Mistakes to Avoid When Building Your Finance Tech Stack

Building the Right Finance Tech Stack: From Legacy ERPs to Curated, Integrated Apps (Finance Focus Episode 3)

In this Finance Focus episode, Brendan Ritchie sits down with First Focus CEO Ross Sardi and Bolden founder Jarrod Morris to unpack a question many Australian businesses are asking: what does the right finance tech stack look like today? Drawing on hands-on experience across organisations in the 50–200 seat range, they explain why so many teams feel stuck with slow, manual finance processes, how to rebuild trust in data, and where modern best-in-class apps beat legacy ERPs. If your finance function is still wrestling with spreadsheets, re-keying the same data twice, or waiting days for reports, this conversation is for you.

Key takeaways

• Manual work and duplicated data are symptoms of a poorly integrated stack, not “just the way things are.”
• A single source of truth and single data entry point unlock faster, more confident decision-making.
• Legacy ERPs can slow teams down with heavy customisation, vendor dependency and limited agility.
• Best-in-class cloud apps, integrated via native connectors and APIs, now cover the mid-market well.
• Map processes first. Choose technology only after swim-lane roles, workflows and decisions are clear.
• Separate business operations (subledgers) from finance operations (general ledger) to improve ownership and accuracy.
• Design for adoption. Avoid “app bloat” by aligning tools to roles and minimising context-switching.
• Real-world stacks combine platforms like Xero + Unleashed + Lightspeed + Shopify + Causal for robust reporting.
• Start with fast wins, measure ROI in capacity created and decision speed, then scale gradually.
• AI and automation amplify value when your data foundation is clean, connected and trusted.

Watch the episode

Watch on YouTube

Why so many finance teams feel stuck

Most finance leaders don’t start by shopping for software. They start with a feeling. As Jarrod puts it, the common trigger is a simple statement from an executive or board member: “there has to be a better way.” That gut feel usually reflects deeper issues, frustration with manual processes, long report turnaround times, and a general lack of trust in the numbers.

These symptoms often share the same root cause: a stack that doesn’t integrate. Data gets entered multiple times across systems. People become the “glue” between broken workflows. Reports arrive late, or not at all. And when business moves quickly, as it does in Australian services, retail and SaaS, slow tends to equal poor outcomes. In this environment, Finance must enable fast, confident decisions. You only get that if your tools are connected and your data is consistent.

Data trust hinges on a single source of truth

Trust is binary. Staff either trust the data or they don’t. When they don’t, everything slows down. The team asks for extracts instead of self-serving a dashboard. Leaders wait for reconciliations instead of acting today. The fix is not a new report—it’s a design principle: single source of truth and single point of data entry.

• Single source of truth: choose where each master record lives. Make that system authoritative.
• Single data entry: enter once, flow everywhere. Remove re-keying and copy-paste from the process.
• Role-based access: give each function live visibility to the information they need, without bottlenecks.

When people can see accurate, timely information without waiting, decision speed improves. In a market where pricing, costs and demand shift rapidly, that speed matters as much as accuracy.

The ERP headache—and where ERPs still fit

Many mid-market organisations still rely on a legacy ERP. The promise was one platform to run everything. The reality for many? Costly implementation, heavy customisation, vendor-controlled changes and slow reporting. Ross and Jarrod see the same pattern: “capable often means complex,” and complex slows you down.

That said, ERPs aren’t dead. They can still shine in specialised industries with deep, niche requirements and mature ecosystems. The lesson isn’t “never use an ERP.” It’s “use the right tool for the job,” and in the 50–200 seat bracket, modern cloud apps now cover the majority of needs with less friction and lower total cost of ownership.

From all-in-one to curated best-in-class apps

APIs, native integrations and accessible pricing have changed the equation. Instead of wrestling an ERP into shape, you can build a curated stack, best-in-class apps that each do one thing well, stitched together cleanly. This approach keeps you nimble. If a component no longer fits, swap it out without ripping up the whole stack.

The team notes the shift over the last few years: from ERPs trying to be everything, to dozens of hyperspecialised apps, to today’s more balanced middle ground. You still pick anchors, like your general ledger and PSA, but you round them out with targeted tools for inventory, POS, e-commerce, forecasting and consolidation.

Process before platform: the non-negotiable first step

The most consistent mistake in technology projects is starting with software. Bolden’s method is different: run a structured digital discovery first. Map the existing workflows in swim-lanes, assign roles, and define the decisions you need to support. Only then select the tools that best fit those processes.

• Clarify outcomes: which decisions must be faster or more accurate? Forecasts, purchasing, pricing, hiring?
• Map the workflow: who raises sales orders, who approves POs, who completes receivables?
• Identify single entry points: decide where each record is born and how it moves.
• Quantify the ROI: estimate hours saved, error reduction and decision velocity improvements.

This approach avoids “solution in search of a problem” projects and anchors technology spend to measurable outcomes and accountability.

Separating business operations and finance operations

A powerful framing from Jarrod is the distinction between subledgers and the general ledger. Business units can own their subledgers, sales orders, inventory movements, POS transactions, while Finance owns the GL. As long as data flows reliably into the GL, you get both ownership and integrity.

• Business units: accountable for accurate, timely subledger data via fit-for-purpose tools.
• Finance: accountable for the GL, consolidation, compliance and forward-looking analysis.
• Integration: automated, auditable data flows from subledgers to GL and to reporting layers.

This structure clarifies responsibilities, reduces handoffs, and improves data quality at the source.

Designing for adoption and avoiding app bloat

App bloat happens when every problem gets a new tool. The result is context-switching and confused staff. Brendan and Ross emphasise role-based design. Each team should “live” primarily in one core application, with supporting tools wrapped around it. Keep the everyday workflow simple and consistent; push complexity into integrations and automation.

