business can have a crowded catalogue and still offer a lonely shopping experience.

NowVertical built recommendations across users, companies, products and content for Informa’s Marketplace 365 services. Its published implementation combines content similarities, collaborative filtering and behavior analysis. It also uses search and language-processing techniques to connect people with relevant material. The aim is recognizably commercial: better discovery, more engagement and more opportunities to sell.

The distinction is subtle but useful. A recommendation system needs a definition of relevance, and that definition changes with the setting. An exhibitor browsing an industry event has different intentions from a subscriber considering renewal. Sector knowledge helps determine what a model should optimize and what a team should do with its answer. NowVertical’s historical language of “vertical intelligence” reflects that emphasis on the business context.

03 / Buying the expertise was the easy part

The company’s own structure has undergone a similar joining-up exercise. Founded in 2017, NowVertical began trading on the TSX Venture Exchange in July 2021. Acquisitions brought in software, specialists and customer relationships. Acrotrend and Smartlytics, two UK-based analytics businesses, joined in January 2023. Their announced combined upfront gross consideration was US$6.45 million, with additional earn-out terms.

Acrotrend supplied a particularly useful strand of customer analytics experience. Its founders, Sandeep Mendiratta and Shailesh Mallya, had met at Deloitte in Mumbai. Their business developed around customer data and long-standing enterprise relationships. That history helps explain why NowVertical’s portfolio contains both technical capabilities and practical knowledge of retention, marketing and customer lifetime value.

Ownership, however, does not automatically produce collaboration. NowVertical’s February 2026 integration announcement explicitly identifies the earlier difficulty: fragmented brands limited enterprise visibility and made larger account engagements harder to scale. The response was “One Brand, One Business,” bringing nine acquisitions under a common identity and aligning delivery, accounts and partnerships.

“Unifying our brand marks a clear transition from integration to value realization.”Sandeep Mendiratta, then CEO · February 2026

The lesson for another acquisitive business is quite concrete. A customer relationship in one country should make relevant expertise in another country easier to buy. Integration has to reach the sales conversation and the delivery team. Changing the sign above the door is the pleasant part.

04 / The price tag belongs in the story

NowVertical’s June 2026 announcement puts a number on one current assignment: a US$1.2 million modernization mandate for an unnamed global media and telecoms group. The scope covers several hundred business-critical reporting views across customer, finance, billing and supply-chain functions, moved onto Google Cloud. That is a particular enterprise contract, not a price list.

The financing also deserves careful arithmetic. In May 2025, HSBC provided access to US$18 million in credit facilities: a US$6 million term loan and a US$12 million revolving facility. Another US$8 million could become available with lender approval. The announcement put the initial blended interest rate at approximately 7.25%. Capital helped consolidate existing debt as well as support working capital and future acquisitions.

The operating record remains mixed. Reported 2025 revenue was US$37.3 million, down from a comparable US$39.4 million in 2024. The company attributed the decrease chiefly to Argentine currency effects, lower reseller revenue and restructuring in Chile. Revenue quality also changed: the top 30 strategic accounts produced US$25 million, representing 67% of annual sales. Deeper relationships were becoming more central to the business.

By Q2 2026, revenue was US$9.7 million, 18% above the corresponding quarter, with a 50% gross margin. Management identified the weaker prior-year currency and reseller comparison as contributing factors. Those details keep the growth figure in proportion. Co-founder Andre Garber became interim CEO in June, and Philip Jones was appointed CFO in August.

05 / Give the assistant something dependable to read

NowVertical’s product propositions make the sequence more explicit. GlassBox addresses undocumented legacy data and transformation logic. It promises to reveal lineage and help convert workloads to BigQuery and Google Cloud, with human review and reconciliation. Its appeal is easy to understand: a report can be business-critical long after everyone has forgotten how it was assembled.

NowUnlock AI, launched in March 2026, organizes enterprise adoption into Impact, Momentum and Scale pathways. The stated progression runs from readiness and an initial viable agent to operational deployment and broader governance. Meanwhile, NowHub-Finance combines financial data and reporting with a conversational agent built using Azure OpenAI. The agent introduced in December 2024 was offered to existing users at no additional charge.

These offerings sit within a partner network. NowVertical established a Microsoft Center of Excellence in 2024 and gained Google Cloud’s Generative AI specialization in 2025. In September 2026, it expanded a dedicated Google Cloud practice into Brazil alongside its established Qlik business. The partners provide platforms and routes to market; NowVertical supplies the implementation around a customer’s circumstances.

For buyers, the alternatives include a large systems integrator, a specialist data consultancy or an internal team. NowVertical’s proposition rests on combining industry experience, proprietary accelerators and international delivery. That combination is most relevant when the problem crosses systems and departments. A small business with clean records and a simple reporting need may have little reason to buy such an engagement.

06 / Borrow the sequence

The transferable idea is to start with a decision. Who needs to make it? What would count as an improvement? Which records are required? In the Economist retention case, NowVertical’s work connected billing, CRM and engagement data, then developed renewal metrics and offers for different customer cohorts. Workshops involved retention and finance teams before the reporting and recommendations were built.

That order places a useful burden on an AI project: somebody must own the result. If the records are unreliable, the prediction will be suspect. If a team cannot intervene, identifying an opportunity changes little. If every business unit defines the same metric differently, a faster answer can simply accelerate an argument.

NowVertical’s company and its customer work arrive at the same question. Can separate parts be made to cooperate well enough to produce a useful outcome? The model, the migration and the brand each have a role. The test comes when someone has to decide what to do next.