A few months ago I mapped out the tech stack of a 150-person IT services firm. They had fourteen tools. A CRM for pipeline. A PSA for time tracking. A spreadsheet for pricing. A different spreadsheet for resource planning. A Word template library for proposals. An invoicing system. A project management tool that half the company used and the other half ignored. A BI tool that pulled from three of these systems but not the other eleven. And a handful of others I’m probably forgetting.
Fourteen tools. Zero shared data model. And a full-time person whose job — I’m not exaggerating — was to copy data between systems so that leadership could get a weekly report that was only sort-of accurate.
This isn’t unusual. It’s the norm.
The Point Solution Problem
The professional services industry has been buying software the same way for two decades: identify a problem, buy a tool for that problem, repeat. Need to track time? Buy a PSA. Need to manage the pipeline? Buy a CRM. Need to create proposals? Buy a CPQ. Need to forecast revenue? Buy a BI tool.
Each of these tools solves its narrow problem reasonably well. The PSA tracks time. The CRM tracks deals. The CPQ generates quotes. In isolation, they work.
But services businesses don’t operate in isolation. An engagement starts as a deal in the CRM, gets scoped and priced in a spreadsheet, moves to the PSA for delivery, generates invoices in the billing system, and produces data that should inform the next deal — but doesn’t, because none of these systems talk to each other in any meaningful way.
The result is an archipelago. Islands of data, separated by oceans of manual processes, connected by rickety bridges of CSV exports and copy-paste.
What Gets Lost Between the Islands
The data that falls into the gaps between point solutions is often the most valuable data in the organization:
- Estimate-to-actual feedback. The pricing tool generates an estimate. The PSA tracks actuals. But nobody connects them systematically. So the same pricing mistakes get made over and over, because the learning loop is broken.
- Resource intelligence. The CRM knows what deals are coming. The PSA knows who’s available. But the two systems don’t share this information, so resource planning happens in meetings and spreadsheets instead of in real time.
- Engagement context. The proposal contained assumptions about scope and complexity. The delivery team may or may not have read the proposal. Even if they did, those assumptions aren’t tracked in the PSA — they live in a PDF attachment that nobody opens after kickoff.
- Margin visibility. You’d think a services firm would know its margin on every engagement. Most don’t — at least not until weeks after the engagement is over, when finance reconciles the numbers. By then it’s too late to do anything about it.
“Point solutions solve point problems. But the problems in professional services aren’t points — they’re connected systems. You need connected solutions.”
The Operating System Analogy
Think about what an operating system does for a computer. It doesn’t just run applications — it provides shared services that every application can use. A file system. Memory management. Networking. Input/output. Without the OS, every application would need to implement all of this independently. They wouldn’t be able to share data, coordinate resources, or communicate.
That’s exactly where professional services firms are today. Every tool implements its own data model, its own user management, its own reporting. They can’t share data easily. They can’t coordinate processes. And they definitely can’t learn from each other.
What services firms need isn’t another application. They need an operating system — a shared foundation that provides common services: a unified data model, a shared engagement lifecycle, a consistent view of resources, and a learning layer that connects what was estimated to what was delivered.
The Architecture Argument
From a systems perspective, the case for unification is straightforward. Let me walk through the architectural advantages:
Single data model
When your service catalog, pricing logic, resource data, and delivery tracking all share the same data model, queries that would require joining data from four different systems become trivial. “Show me every engagement we’ve delivered for financial services clients in the last two years, with estimated vs. actual margin” — that’s a simple query in a unified system. In a point-solution stack, it’s a multi-day data engineering project.
Transactional integrity
When a deal is won and transitions from sales to delivery, multiple things need to happen atomically: resources get allocated, the project gets created, the budget gets set, notifications go out. In a point-solution stack, these are separate operations in separate systems, and if one fails, you’re in an inconsistent state. In a unified system, it’s a single transaction. Either everything happens or nothing does.
Consistent permissions
One of the ugliest problems with multi-tool stacks is permission management. Each tool has its own user model, its own roles, its own access controls. A partner might have access to margin data in the billing system but not in the PSA. An engagement manager might see resource costs in one tool but not another. A unified system has one permission model. One source of truth for who can see what.
The learning layer
This is the part that excites me most as a technologist. When all engagement data — from initial scope through pricing, delivery, and completion — lives in one system, you can build learning loops that are impossible with disconnected tools. The system can observe that engagements scoped at X hours for Y type of work consistently take 1.3X hours. It can surface that a particular pricing configuration leads to margin erosion. It can recommend resource allocations based on what worked well in similar past engagements.
None of this is magic. It’s pattern matching on structured data. But it requires the data to be structured, and it requires the data to be in one place. Point solutions can’t do this.
Unpopular Opinion
Best-of-breed is a trap for services firms under 500 people.
I know this goes against the conventional wisdom. The standard advice is to pick the best tool for each job and integrate them. And for large enterprises with dedicated integration teams, that can work. But for the 50-to-500 person firm — which is the majority of professional services — the integration tax kills you. You spend more time and money connecting tools than you’d spend just using a unified platform that does 80% of what each point solution does.
The tools you pick might each be a 9/10 in their category. But the integration between them is a 3/10. And the intelligence you can extract from disconnected data is close to zero. I’d rather have a unified 8/10 with real data continuity than a collection of 9/10s that can’t communicate.
What “Unified” Doesn’t Mean
I want to be clear about what I’m not arguing for:
- Not a monolith. A unified platform doesn’t mean one giant application that does everything. It means a shared data model and shared services, with purpose-built interfaces for different functions. The sales team sees a sales-oriented view. The delivery team sees a delivery-oriented view. But they’re looking at the same underlying data.
- Not a replacement for everything. Your CRM isn’t going anywhere. Your accounting system isn’t going anywhere. A unified engagement platform sits between your sales system and your back-office, covering the operational core of how services get scoped, priced, sold, and delivered.
- Not a lock-in play. Good architecture means clean APIs and well-defined data boundaries. Data should flow in and out easily. If a system makes it hard to extract your data, that’s a red flag — regardless of whether it’s a point solution or a platform.
The Migration Reality
I won’t pretend moving from a point-solution stack to a unified platform is painless. It’s not. The data migration alone takes planning. The change management takes effort. And there’s always a period where the new system doesn’t do that one thing the old system did perfectly (even if the old system did everything else poorly).
But the cost of not migrating is compounding. Every month you operate on disconnected systems, you’re generating data that can’t be used, making decisions without full context, and paying the integration tax in human hours. The firms that make the transition early capture the compounding benefit of unified data sooner.
The firms that wait are building a bigger migration problem every quarter.
Bottom Line
Here’s what you can do tomorrow: count your tools. Not the ones on the IT asset list — the ones people actually use to get a deal from pipeline to delivered. Include the spreadsheets, the shared drives, the email chains that function as approval workflows. Then trace one engagement from first contact to final invoice and write down every time data moved between systems — manually or automatically. That number is your integration tax. If it’s more than three hand-offs, you’re paying more for disconnection than you would for a unified platform. That’s the arithmetic. It’s not complicated. It’s just usually invisible.
Ready to encode your services?
See how Servantium brings CPQ discipline to professional services.