• Minimise the number of apps any one role must use daily.
• Standardise navigation patterns and naming conventions across tools.
• Use SSO and consistent permissions so access feels seamless.
• Pair go-live with targeted training and quick-reference guides.

A real-world example stack

Consider an apparel group that has grown through acquisitions. They don’t trust outputs from their legacy inventory and POS platform. The solution is a curated stack anchored by a modern GL and integrated subledgers:

• General ledger: Xero as the finance backbone.
• Inventory: Unleashed for product, costing and stock movements.
• Point of sale: Lightspeed to run in-store transactions.
• E-commerce: Shopify to support online sales channels.
• Strategic finance & reporting: Causal to consolidate entities, scenario test and dashboard.

With native integrations handling most flows and APIs filling gaps, the business eliminates re-keying, lifts data trust, and gains live reporting. Finance can run scenarios quickly – what if we adjust price, expand a range, or change supplier terms and leaders can decide with confidence.

A practical blueprint for 50–200 seat Australian organisations

Every business is different, but a proven pattern is emerging in the Australian mid-market. Use this as a starting point, then adapt based on your workflows, industry and compliance needs.

• Anchor systems: pick a GL (e.g. Xero) and, if relevant, a PSA for services (e.g. ConnectWise).
• Operational subledgers: inventory (Unleashed), POS (Lightspeed), e-commerce (Shopify), subscriptions/billing (fit to your model).
• Planning & consolidation: a strategic finance layer (e.g. Causal) for budgeting, multi-entity rollup and scenarios.
• Integration fabric: prefer native connectors; supplement with iPaaS or lightweight workflow tools where needed.
• Analytics: centralise clean data in your reporting layer; standardise metrics and definitions.
• Security & identity: SSO, MFA and role-based access, aligned with Australian privacy requirements.

Implementation roadmap: phase for outcomes

Move quickly, but sequence work to manage risk and maximise learnings.

1. Discovery (2–4 weeks): stakeholder interviews, swim-lane mapping, define target decisions and KPIs. Document single-entry points and source-of-truth systems.
2. Architecture (1–2 weeks): select anchor apps, choose integrations (native vs API), design the data model and permission structure.
3. Pilot & quick wins (4–8 weeks): automate one or two high-friction processes (e.g. sales order to invoice, PO to bill matching). Prove value early.
4. Scale & adoption (ongoing): roll to adjacent workflows, retire spreadsheets, embed dashboards in team rituals.
5. Optimise & extend: add forecasting scenarios, connect additional entities, and build AI-assisted tasks once the data layer is clean.

Governance, controls and the cost conversation

Modern stacks spread cost across line items instead of one ERP licence. That can feel messy until you line spend up against outcomes. Keep governance tight and transparent.

• Ownership: appoint system owners for each app and an overall data steward in Finance.
• Change control: small backlog, fortnightly releases, and clear comms to affected teams.
• Cost tracking: tag subscription spend, measure hours saved and cycle-time reductions.
• Risk: apply least-privilege access, log integrations, and back up critical datasets.

When stakeholders can see time saved, error rates reduced and faster decisions, the value story becomes obvious. Compare that to the hidden cost of waiting days for answers.

Where AI and automation add real value

AI isn’t a silver bullet. It amplifies whatever data foundation you already have. In a curated stack with single entry and consistent master data, AI can help you enrich records, detect anomalies, or summarise trends. In a patchwork of duplicated data, AI will merely accelerate confusion.

• Use AI to flag exceptions in AR, AP and inventory rather than generate more noise.
• Automate reconciliations and allocations where rules are clear and auditable.
• Feed scenario modelling tools with clean, timely data to test pricing, headcount and capex assumptions.

Metrics that prove the stack is working

If you can’t measure the improvement, you won’t sustain it. Tie your tech investment to clear metrics the board will care about.

• Decision lead time: days to decision, now vs baseline.
• Manual touches per transaction: automated share of the workflow.
• Close time: days to month-end close and variance rates.
• Data trust: survey score from users on accuracy and accessibility.
• Capacity released: hours per month shifted from manual processing to analysis.
• Working capital impacts: DSO/DPO/DOH improvements attributable to better visibility.

Adoption is the multiplier

The next Finance Focus episode dives deeper into adoption, and for good reason. Even the best architecture fails without people using it daily. Design the system around roles, train to the workflow, and make the right way the easy way. When adoption is strong, data quality rises, reporting becomes self-service, and Finance can move from historian to strategic partner.

A straightforward action plan

• Run a short discovery to map current workflows and define target decisions.
• Nominate source-of-truth systems and single-entry points for each record type.
• Choose an anchor GL and 1–2 best-in-class subledgers that solve your biggest bottlenecks.
• Implement native integrations first; reserve custom work for true gaps.
• Launch with one team, prove the win, then scale to adjacent processes.
• Embed dashboards in weekly rituals so insights drive behaviour.

Do this, and you’ll feel the shift quickly. The finance team spends less time chasing data and more time advising the business. Leaders stop waiting for reports and start asking better questions. Most importantly, you create a stack that moves at the speed of your market.

Wrap-up

Episode 3 makes the case for a curated, integrated finance stack designed around process, not vendor promises. For Australian organisations in the mid-market, the tools now exist to do this well, without the weight of a legacy ERP. Start with process mapping, set a single source of truth, choose best-in-class apps, and measure value in decision speed, accuracy and capacity. Then keep iterating. When you’re ready to tackle adoption, the multiplier effect, join us for Episode 4.

Follow First Focus

LinkedIn: First Focus IT

Facebook: First Focus IT

Instagram: @firstfocusit